1 INFO-VAX	Wed, 12 May 2004	Volume 2004 : Issue 263       Contents:@ Re: %MOUNT-F-MAPHDRBAD, storage map header is bad; volume locked? Re: Alpha is phased out .... UltraSPARC V in the same way ..... ? Re: Alpha is phased out .... UltraSPARC V in the same way ..... ? Re: Alpha is phased out .... UltraSPARC V in the same way .....  Alpha/VMS news on The InquirerO Re: DCLER (DCL Enhancement Request) for DIRECTORY  /SORT=[CREATED,MODIFIED,...] P Re: DCLER (DCL Enhancement Request) for DIRECTORY /SORT=[CREATED,MODIFIED,...] /" decserver 900TM, PPP with security Re: DFO and SET FILE/NOMOVE  Re: EVE customizations Re: EVE customizations/ Re: Gigantic hobbyist cluster opportunity   :-) / Re: Gigantic hobbyist cluster opportunity   :-) / Re: Gigantic hobbyist cluster opportunity   :-)  Help required with CLD file : Re: HP to can HP/UX? Intel pulls out of IA-64 prematurely?: Re: HP to can HP/UX? Intel pulls out of IA-64 prematurely?: Re: HP to can HP/UX? Intel pulls out of IA-64 prematurely?: Re: HP to can HP/UX? Intel pulls out of IA-64 prematurely?: Re: HP to can HP/UX? Intel pulls out of IA-64 prematurely?: Re: HP to can HP/UX? Intel pulls out of IA-64 prematurely?: Re: HP to can HP/UX? Intel pulls out of IA-64 prematurely?: Re: HP to can HP/UX? Intel pulls out of IA-64 prematurely?K Re: HP World Magazine: For Business Continuity...the answer may be  OpenVMS J Re: HP World Magazine: For Business Continuity...the answer may be OpenVMSP Re: HP World Magazine: For Business Continuity...the answer may be OpenVMS OpenV( HP/Compaq/DEC Basic - Alpha Optimisation, Re: HP/Compaq/DEC Basic - Alpha Optimisation/ Re: Intel jumps Itanic ... HP to drop its unix? # Re: LIB$INITIALIZE: simple example? # Re: LIB$INITIALIZE: simple example? , Re: Newbie Queue Manager in cluster question, Re: Newbie Queue Manager in cluster question, Re: OT You'll never guess what HP advertised, Re: OT You'll never guess what HP advertised, Re: OT You'll never guess what HP advertised& OT: Sun completes  Niagara chip design* Re: OT: Sun completes  Niagara chip design' Re: Parity errors when HSG80 restarted?  Processor Based License Model ! Re: Processor Based License Model ! Re: Processor Based License Model ! Re: Processor Based License Model ! Re: Reserved TCP/IP Port Numbers? % RZ2DD-LS performance specifications ?  Re: SHOW ROUTE	 SPAM, etc ! Re: SUN fails to advertise VMS... / TCP/IP routing - which is the correct behaviour  Re: TNSPING for OpenVMS Alpha . VMS Linker Question - multiple defined symbols Re: XDMP and TCPIP 5.4 Re: XDMP and TCPIP 5.4) Re: You'll never guess what HP advertised ) Re: You'll never guess what HP advertised ) Re: You'll never guess what HP advertised ) Re: You'll never guess what HP advertised ) Re: You'll never guess what HP advertised ) Re: You'll never guess what HP advertised ) Re: You'll never guess what HP advertised ) Re: You'll never guess what HP advertised ) RE: You'll never guess what HP advertised ) Re: You'll never guess what HP advertised ) Re: You'll never guess what HP advertised ) Re: You'll never guess what HP advertised ) Re: You'll never guess what HP advertised ) Re: You'll never guess what HP advertised ) Re: You'll never guess what HP advertised  [OT]: What's Sun really saying?   F ----------------------------------------------------------------------  % Date: Wed, 12 May 2004 22:30:12 +0800 , From: Paul Repacholi <prep@prep.synonet.com>I Subject: Re: %MOUNT-F-MAPHDRBAD, storage map header is bad; volume locked - Message-ID: <874qqlwwej.fsf@prep.synonet.com>   ' info@it-bcsb.de (Klaus-D. Bohn) writes:   B > perhaps anyone of you could help me or could give me some hints.  H > A wrong command "mount /shadow = ..." has destroyed my data on a disk.C > I could stop the shadow copy by 0%. But i can't mount the disk by G > normal way and get the error message "%MOUNT-F-MAPHDRBAD, storage map = > header is bad; volume locked". Mount /foreign works normal.   < Ah, you too have discovered the mount of instant death :|...  D First off, get a BACKUP/PHYS of the disk so you can have a chance of recovering if it goes wrong.  D Next, try MOUNT/OVER=LIMIT and see how that goes. Then try ANAL/DISK and see if it gets anywhere.  8 > What is the best way to recover the data on this disk.   Carfully...   I You have most certainly lost the boot block, home block and the beginning 5 of the indexf.sys. Recovering from this is not easy.     Is the data worth it?    --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  % Date: Wed, 12 May 2004 12:41:11 +0100 9 From: Andrew Harrison <andrew_._remove_harrison@su_n.com> H Subject: Re: Alpha is phased out .... UltraSPARC V in the same way .....0 Message-ID: <c7t2go$cqp$1@new-usenet.uk.sun.com>   David Svensson wrote:       7 >>When did the estimate of 100000 which vanishingly few 2 >>people thought was likely to happen turn into an >>actual sale of 100000 CPU's. >  > 	 > At IDF: B > "Intel CEO Craig Barrett noted that Intel sold more than 100,000  > Itanium processors last year". >   8 Yes and absolutely no one thought that meant that Intels9 partners had sold 100000 Itanium CPU's. Historically more 9 Itanium systems have been shipped into the sales channel, 8 into porting centers and demo rooms than have been sold.   Except obviously you.  > 3 >>And where did the 13000 Itanium CPU's sold by SGI 
 >>come from ?  >  > 4 > What do you mean? I guess Intel manufactured them. >  > = >>>>>>Itanium is for HP, the sooner the rest of the Intel OEM ? >>>>>>community wake up to that fact the better (for them), IBM C >>>>>>already have and its only a matter of time before NEC, Unisys @ >>>>>>and the rest give up on what have been a major problem for! >>>>>>HP and a disaster for them.  >>>>>  >>>>>  >>>>>That is FUD. 1 >>>>>IBM have not changed their Itanium plans and + >>>>>NEC is doing pretty well with Itanium.  >>>>7 >>>>This is not FUD, IBM have chosen not to use Itanium < >>>>for their next generation high end Intel based SMP boxes? >>>>this was key design win that Itanium needed. IBM considered < >>>>Itanium but rejected it because it was too expensive and >>>>not fast enough. >>>  >>> H >>>This has been discussed before... IBM has as far as I know never beenF >>>that keen on Itanium for various reasons. I don't expect them to do >>>more than what they do now. >># >>Sorry but thats the wrong answer.  >  >  > Not from my point of view. >   / And why does that count in this discussion. Its , not your views that matter its the facts and+ the facts don't support your view hence the  first point.   > = >>Had IBM's only option for 64bit been Itanium then I suspect > >>that their decision would have been different, only having a? >>32bit OS and 32bit platform would for a large SMP system with 1 >>up to 64 CPU's have become an increasing issue.  >  > 6 > I don't think IBM would have build one in that case.H > And besides, they push POWER as their high-end offering more nowadays. > 8 Sorry but also untrue. Where do you dream up this stuff.  7 IBM have a number of divisions, the Power boys push the 7 P Series, eSeries team push Intel/AMD based servers etc  etc.2 http://www.pc.ibm.com/us/eserver/xseries/x445.html  3 Is the larges eSeries machine you will note that it 6 supports 32 x86 CPU's exactly the same number of CPU's as the P690.   Regards  Andrew Harrison    > E >>However its also pretty clear that IBM knew about the Intel x86-64  F >>announcements before they made their decision hence their apparentlyB >>odd choice of going with what at the time appeared to be a 32bit >>platform only. >  > 
 > Ofourse.   ------------------------------  % Date: Wed, 12 May 2004 12:54:42 +0100 9 From: Andrew Harrison <andrew_._remove_harrison@su_n.com> H Subject: Re: Alpha is phased out .... UltraSPARC V in the same way .....0 Message-ID: <c7t3a3$d28$1@new-usenet.uk.sun.com>   Rob Young wrote:   > In article <c7qak9$e51$2@new-usenet.uk.sun.com>, Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> writes:   >>boxes. >>/ >>Come on Rob even you can do better than that.  >> >  > 2 > 	Gee Andrew - your "awe come on Robs" get old.   >   2 If you remember you sucess rate in our discussions/ is tiny in fact almost non existant, because of 0 that you may well have been on the receiving end. of more oh come on Rob's than you may like but you should also be used to it.  2 Improving the quality of you postings would reduce& the frequency of the oc come on Rob's.  + > 	You are missing the point - nothing new.  > 	   5 So provide examples of me missing the point why don't  you.  @ > 	Pretend I'm arguing for and against.  Or I'm arguing against. > 	Or I'm arguing for. > @ > 	At 2% of market share it doesn't matter what NEC is doing one > 	way or another. >   = Course it does, 2% of the total server market if all based on < Itanium would make NEC the market leader in Itanium systems.   Regards  Andrew Harrison    ------------------------------  % Date: Wed, 12 May 2004 13:32:16 +0100 9 From: Andrew Harrison <andrew_._remove_harrison@su_n.com> H Subject: Re: Alpha is phased out .... UltraSPARC V in the same way .....0 Message-ID: <c7t5gi$dsf$1@new-usenet.uk.sun.com>   David Svensson wrote:    > Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> wrote in message news:<c7qapp$e51$3@new-usenet.uk.sun.com>...  >  >>David Svensson wrote:  >> >>>Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> wrote in message news:<c7nlk3$g9i$1@new-usenet.uk.sun.com>... >>>  >>> F >>>>Even Dell who have an exclusive deal with Intel are hardly pushing >>>>Itanium are they.  >>>  >>> 5 >>>They are about to ship a 4-way Itanium server now.  >>1 >>Wow, how many years have HP been shipping 4 way  >>Itanium boxes for. >>0 >>If you can't work out why this point is rather1 >>damaging to yours then have a think about Dells  >>market strategy. >  > N > They have until now only a 2-way server, now they are adding a 4-way server.5 > I see this as a bit positive for Itanium from Dell.   6 How so when Dell have an exclusive deal with Intel for4 CPU's. Currently Itanium is their only 64bit option,0 which they have introduced at a slower rate than5 any other Intel OEM depite being the only major Intel ! OEM that can only use Intel CPU's    regards  Andrew Harrison    ------------------------------  % Date: Wed, 12 May 2004 13:05:52 +0100 E From: Tim ffrench-Lynch <tim-DOT-ffrench-HYPHEN-lynch@baesystems.com> ' Subject: Alpha/VMS news on The Inquirer . Message-ID: <40A21320.7C4983AA@baesystems.com>  * At last a little possitive coverage on VMS  ) http://www.theinquirer.net/?article=15859    Tim    ------------------------------  % Date: Wed, 12 May 2004 13:31:55 -0400 < From: "Peter Weaver" <WeaverConsultingServices@sympatico.ca>X Subject: Re: DCLER (DCL Enhancement Request) for DIRECTORY  /SORT=[CREATED,MODIFIED,...]* Message-ID: <2gf5cdF24v79U1@uni-berlin.de>   Michael Austin wrote:  > ... 9 > Ditto!!  looks like the Unix cut command on steroids...   C You might notice in the thread "Suggestion for TYPE command" that I  posted the solution;  C $ merge/stable/nocheck sys$manager:operator.log tt:/spec=sys$input: ( /field=(name=first50,position=1,size=50)
 /data=first50    and someone else posted;   bash$ cut -c 10-50 login.com  > Both do the same thing, but merge works on VMS out of the box.   > ... B > Here is a more efficient way to do this without using tmp files: >...* > Ya just gotta love the pipe command!! :) >...  ; You must have a more efficient version of PIPE than I do :)    --   Peter Weaver Weaver Consulting Services Inc.  Canadian VAR for CHARON-VAX  www.weaverconsulting.ca    ------------------------------  # Date: Wed, 12 May 2004 14:42:57 GMT 1 From: Michael Austin <maustin@firstdbasource.com> Y Subject: Re: DCLER (DCL Enhancement Request) for DIRECTORY /SORT=[CREATED,MODIFIED,...] / 2 Message-ID: <40A23860.FB60255E@firstdbasource.com>   "David J. Dachtera" wrote:   > Peter Weaver wrote:  > >  > > Paul Sture wrote:  > > >...D > > > Peter Weaver posted this neat trick with dates on 30-Apr-2004: > > > D > > > Make sure you run it with SET TERM/WIDTH=132 to avoid wrapping	 > > > ...  > > J > > Actually, if the terminal is set to 80 the wrap should be OK, but hereJ > > is an "Improved" version that handles the terminal wrap better as longK > > as you have the terminal width set to 80. Watch the two long lines, the G > > comments tell you how many spaces should be there. (It sure is nice . > > seeing VMS postings in comp.os.vms again!) > > " > > $ ASSIGN LIB$DATE_FORMAT_037,- > >    LIB$TIME_FORMAT_001 - > >    LIB$DT_FORMAT/USER_MODE) > > $ directx -  ! Ignore any DIR symbols  > >    /date=modified-$ > >    /width=(file:80,display:132)-: > >    /out=sys$scratch:sort_date_'f$getjpi("","PID")'.tmp> > > $ sort sys$scratch:sort_date_'f$getjpi("","PID")'.tmp tt:-! > >     /specification=sys$input: . > > /field=(name=filename1,position:1,size:19)/ > > /field=(name=filename2,position:20,size:60) / > > /field=(name=datefield,position:83,size:22) 3 > > /condition=(name=shortname,test=(filename2 eq " . > >                            ")) ! 60 spaces2 > > /condition=(name=longname,test=(filename2 ne "- > >                           ")) ! 60 spaces " > > /include=(condition=shortname, > > key=datefield, > > data=filename1,  > > data="  ", ! 2 spaces  > > data=datefield) ! > > /include=(condition=longname,  > > key=datefield, > > data=filename1,  > > data=filename2, . > > data="                      ", ! 22 spaces > > data=datefield) L > > $ delete/nolog/noconfirm sys$scratch:sort_date_'f$getjpi("","PID")'.tmp. > H > Thanx for that. That's a far more lucid explanation of how to use SORTC > /SPECIFICATION than I have ever seen in a manual or on-line HELP.  >   7 Ditto!!  looks like the Unix cut command on steroids...   @ Here is a more efficient way to do this without using tmp files:   $ create sys$login:spec.inc +  /field=(name=filename1,position:1,size:19) ,  /field=(name=filename2,position:20,size:60),  /field=(name=datefield,position:83,size:22)0  /condition=(name=shortname,test=(filename2 eq "+                             ")) ! 60 spaces /  /condition=(name=longname,test=(filename2 ne " *                            ")) ! 60 spaces  /include=(condition=shortname,   key=datefield,   data=filename1,  data="  ", ! 2 spaces  data=datefield)  /include=(condition=longname,  key=datefield,   data=filename1,  data=filename2,+  data="                      ", ! 22 spaces   data=datefield)- now create a command procedure called dirsort  $create dirsort.com  $! DIRSORT.COM $ ASSIGN LIB$DATE_FORMAT_037,-     LIB$TIME_FORMAT_001 -      LIB$DT_FORMAT/USER_MODE  $ directx :== directory * $ pipe directx -  ! Ignore any DIR symbols     /date=modified- $     /width=(file:80,display:132) | -=    sort sys$pipe sys$output /specification=sys$login:spec.inc   ( Ya just gotta love the pipe command!! :)     Michael Austin http://www.firstdbasource.com I Going on Vacation?  Need a Consultant to sit in or be on call?  Give me a  call.    >  > -- > David J. Dachtera  > dba DJE Systems  > http://www.djesys.com/ > * > Unofficial Affordable OpenVMS Home Page:! > http://www.djesys.com/vms/soho/    ------------------------------  # Date: Wed, 12 May 2004 17:04:49 GMT . From: "Jerry Alan Braga" <jabraga@flanagan.ca>+ Subject: decserver 900TM, PPP with security ( Message-ID: <ROsoc.2478$RM.994@edtnps89>  J How does one go about setting up an decserver 900TM with PPP on ports thatK have security.   I have been able to get the PPP configuration to work from H a windows 98/2000/XP using a vpn connection over the modem but only withJ security disabled on the port.  I haved looked at the port config and haveG tried both pap and chap but with no luck.  I also configured the server K security as best as I can according the the operations manual (realm, local E user with password, the access type etc..)  but everytime a attempt a L connection from the client is fails with invalid username/pass.  I have evenL tried in XP to force the security method using the advanced button to be PAP or CHAP specifically.    ------------------------------  # Date: Wed, 12 May 2004 17:58:53 GMT 3 From: hammond@not@peek.ssr.hp.com (Charlie Hammond) $ Subject: Re: DFO and SET FILE/NOMOVE2 Message-ID: <xBtoc.1322$Lf2.1213@news.cpqcorp.net>  % In article <c7ijbm$u48$1@online.de>,  R helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply) writes:   > G >Right.  However, if I boot from the CD, I don't have shadowing (and as D >far as I know can't get it), so I would have to run it against each@ >member individually.  In that case, however, I would have to do@ >MOUNT/OVER=SHADOW and reconstruct the shadow set later.  Ugly.   C If you need to do _ANYTHING_ to a shadow set while booted from the  A operating system kit, you FIRST need to break the shadow set into ? a non-shadowed disk.  Then, after whatever, you can re-form the > shadow set.  The process for doing this for the system disk isB set out in the upgrade and installation manual.  (Because you must do this for an upgrade.)  G Please don't take shortcuts.  It is quite easy to operate on one volume G of a broken shadow set and then re-build the set from the other volume, ! thereby loosing whatever you did.    --  J       Charlie Hammond -- Hewlett-Packard Company -- Ft Lauderdale  FL  USAF           (hammond@not@peek.ssr.hp.com -- remove "@not" when replying)J       All opinions expressed are my own and not necessarily my employer's.   ------------------------------  % Date: Wed, 12 May 2004 14:28:06 +0200 ( From: Andreas Davour <ante@update.uu.se> Subject: Re: EVE customizations 0 Message-ID: <cs9wu3h25k9.fsf@Tempo.Update.UU.SE>  = koehler@eisner.nospam.encompasserve.org (Bob Koehler) writes:   ] > In article <cs9vfj7x7hq.fsf@Tempo.Update.UU.SE>, Andreas Davour <ante@update.uu.se> writes:  > I >> Otherwise I find you answer puzzling. I was asking about how to use an K >> editor for coding, so of course that was the fuctionallity I was after!  6 >> (I wasn't talking about coding standards though...) > E >    I get tired of fighting editors that do thingds to my text or my 
 >    code.  7 For some reason I guess you wouldn't like INTERLISP. :)    /andreas   ------------------------------  % Date: Wed, 12 May 2004 14:33:20 +0200 ( From: Andreas Davour <ante@update.uu.se> Subject: Re: EVE customizations 0 Message-ID: <cs9pt9925bj.fsf@Tempo.Update.UU.SE>  = koehler@eisner.nospam.encompasserve.org (Bob Koehler) writes:   f > In article <mddu0ysoyra.fsf@panix5.panix.com>, Rich Alderson <news@alderson.users.panix.com> writes: >>  Q >> As is Emacs, which does not *force* any coding standards on anyone, but rather R >> simply allows those who wish to, to use them.  It's just more easily programmed >> than EVE. > A >    Please, if you can get emacs to consistently obey a C coding * >    standard other than gnu, tell us how. > J >    Our emacs gurus have set up emacs to do C our way every way they knowM >    how, and every time it eventually falls back to gnu style in the middle   >    of editing a file.   E If you mean the "c-mode" happens to you without you know why, then it  should be a simple problem.   H Check your hooks. There are some mode-hooks in there that reacts to yourD file and enters the c-mode. Ask at a emacs group if that don't help.  F Otherwise, change the c-mode code to conform to your coding standards,4 with your ideas about proper indention and so forth.   /andreas   ------------------------------   Date: 12 May 2004 06:40:42 GMT+ From: Thierry Dussuet <thierry@MARS.Family> 8 Subject: Re: Gigantic hobbyist cluster opportunity   :-)/ Message-ID: <slrnca3hna.mc.thierry@MARS.Family>   = On 2004-05-12, JF Mezei <jfmezei.spamnot@teksavvy.com> wrote:  > Alan Adams wrote: 1 > http://toyvax.glendale.ca.us/~vance/vaxbar.html  > O > (I still think that the first process in the SHOW SYS output should be in HIB  > state instead of LEF.   D How about it waiting for the door to open/close so it can switch the light on/off? :-)    Thierry    ------------------------------    Date: 12 May 2004 07:30:37 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) 8 Subject: Re: Gigantic hobbyist cluster opportunity   :-)3 Message-ID: <CqWQj3vqKKlv@eisner.encompasserve.org>   ] In article <slrnca3hna.mc.thierry@MARS.Family>, Thierry Dussuet <thierry@MARS.Family> writes:  > F > How about it waiting for the door to open/close so it can switch the > light on/off? :-)   @    Be carefull or you might end up in the pink scheduling state.   ------------------------------  % Date: Wed, 12 May 2004 21:35:45 +0800 , From: Paul Repacholi <prep@prep.synonet.com>8 Subject: Re: Gigantic hobbyist cluster opportunity   :-)- Message-ID: <878yfxwyxa.fsf@prep.synonet.com>   2 ace join_to ware@iinet.net.au (Tony Epton) writes:   >>4 >>  All installed at ROOSEVELT OKLAHOMA ALL WORKING!  A > I wonder what the shipping would be to Perth, Western Australia   . Bugger off Tony, the lounge is full already...   --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------    Date: 12 May 2004 09:07:13 -0700' From: oj_287@yahoo.co.uk (Ron Atkinson) $ Subject: Help required with CLD file= Message-ID: <f0fbb2b7.0405120807.69447354@posting.google.com>   < I want to be able to enter a command in the following format   	> APP CMD /QUAL=<txt>  F where the first two characters of the <txt> value are fixed (i.e. AB).  C In my CLD file the value of QUAL is set to $quoted_string type i.e.    DEFINE SYNTAX 	QUAL_SYN  IMAGE 		"QUAL.EXE"* QUALIFIER	QUAL VALUE (TYPE=$QUOTED_STRING)  D I guess I want to be able to define a user-defined keyword where theD first two characters are set to AB and the rest of the string can be
 free text.  F I cannot create a new QUAL.EXE file. So is this possible under VMS 6.2# purely within the CLD file itself ?    ------------------------------  % Date: Wed, 12 May 2004 09:35:56 +0200 * From: "Karsten Nyblad" <nospam@nospam.com>C Subject: Re: HP to can HP/UX? Intel pulls out of IA-64 prematurely? , Message-ID: <c7sk50$69c$1@news.cybercity.dk>  3 "Jan-Erik Sderholm" <aaa@aaa.com> wrote in message ! news:40A13FEE.F93EF653@aaa.com... F > Isn't that much like "Unix" situation way back when there was a kindG > of open source, common source. Before Digital, SUN, IBM, SCO, whoever D > begun to add there own stuff. So in 10-20 years from now, we couldD > very well have a SUN-Linux, a HP-Linux, a IBM-Linux and so on. All? > just as different as Tru64, UP-UX, AIX ans Solaris are today.  >  > Because :  > G > - It's the *differences* between companies that generates the income.  >   Not the similarities.  >  > Jan-Erik.   H It will be difficult because any changes to the kernel or to most of theK applications has to be open source.  Thus the competitors can have the same H features a few months later.  The differences has to be in closed source
 applications.   L In the early Unix days the Internet was not as big as it is today.  Further,I open source was not an established way of developing software.  There are E many, many more people that write open source today.  If some company I develop a killer application and start getting market share with it, then L you can be sure that there will be three open source projects trying to make a clone.   Karsten Nyblad ibpit1202 at sneakemail dot com    ------------------------------  % Date: Wed, 12 May 2004 12:16:07 +0100 9 From: Andrew Harrison <andrew_._remove_harrison@su_n.com> C Subject: Re: HP to can HP/UX? Intel pulls out of IA-64 prematurely? 0 Message-ID: <c7t11o$car$1@new-usenet.uk.sun.com>   JF Mezei wrote:  > P > What the article really says is that inside HP, there are some who don't thinkP > too much of that IA64 thing, and there are some who feel HP-UX isn't worth the& > effort and they should just go Unix. > N > However, there are interesting commercial implications about Linux. Lest sayM > IBM or HP or Sun were to abandon their proprietary Unix and go Linux. Would P > they give money to Redhat/SuSe/etc, or would they decide to maintain their own+ > version of Linux for their own platform ?  > C Its unlikely to happen in the near term. If for example you compare E Solaris 10 with the 2.6 Linux Kernel you realise that the gap between A what Linux is capable of and what a commercial UNIX is capable of # doesn't seem to have narrowed much.   F Nor is it likely that Linux on Power/SPARC/Itanium will be sucessfull,D Linux plays in the 1-4 way server space and the desktop and the vastI majority of applications available for Linux only run on x86. The sucess  = of Power/SPARC/Itanium as a Linux platform would also require B Power/SPARC/Itanium to become the commodity platform in that space( which is also highly unlikely to happen.  C That doesn't mean that HP, Sun and IBM won't offer Linux solutions, A Sun for example uses SuSE as the basis of our JDS desktop OS. But G that is targetted at the commodity desktop which is currently dominated  by x86.   C The fly in the ointment is Intels recent announcments that they are D going multi-core and dropping Tejas etc in favour of lower speed butD dual + core CPU's. This currently does not play to Linux's strengthsE which are not in heavily threaded apps which you will need to exploit , the new Intel CPU's to their best advantage.  > AMD I am sure have identified the opportuinity, they currently? out-perform Prescott with Athlon-64 and XeonDP/MP with Opteron, ? they will have to go dual core later than Intel and with faster < cores. This could result in the platform of choice for Linux" shifting from Intel x86 to AMD-64.  ? This would be interesting for Sun as we now own the largest AMD + server design house outside AMD themselves.    Regards  Andrew Harrison    ------------------------------  % Date: Wed, 12 May 2004 12:25:06 +0100 9 From: Andrew Harrison <andrew_._remove_harrison@su_n.com> C Subject: Re: HP to can HP/UX? Intel pulls out of IA-64 prematurely? 0 Message-ID: <c7t1ij$cfq$1@new-usenet.uk.sun.com>   david20@alpha2.mdx.ac.uk wrote:   e > In article <40A13FEE.F93EF653@aaa.com>, Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com> writes:  >  >>JF Mezei wrote:  >>M >>>If they decide to go with a "linux vendor", that linux vendor then becomes O >>>more or less a proprietary vendor of a Unix OS. So there would be commercial = >>>versions of Linux and there would be open source versions.  >>> 7 >>>So to end up with the same paradigm as you have now.  >>F >>Isn't that much like "Unix" situation way back when there was a kindG >>of open source, common source. Before Digital, SUN, IBM, SCO, whoever D >>begun to add there own stuff. So in 10-20 years from now, we couldD >>very well have a SUN-Linux, a HP-Linux, a IBM-Linux and so on. All? >>just as different as Tru64, UP-UX, AIX ans Solaris are today.  >> >>Because :  >>G >>- It's the *differences* between companies that generates the income.  >> Not the similarities. >> >  > M > The difference is that all those companies made changes to the UNIX kernels  > they inherited. N > Linux is a registered trademark of Linus Torvalds and he keeps a pretty firmO > grip on the Kernel. As long as the companies want to keep on calling it Linux J > the Kernel will remain compatible between different companies offerings. >   F Sadly that has very little impact on the compatibility of applicationsA across Linux distributions. Having the same Linux kernel is not a B guarantee that an application will run unmodified beween two Linux distributions.  > In reality you have Mandrake, RedHat, Debian, SuSE, Fedora etc@ each of which is slightly different from the other each of whichA requires binarys targetted to that platform. Sometimes you do get  lucky but sometimes you don't.   Regards  Andrew Harrison    ------------------------------  % Date: Wed, 12 May 2004 13:59:56 +0200   From: "Dr. Dweeb" <dr@dweeb.com>C Subject: Re: HP to can HP/UX? Intel pulls out of IA-64 prematurely? , Message-ID: <c7t3js$p2p$1@news.cybercity.dk>   Andrew Harrison wrote:! > david20@alpha2.mdx.ac.uk wrote:  > 3 >> In article <40A13FEE.F93EF653@aaa.com>, Jan-Erik 5 >> =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com> writes:  >> >>> JF Mezei wrote:  >>> G >>>> If they decide to go with a "linux vendor", that linux vendor then E >>>> becomes more or less a proprietary vendor of a Unix OS. So there B >>>> would be commercial versions of Linux and there would be open >>>> source versions.  >>>>9 >>>> So to end up with the same paradigm as you have now.  >>> H >>> Isn't that much like "Unix" situation way back when there was a kindA >>> of open source, common source. Before Digital, SUN, IBM, SCO, E >>> whoever begun to add there own stuff. So in 10-20 years from now, G >>> we could very well have a SUN-Linux, a HP-Linux, a IBM-Linux and so B >>> on. All just as different as Tru64, UP-UX, AIX ans Solaris are
 >>> today. >>> 
 >>> Because :  >>> A >>> - It's the *differences* between companies that generates the ! >>> income. Not the similarities.  >>>  >> >>F >> The difference is that all those companies made changes to the UNIX
 >> kernels >> they inherited.C >> Linux is a registered trademark of Linus Torvalds and he keeps a  >> pretty firmG >> grip on the Kernel. As long as the companies want to keep on calling  >> it Linux @ >> the Kernel will remain compatible between different companies
 >> offerings.  >> > H > Sadly that has very little impact on the compatibility of applicationsC > across Linux distributions. Having the same Linux kernel is not a D > guarantee that an application will run unmodified beween two Linux > distributions. > @ > In reality you have Mandrake, RedHat, Debian, SuSE, Fedora etcB > each of which is slightly different from the other each of whichC > requires binarys targetted to that platform. Sometimes you do get   > lucky but sometimes you don't. >   % nee nee nee nee -- nee nee nee nee --   L Oh no, I am entering the twighlight zone, another space and dimension, beingK transported to a time when application compatibility accross Unix platforms A was a claimed and largely non-existent advantage of Unix* per se.     Is this deja vu all over again ?  	 > Regards  > Andrew Harrison    ------------------------------  % Date: Wed, 12 May 2004 14:04:02 +0200 * From: "Karsten Nyblad" <nospam@nospam.com>C Subject: Re: HP to can HP/UX? Intel pulls out of IA-64 prematurely? , Message-ID: <c7t3rl$pia$1@news.cybercity.dk>  F "Andrew Harrison" <andrew_._remove_harrison@su_n.com> wrote in message* news:c7t11o$car$1@new-usenet.uk.sun.com...E > Its unlikely to happen in the near term. If for example you compare G > Solaris 10 with the 2.6 Linux Kernel you realise that the gap between C > what Linux is capable of and what a commercial UNIX is capable of % > doesn't seem to have narrowed much.  > H > Nor is it likely that Linux on Power/SPARC/Itanium will be sucessfull,F > Linux plays in the 1-4 way server space and the desktop and the vast? > majority of applications available for Linux only run on x86.   D When the 2.6 kernel was released, it was claimed that the new kernelI supported up to 32 processors.  Do you think that is bull, i.e., that the L kernel will not perform under normal work loads as, e.g., a database server?I Or are you simply saying that companies have not started to used Linux on  server larger than 4 ways yet?   Karsten Nyblad ibpit1202 at sneakemail dot com    ------------------------------  % Date: Wed, 12 May 2004 15:49:22 +0100 9 From: Andrew Harrison <andrew_._remove_harrison@su_n.com> C Subject: Re: HP to can HP/UX? Intel pulls out of IA-64 prematurely? 0 Message-ID: <c7tdhi$gh3$1@new-usenet.uk.sun.com>   Jan-Erik Sderholm wrote:    > JF Mezei wrote:  > L >>If they decide to go with a "linux vendor", that linux vendor then becomesN >>more or less a proprietary vendor of a Unix OS. So there would be commercial< >>versions of Linux and there would be open source versions. >>6 >>So to end up with the same paradigm as you have now. >  > F > Isn't that much like "Unix" situation way back when there was a kindG > of open source, common source. Before Digital, SUN, IBM, SCO, whoeverUD > begun to add there own stuff. So in 10-20 years from now, we couldD > very well have a SUN-Linux, a HP-Linux, a IBM-Linux and so on. All? > just as different as Tru64, UP-UX, AIX ans Solaris are today.= >   7 That is exactly what the Linux market looks like today.e  = RedHat, Debian, SuSE, Mandrake all similar all just differentm enough to make life a PITA.s  @ Because as you say its the differences between the distributions that generates the income.   Regardse Andrew Harrison-   ------------------------------  % Date: Wed, 12 May 2004 15:55:10 +0100A9 From: Andrew Harrison <andrew_._remove_harrison@su_n.com>AC Subject: Re: HP to can HP/UX? Intel pulls out of IA-64 prematurely?t0 Message-ID: <c7tdse$gmo$1@new-usenet.uk.sun.com>   Dr. Dweeb wrote:   > Andrew Harrison wrote: >>@ >>In reality you have Mandrake, RedHat, Debian, SuSE, Fedora etcB >>each of which is slightly different from the other each of whichC >>requires binarys targetted to that platform. Sometimes you do get?  >>lucky but sometimes you don't. >> >  > ' > nee nee nee nee -- nee nee nee nee --W > N > Oh no, I am entering the twighlight zone, another space and dimension, beingM > transported to a time when application compatibility accross Unix platforms C > was a claimed and largely non-existent advantage of Unix* per se.i > " > Is this deja vu all over again ? >   = Yes except its worse because all the distributions are calledE= Linux all run on x86 and this gives people the "strange" idean that they might be compatible.  ? At least you had Solaris, AIX, HP-UX, OSF-1 all different namesT< all running on different platforms, much less expectation of compatibility.   Regards  Andrew Harrisona   ------------------------------  % Date: Wed, 12 May 2004 16:03:31 +0100r9 From: Andrew Harrison <andrew_._remove_harrison@su_n.com> C Subject: Re: HP to can HP/UX? Intel pulls out of IA-64 prematurely? 0 Message-ID: <c7tec4$guf$1@new-usenet.uk.sun.com>   Karsten Nyblad wrote:Z  H > "Andrew Harrison" <andrew_._remove_harrison@su_n.com> wrote in message, > news:c7t11o$car$1@new-usenet.uk.sun.com... > E >>Its unlikely to happen in the near term. If for example you compareeG >>Solaris 10 with the 2.6 Linux Kernel you realise that the gap betweenPC >>what Linux is capable of and what a commercial UNIX is capable ofg% >>doesn't seem to have narrowed much.v >>H >>Nor is it likely that Linux on Power/SPARC/Itanium will be sucessfull,F >>Linux plays in the 1-4 way server space and the desktop and the vast? >>majority of applications available for Linux only run on x86.l >  > F > When the 2.6 kernel was released, it was claimed that the new kernelK > supported up to 32 processors.  Do you think that is bull, i.e., that therN > kernel will not perform under normal work loads as, e.g., a database server?K > Or are you simply saying that companies have not started to used Linux ong  > server larger than 4 ways yet? >   H It will all depend on the workload, Sun's experience of adding threadingE to UNIX was that this was an iterative process in terms of tuning thelF threading to get good scalability for most workload, this process tookD the best part of a decade with rather more engineering resources andB hardware/ISV benchmark resources than are available to develop the
 Linux kernel.   A Linux 2.6 is roughly where very early versions of Solaris were inm
 this process..  F You will get workloads that should scale well, HPC for example because@ typically it is mostly user and not kernel space. This should be OK for SGI but bad for HP.  > To give you an idea of how far adrift it is SGI published some; scalability numbers for a standard commercial type workloadt= which Sun would expect to scale linearly on Solaris and which = is also pretty benign, SGI got 13x throughput using 28 CPU's.   < I will leave you to decide if this is good, bad or terrible.     Regardss Andrew Harrisonh   ------------------------------    Date: 12 May 2004 05:51:50 -07002 From: williamwebb@openvms-rocks.com (William Webb)T Subject: Re: HP World Magazine: For Business Continuity...the answer may be  OpenVMS= Message-ID: <bf98c417.0405120451.5a06ea8e@posting.google.com>   ~ "David J. Dachtera" <djesys.nospam@NeOaSrPtAhMlNiOnWk.net> wrote in message news:<40A18AC5.C8A62BF0@NeOaSrPtAhMlNiOnWk.net>... > Robert Deininger wrote:o > > M > > In article <b8876400c5f6265491ad8f3db64c424a@news.teranews.com>, JF Mezei ) > > <jfmezei.spamnot@teksavvy.com> wrote:s > >  > > >Robert Deininger wrote:J > > >> I don't know exactly _why_, but it is the case.  Digital built goodA > > >> systems, but they were not very good at controlling costs.p > > >hM > > >Aren't Alpha systems such as DS10s (and I would assume DS15s) built from- > > >industry standard parts ? > >  > > Partly.  > > 0 > > >They use industry standard PCI bus, right ?K > > The proper punctuation, in this case, is "industry standard".  PCI is a  > > nasty standard > ; > I believe "specification" would be more appropriate here.n > ; > >, with lots of wiggle room.  Many hours have been wastede$ > > getting PCI stuff to work right. > F > Likewise with SCSI. Witness: getting VMS to play nice with so-called > "standard" SCSI devices. e > F > Every vendor takes a slightly difference view of the specifications.  > Hence, our technical dilemmas.  D I've pointed out in the past that neither the first nor the second S5 in the acronym "SCSI" represents the word "Standard".   * That fact alone should tell you something.   :^)o   WWWebb   ------------------------------  + Date: Wed, 12 May 2004 06:16:42 +0000 (UTC)l6 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER)S Subject: Re: HP World Magazine: For Business Continuity...the answer may be OpenVMSa1 Message-ID: <newscache$jy7lxh$k712$1@news.sil.at>F  x In article <40A18A13.7B85945@NeOaSrPtAhMlNiOnWk.net>, "David J. Dachtera" <djesys.nospam@NeOaSrPtAhMlNiOnWk.net> writes: >JF Mezei wrote: >> Robert Deininger wrote:H >> > I don't know exactly _why_, but it is the case.  Digital built good? >> > systems, but they were not very good at controlling costs.u >> eK >> Aren't Alpha systems such as DS10s (and I would assume DS15s) built from>I >> industry standard parts ? They use industry standa<rd PCI bus, right ?t< >> Industry standard power supply, drives, connectors, etc ?   Unfortunateley not really.M The power supply of my DS10 died (three 3 days after the end of the warranty)fI and it costed me over 150.- to replace it (and I even went to the serviceeO center to avoid shipping costs). It has no similarities with the power suppliesi I've seen so far for the PCs.n  A >You have to be careful how you use the term "industry standard".o >oF >Industry standards are defined by standards setting bodies like ANSI,H >ISO, CCITT, IEEE and so on. Standards are not set either arbitrarily orH >unilaterally as is the case with Micro$hit, nor are they defined by howI >many people use what brand of (x). Standards are defined by the industry 4 >at large through research, review, and negotiation. >fE >While the items you cite may employ protocols defined by one or moresC >standards bodies, I sincerely doubt whether any specific device or.E >vendor is defined by any standards body as a standard specification.o >i  >Widely used = de facto standard >'* >Defined specification = industry standard  L I haven't seen defined specifications for hardware (only for protocols, ...)E which would make a difference for what JF wrote. I see it as hardwarer# standards are de facto standards...d   -- e Peter "EPLAN" LANGSTOEGERt% Network and OpenVMS system specialisto E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------  % Date: Wed, 12 May 2004 15:44:22 +0100p9 From: Andrew Harrison <andrew_._remove_harrison@su_n.com>eY Subject: Re: HP World Magazine: For Business Continuity...the answer may be OpenVMS OpenVe0 Message-ID: <c7td8a$gf2$1@new-usenet.uk.sun.com>   Robert Deininger wrote:rK > In article <d8ece8766feb0230d57ad42fe1bde30b@news.teranews.com>, JF Mezeid' > <jfmezei.spamnot@teksavvy.com> wrote:A >  >  >>Robert Deininger wrote:  >>J >>>The components used in a DS10 probably haven't gotten much cheaper, andI >>>DS15 is in the same ballpark.  I think $6-$7k is too expensive for VMSi >>>entry level.  >>H >>>Obviously, HP has some flexibility in pricing.  But they can build an# >>>rx1600 for far less than a DS15,l >> >>M >>Ok, please explain why a DS10 would have to be more expensive to HP than ani >>itanium equivalent.o >  > E > I don't know exactly _why_, but it is the case.  Digital built goodgK > systems, but they were not very good at controlling costs.  Compaq didn't L > really change the alphaserver HW business.  HP HW groups seem to be betterL > at designing and building systems at lower cost.  That is reflected in the > selling prices.e >   : There is no reason why the DS15 could not be produced at a9 competitive price to the rx1600. Sun's pricing of the Sun19 V210/V240 show that it is perfectly possible to build andi2 non "commodity" server to a similar price point as commodity x86 systems.    I > I happened to know the approximate cost to manufacture DS15 when it wasnL > released.  There isn't all that much profit in the sale of these systems. I > I strongly suspect DS15 costs more to manufacture than the list selling  > price of rx1600. >   ? Thats not really the right answer, DS15 margins may well be low@@ but that doesn't mean that they ave to be it just means you hvae< to work a bit harder at designing cost out of the system and% improving your manufacturing process.>   Regards  Andrew Harrison    ------------------------------    Date: 12 May 2004 06:25:46 -07000 From: chris@townleyc.demon.co.uk (Chris Townley)1 Subject: HP/Compaq/DEC Basic - Alpha Optimisations= Message-ID: <93b50805.0405120525.74111de6@posting.google.com>e  F Is there any documentation available about the optimisations performedD at the various levels? I have seen the minimal information available in the HP Basic user guide.o  @ I have a need to use optimisation on certain areas of code, so IB started by changing my build routines to use level 4 for all. I am& certianly seeing some strange results.  C This is DEC BASIC V1.3-000 running under OpenVMS 6.2 I have a large B code base that was originally on VAX (actually PDP first), but now6 runs fine on Alpha, but compiled without optimisation.   --   ChrisT   ------------------------------  % Date: Wed, 12 May 2004 13:32:29 +0000i- From: David B Sneddon <dbsneddon@bigpond.com>.5 Subject: Re: HP/Compaq/DEC Basic - Alpha Optimisationt* Message-ID: <40A2276D.9000508@bigpond.com>  # Chris Townley was overheard to say: H > Is there any documentation available about the optimisations performedF > at the various levels? I have seen the minimal information available > in the HP Basic user guide.D > B > I have a need to use optimisation on certain areas of code, so ID > started by changing my build routines to use level 4 for all. I am( > certianly seeing some strange results. > E > This is DEC BASIC V1.3-000 running under OpenVMS 6.2 I have a largetD > code base that was originally on VAX (actually PDP first), but now8 > runs fine on Alpha, but compiled without optimisation. >    Chris,  @ What type of "strange results" are you seeing?  I saw some weird@ behaviour in some routines when compiling with /OPTIMIZE=LEVEL=4; (the default) and overcame them using the OPTION statement. D (I can't recall what OPTION statement I used, I am at home, the code% is at work, but will check tomorrow.)    Regards, Dave.v -- nI David B Sneddon (dbs)    VMS Systems Programmer     dbsneddon@bigpond.comcI Sneddo's quick guide ...          http://www.users.bigpond.com/dbsneddon/ I DBS freeware at ...   http://www.users.bigpond.com/dbsneddon/software.htmCI "Life is what happens to you while you're busy making other plans" Lennong   ------------------------------  % Date: Wed, 12 May 2004 12:29:17 +0100 9 From: Andrew Harrison <andrew_._remove_harrison@su_n.com> 8 Subject: Re: Intel jumps Itanic ... HP to drop its unix?0 Message-ID: <c7t1qe$ch8$1@new-usenet.uk.sun.com>   Bob Ceculski wrote:1  . > does HP now have to steer the itanic itself? > + > http://www.theinquirer.net/?article=15834b    / This appears to be just a rumour, however if itd. was true the answer would be that HP could not/ afford to do Itanium themselves without a major , increase in their R&D budget which currently) stands at 4 billion a year or a major cut , back in soem of the programs paid for by the profits from toner.N  . Given that you are an OpenVMS advocate this is0 not a scenario that you even want to contemplate1 unless you actually think that HP would axe HP-UX- instead of OpenVMS.:   Regards1 Andrew Harrison=   ------------------------------    Date: 12 May 2004 07:59:54 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen) , Subject: Re: LIB$INITIALIZE: simple example?3 Message-ID: <oUTmhE0RF85+@eisner.encompasserve.org>n  B In article <04051115194795@antinode.org>, sms@antinode.org writes:% > From: hoff@hp.nospam (Hoff Hoffman)t  - >> :/* (What does this stuff actually do?) */  > L >>   It sets up a program section specifically known to the image activator. > K >    If you say so.  I assume that the name LIB$INITIALIZE is significant. =A > Is LIB$INITIALIZ just an aligner/filler where any name will do?a  D There are VMS-supplied PSECTS which alphabetize just before and justB after LIB$INITIALIZE.  Anything between those PSECTS is taken as aC pointer to some code which will be executed by the image activator.n  ? Note that this happens _before_ the debugger is initialized, sooA debugging inside your LIB$INITIALIZE code is not straightforward.S   ------------------------------  + Date: Wed, 12 May 2004 10:27:46 -0500 (CDT)l From: sms@antinode.org, Subject: Re: LIB$INITIALIZE: simple example?) Message-ID: <04051210274664@antinode.org>:  - From: Kilgallen@SpamCop.net (Larry Kilgallen)i  M > >    If you say so.  I assume that the name LIB$INITIALIZE is significant. lC > > Is LIB$INITIALIZ just an aligner/filler where any name will do?e  F > There are VMS-supplied PSECTS which alphabetize just before and justD > after LIB$INITIALIZE.  Anything between those PSECTS is taken as aE > pointer to some code which will be executed by the image activator.t  H    I'd've picked LIB$INITIALIZD instead of LIB$INITIALIZ, but presumably that's close enough.  A > Note that this happens _before_ the debugger is initialized, so C > debugging inside your LIB$INITIALIZE code is not straightforward.c  8    I noticed.  printf() worked.  Thanks for the details.  H ------------------------------------------------------------------------  4    Steven M. Schweda               (+1) 651-699-98183    382 South Warwick Street        sms@antinode-orgh    Saint Paul  MN  55105-2547s   ------------------------------  % Date: Wed, 12 May 2004 06:50:58 -0400   From: John Santos <JOHN@egh.com>5 Subject: Re: Newbie Queue Manager in cluster questionV6 Message-ID: <1040512063922.23626A-100000@Ives.egh.com>  ' On Mon, 10 May 2004, Chuck Chopp wrote:c   > JF Mezei wrote:t > , > > OK, I can now better refine my question: > > \ > > the TCPIP$SMTP queues run on node2, and are "server" queues which do not have autostart.8 > > The QMAN$MASTER file resides on node1's system disk. > > V > > When Node1 fails and restarts, TCPIP$SMTP queues on node2 remain in stopped state. > > R > > Node1 can't blindly restart them since node1 isn't sure of the status of node2K > > (whether the TCPIP stack on NODE2 is up and running, if smtp service is  > > enabled etc).w > > O > > Is there a very elegant way to ensure that the state (started/stopped) of atK > > queue is restored once the queue manager is again granted access to its  > > database ? > > I > > Or must node1 submit a job to run on node2 and do whatever checks arerI > > necessary to determine whether the queue should be restarted or not ?u >  > K > Seeing that the logical name QMAN$MASTER is being used, you're using the aI > massively revised queue manager that was implemented as something like eK > OpenVMS v6.x, right?  Doesn't that queue manager implemenation allow for  K > multiple instances of the queue manager to be running, each with its own aO > separate queue manager databases?  Maybe there is a need for a common set of .J > files regardless of how many queue manager instances are running, but I L > thought those instances were allowed to be 100% independent of each other I > just because of situations like this.  Now I'll have to ferret out the  N > specifics in the OpenVMS docs themselves to satisfy my own curiosity.  What N > I was going to suggest, should this turn out to be true, is that you create K > separate instances of the queue manager and assign certain queues to the  J > their own instance that has its qman files located on node2.  I suppose D > it'll take some reading to determine if that can actually be done. >  >  > -- r
 > Chuck Chopps  G You can certainly do this.  However, the downside is the queuing systemaC on each node is completely separate from the other nodes.  In othereE words, you can't SUBMIT jobs on NODEA to run on NODEB, or have commonvA queues that can execute on which ever node is least busy, and alltC the various queue manipulation commands (SUBMIT, PRINT, SHOW ENTRY,iB DEFINE/FORM, ASSIGN/MERGE, etc., etc.) apply only to queues on theD local node.  You have to SET HOST to the other node or use SYSMAN to+ do these things to the other set of queues.   E We have a cluster (for software development only, no production work),E with 2 Alphas and a VAX, and no common disks, just local disks, so we C have it configured this way.  With network printers, we just have aUA print queue on each node for each printer, so that's no big deal.rE It would be nice to have batch queues that could run on any node, butsD for example, if you are compiling something, you usually want eitherB the VAX compilers or the Alpha compilers, so you care which system? the batch executes on.  And we want each node to be able to runa@ without the others, so we don't want to make one node the "queue> master" node and not to be able to run any batch jobs or print= anything when that node is down.  So it hasn't been worth the = effort to merge all the queues into one queuing system.  (The @ VAX and one of the Alpha's were originally stand-alone systems.)  @ So we've been running with separate queuing systems on each node( for years and it has always worked fine.     -- e John SantosK Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  % Date: Wed, 12 May 2004 09:11:56 -0400  From: norm.raphael@metso.com5 Subject: Re: Newbie Queue Manager in cluster question Q Message-ID: <OFE03971D7.3A03ADE1-ON85256E92.00483E48-85256E92.0048908B@metso.com>   ; John Santos <JOHN@egh.com> wrote on 05/12/2004 06:50:58 AM:n  ) > On Mon, 10 May 2004, Chuck Chopp wrote:n >u > > JF Mezei wrote:  > >r. > > > OK, I can now better refine my question: > > >yA > > > the TCPIP$SMTP queues run on node2, and are "server" queuesu > which do not have autostart.: > > > The QMAN$MASTER file resides on node1's system disk. > > > F > > > When Node1 fails and restarts, TCPIP$SMTP queues on node2 remain > in stopped state.- > > >nD > > > Node1 can't blindly restart them since node1 isn't sure of the > status of node2tJ > > > (whether the TCPIP stack on NODE2 is up and running, if smtp service is > > > enabled etc).u > > >r: > > > Is there a very elegant way to ensure that the state > (started/stopped) of aI > > > queue is restored once the queue manager is again granted access toR itsg > > > database ? > > >pK > > > Or must node1 submit a job to run on node2 and do whatever checks are5K > > > necessary to determine whether the queue should be restarted or not ?  > >/ > >aH > > Seeing that the logical name QMAN$MASTER is being used, you're using thesJ > > massively revised queue manager that was implemented as something likeH > > OpenVMS v6.x, right?  Doesn't that queue manager implemenation allow for H > > multiple instances of the queue manager to be running, each with its own7B > > separate queue manager databases?  Maybe there is a need for a > common set ofgK > > files regardless of how many queue manager instances are running, but Ic  G > > thought those instances were allowed to be 100% independent of each  other J > > just because of situations like this.  Now I'll have to ferret out theI > > specifics in the OpenVMS docs themselves to satisfy my own curiosity.o WhatH > > I was going to suggest, should this turn out to be true, is that you createH > > separate instances of the queue manager and assign certain queues to the K > > their own instance that has its qman files located on node2.  I supposew  F > > it'll take some reading to determine if that can actually be done. > >: > >> > > -- > > Chuck ChoppO >EI > You can certainly do this.  However, the downside is the queuing system E > on each node is completely separate from the other nodes.  In other G > words, you can't SUBMIT jobs on NODEA to run on NODEB, or have common0C > queues that can execute on which ever node is least busy, and all E > the various queue manipulation commands (SUBMIT, PRINT, SHOW ENTRY,,D > DEFINE/FORM, ASSIGN/MERGE, etc., etc.) apply only to queues on theF > local node.  You have to SET HOST to the other node or use SYSMAN to- > do these things to the other set of queues.2 > G > We have a cluster (for software development only, no production work)eG > with 2 Alphas and a VAX, and no common disks, just local disks, so weoE > have it configured this way.  With network printers, we just have amC > print queue on each node for each printer, so that's no big deal. G > It would be nice to have batch queues that could run on any node, buteF > for example, if you are compiling something, you usually want eitherD > the VAX compilers or the Alpha compilers, so you care which systemA > the batch executes on.  And we want each node to be able to runrB > without the others, so we don't want to make one node the "queue@ > master" node and not to be able to run any batch jobs or print? > anything when that node is down.  So it hasn't been worth the-? > effort to merge all the queues into one queuing system.  (TherB > VAX and one of the Alpha's were originally stand-alone systems.) >uB > So we've been running with separate queuing systems on each node* > for years and it has always worked fine. >w >e  I When we had separate queue management processes on nodes in a VMScluster,xI we used submit/remote to run a command procedure on a queue on the target.H node that resubmitted the job on the target node to the desired queue onE that node with all the particulars (since submit/remote is restrictedP& as to what qualifiers can be invoked).  K That allowed us for example to run VAX applications on a VAX in the cluster F when the data was available from a process on an Alpha in the cluster.   > --
 > John Santosa > Evans Griffiths & Hart, Inc. > 781-861-0670 ext 539 >a   ------------------------------  % Date: Wed, 12 May 2004 10:36:36 +0200r  From: "Dr. Dweeb" <dr@dweeb.com>5 Subject: Re: OT You'll never guess what HP advertisede, Message-ID: <c7snmj$9jl$1@news.cybercity.dk>   Karsten Nyblad wrote:t- > "Dr. Dweeb" <dr@dweeb.com> wrote in messaget) > news:c7rje1$2cuf$1@news.cybercity.dk...y >> Karsten Nyblad wrote:E >>> The market for luxury items is about showing how much money have.m >>F >> That is true for some, but certainly not true for all the items youC >> consider luxury.  You probably think a Mercedes Benz is a luxuryeF >> car, yet it is probably the best value proposition new car purchase >> there is. >>F >> We I inclined, I could prove to you that the best value propositionF >> for a car owner is to buy a Mercedes Benz, rather than your Lada orF >> skoda or whatever it is you thinkn we should all drive.  Anyone whoE >> has done fleet management understands the sums - for Joe Familymanc. >> and the correct preconditions - buy a Merc. >oA > Feel free to think that your decission on which car to drive is # > solely based on rational reasons.a >t  H My reasons are not at all rational and I have made and intend to make noK such claim - they just have nothing to do with the reasons you claim.  YouroL statement, "is about showing how much money have" belongs to early 20th leftK wing economic rhetoric and is not a generally accurate statement.  That youf; and many others believe this to be true is your own affair.x  G >>> It is not about the quality or true value of these products.  Thus,hD >>> rich people buy Maybachs, Ferraris, etc. to show of their money.A >>> The more expensive the car is, the more valuable it is to itse= >>> customers regardless of what it costs to produce the car.  >>F >> You must be a f*****g Dane.  That country is full of envious people! >> with opinions just like yours.l >uC > LOL.  You should watch your language.  It is you who claim that IuG > envy those people.  They can use their money as they want as far as I E > am concerned. People use an amasingly large part of their wages foraG > cars, and that is also true for blue collar workers, students, peoplesG > on social welfare, etc.  An important reason for using all that moneyc' > is to send an image to other persons.a >   D Except for people who actually buy cars to collect (like stamps) theI "important reason", singular, for a first car purchase is in 99.99999% ofdG cases transportation.  All other reasons are subsidiary.  If you do nots/ believe this, try selling a car without wheels.u  H Most car owners have to look at, sit in and travel in their vehicle on aL daily basis.  For the same reason my living room walls are not painted pink,K I do not own a pink car.  Fill in the rest of the attributes.  The primary,eE irrational factors you would say, have to do with aesthetic, comfort,tJ economy, utility and other rational attributes.  The "image" thing you areJ so fond of is out at the end of the scale and tilts the scale left, right,= centre or whatever.  It is A factor, not THE factor.  It is a-G differentiating factor.  It is often THE deciding factor.  As our othernL scandinavian poster commented somewhere else, it is about the "differences".  I People also by $30,000 ergonomically correct beds in which to plant theirnE bodies every night - would you suggest that such luxury purchases are1> irrational and just to show how much money they have as well ?  F That some people dispose of income in crazy ways is a statement of the obvious.  C >> Just what is a MayBach ?? - I will need to resaerch this one :-)t > ? > It is Daimler-Chryslers car in the Bentley/R&R class of cars.f >n  J Out of my league I guess.  Are they actively sold in Europe, or is it a US1 only thing like a Cadillac from the Chrysler arm?c   > Karsten Nyblad! > ibpit1202 at sneakemail dot com    ------------------------------  % Date: Wed, 12 May 2004 11:15:45 +0200>* From: Paul Sture <nospam@sture.homeip.net>5 Subject: Re: OT You'll never guess what HP advertised * Message-ID: <2ge8a3F1rinsU1@uni-berlin.de>   Dr. Dweeb wrote: > C >>>Just what is a MayBach ?? - I will need to resaerch this one :-)e >>? >>It is Daimler-Chryslers car in the Bentley/R&R class of cars.t >> >  > L > Out of my league I guess.  Are they actively sold in Europe, or is it a US3 > only thing like a Cadillac from the Chrysler arm?r >  >   + The best description I've seen of it, is ate  ) http://auto.howstuffworks.com/maybach.htme  I Lots of links to follow from which appears it is not just sold in the US.i   ------------------------------  % Date: Wed, 12 May 2004 12:59:36 +0200e  From: "Dr. Dweeb" <dr@dweeb.com>5 Subject: Re: OT You'll never guess what HP advertised , Message-ID: <c7t02l$keh$1@news.cybercity.dk>   Paul Sture wrote:d > Dr. Dweeb wrote: >>E >>>> Just what is a MayBach ?? - I will need to resaerch this one :-)e >>>eA >>> It is Daimler-Chryslers car in the Bentley/R&R class of cars.h >>>u >> >>E >> Out of my league I guess.  Are they actively sold in Europe, or isi< >> it a US only thing like a Cadillac from the Chrysler arm? >> >> >-- > The best description I've seen of it, is at  >m+ > http://auto.howstuffworks.com/maybach.htm  >cG > Lots of links to follow from which appears it is not just sold in the$ > US.c  J Good link, thanks.  I thought that the name sounded familiar, and the siteI refreshes that blurry memory from my earlier interest in old cars. Thanks= Paul.l  G By the way, they cite "my-bok" as the pronunciation.  Since the word iscL German, I would have expected the "bach" part to be pronounced like Bach (ofK J.S. fame).  You maybe have better insight into the German language than I.v  
 Dr. Dweeb.   ------------------------------  # Date: Wed, 12 May 2004 09:07:14 GMTu- From: JF Mezei <jfmezei.spamnot@teksavvy.com> / Subject: OT: Sun completes  Niagara chip designa@ Message-ID: <ae23ad57d5fe256e0e2f802f6d40e005@news.teranews.com>   This is from a cnet article at:e9 http://news.com.com/2100-1006_3-5210561.html?tag=nefd.topw  M It is the last paragraph that is most interesting. (Note absence of a mention  of Alpha though)   ##M Sun Microsystems has completed the design of its Niagara processor, a crucial K product in the server maker's effort to keep its own UltraSparc chip familye6 competitive, a source familiar with the project said.   K Sun acquired the Niagara design when it bought start-up Afara Websystems ineG 2002. The processor is part of Sun's UltraSparc family--an increasingly N important part after Sun canceled two other UltraSparc models and embraced the> rival "x86" chip lineage of Intel and Advanced Micro Devices.   K Sun "taped out" the Niagara design Tuesday, meaning that its design is donetM and has been sent to the chip fabrication facility for prototypes to be builtaJ and tested, the source said. A Sun representative confirmed the milestone.N Typically, at least a year is required after tape-out to fix manufacturing and= other issues before a chip actually can be used in a server. h  G Niagara is notable for an unusual design that includes eight processing H engines, or cores, each of which is capable of handling four instructionG sequences called threads. Sun variously calls this approach "throughputi% computing" or "chip multithreading." o  L Niagara and a higher-end cousin code-named Rock are still under development,M but Sun canceled development of two other chips this year: the low-end GeminisC model with dual UltraSparc II cores and the high-end UltraSparc V. -  K At the same time, Sun is aggressively adopting AMD's Opteron processor. TheKF Santa Clara, Calif.-based company also is putting more emphasis on itsJ software, which runs on both chip types and in the future also will run on, competitors' servers with other processors.   K Texas Instruments builds Sun's processors. UltraSparc chips compete chieflytM with IBM's Power processors, Intel's Itanium processors and Hewlett-Packard's0 PA-RISC processors.e ##   ------------------------------    Date: 12 May 2004 07:37:47 -0700. From: fabiopenvms@yahoo.com.br (Fabio Cardoso)3 Subject: Re: OT: Sun completes  Niagara chip designa= Message-ID: <f30679fb.0405120637.19bad43a@posting.google.com>y  u JF Mezei <jfmezei.spamnot@teksavvy.com> wrote in message news:<ae23ad57d5fe256e0e2f802f6d40e005@news.teranews.com>...6! > This is from a cnet article at:s; > http://news.com.com/2100-1006_3-5210561.html?tag=nefd.topo > O > It is the last paragraph that is most interesting. (Note absence of a mentiono > of Alpha though) >  > ##O > Sun Microsystems has completed the design of its Niagara processor, a crucialLM > product in the server maker's effort to keep its own UltraSparc chip familym8 > competitive, a source familiar with the project said.  > M > Sun acquired the Niagara design when it bought start-up Afara Websystems innI > 2002. The processor is part of Sun's UltraSparc family--an increasingly P > important part after Sun canceled two other UltraSparc models and embraced the@ > rival "x86" chip lineage of Intel and Advanced Micro Devices.  > M > Sun "taped out" the Niagara design Tuesday, meaning that its design is done O > and has been sent to the chip fabrication facility for prototypes to be builtrL > and tested, the source said. A Sun representative confirmed the milestone.P > Typically, at least a year is required after tape-out to fix manufacturing and? > other issues before a chip actually can be used in a server. U > I > Niagara is notable for an unusual design that includes eight processingwJ > engines, or cores, each of which is capable of handling four instructionI > sequences called threads. Sun variously calls this approach "throughputh' > computing" or "chip multithreading."      D What about the License Model based in number of processors x Core ? ( Would Oracle become much more expensive?   Regardse   FC b    N > Niagara and a higher-end cousin code-named Rock are still under development,O > but Sun canceled development of two other chips this year: the low-end Gemini E > model with dual UltraSparc II cores and the high-end UltraSparc V. o > M > At the same time, Sun is aggressively adopting AMD's Opteron processor. ThesH > Santa Clara, Calif.-based company also is putting more emphasis on itsL > software, which runs on both chip types and in the future also will run on. > competitors' servers with other processors.  > M > Texas Instruments builds Sun's processors. UltraSparc chips compete chieflyuO > with IBM's Power processors, Intel's Itanium processors and Hewlett-Packard's  > PA-RISC processors.t > ##   ------------------------------  # Date: Wed, 12 May 2004 14:10:44 GMT 1 From: Michael Austin <maustin@firstdbasource.com>m0 Subject: Re: Parity errors when HSG80 restarted?2 Message-ID: <40A230D3.B36F479F@firstdbasource.com>   Malcolm,  Q I do not know when you posted the original message, but basically the metadata onnP the mirror/raid/disk is toast.  Been there done that!!! and this is how you undoM it.  The command that should have been executed on the controller was RESTARTnL OTHER from the  controller NOT being restarted.  In order to get out of thisI problem, delete the unit, initialize the mirror/raid/disk  (this does NOT1M overwrite any data!!) add the unit back and assign any selective presentationj connections.   set unit dx dis=all 	 delete dx4 init {rx | mx | diskx}$ add unit dx {rx | mx | diskx}dis=all set unit dx id=<somenumber> - set unit dx ena=conn1,conn2,conn3,conn4 etc..@   on the system, mc sysman set env/clus do mc sysman io auto/log) do mount/system $1$dbax <label> <logical>e exit  L BTW, on the HSG80, I have moved whole raidsets to another entirely differentO controller and different slots, followed these procedures and continued withoutsI problems.  In fact in a another instance, moved all of the disks from oney- controller to another, not just one raid set.s   Michael Austin.y     "David J. Dachtera" wrote:   > Malcolm Dunnett wrote: > >oG > >  I had an odd thing happen yesterday ( at least I think it's odd ).n > >2B > >  VMS 7.3-2 (with the VMS732_FIBRE_SCSI V2.0 patch) on a DS20E,J > > disks on a dual controller RA8000 ( ACS 8.6P-13 ) in Multibus failover? > > mode, KGPSA-B (Emulex LP7000) FC adapters, switched fabric.sI > > All possible FC paths are configured ( 2 adapters in the DS20E, goingLO > > to 2 separate FC fabrics, going to the A/B ports on each HSG80 controller).oK > > There are no other devices in this zone on the fabric switches ( thoughi5 > > there's a separate cluster in a different zone ).o > >lI > >  I restarted one of the controllers on the HSG80 ( using RESTART THIShI > > command from the console ). It restarted normally. The disks appearedaF > > to fail over to the other controller as would be expected ( and asD > > was logged in the system error log ). However my Oracle databaseG > > crashed, saying it got parity errors during a write on all 3 of thelH > > disks where I have mirrored versions of the control file ( all 3 areK > > on the RA8000 ). There didn't appear to be any damage to the data, when P > > I restarted the database it recovered normally ( no media recovery needed ). > > M > >  I thought the failover between controllers on the RA8000 was supposed tohM > > be transparent to the application software, I didn't expect to see parityiI > > errors being reported. I don't recall this being an issue other timesaJ > > I've restarted a controller, but that has been with VMS 7.3-1 systems.H > > Is this a new behaviour/bug in 7.3-2 or indicative of a problem withF > > this configuration ( or just something that's always happened that > > I've not noticed before) ? >dA > *** WARNING *** WARNING *** WARNING *** WARNING *** WARNING ***5 >AI > There is a problem with the FC-SCSI patch V5 for V7.3-1. Do not installaF > this patch, and if you have installed it, back it out IMMEDIATELY!!! > H > Do not attempt to apply the replacement patch (V6) as it has also been > recalled.n >kA > *** WARNING *** WARNING *** WARNING *** WARNING *** WARNING ***s >eJ > Also, contact your support rep. about an update to the HSG firmware, but$ > be ready for some unexpected news. >a > -- > David J. Dachtera  > dba DJE Systemsd > http://www.djesys.com/ > * > Unofficial Affordable OpenVMS Home Page:! > http://www.djesys.com/vms/soho/i   ------------------------------    Date: 12 May 2004 07:16:27 -0700. From: fabiopenvms@yahoo.com.br (Fabio Cardoso)& Subject: Processor Based License Model= Message-ID: <f30679fb.0405120616.3dc21a31@posting.google.com>   I Do you know what can change in the license model if the dual / multi-coren* processors begin to ship in high volumes ?  : For example: the super-Expensive Oracle RDB license model.  D If we buy a machine with 4 processors but each of them with 2 cores.$ Are we buying 4 or 8 "processors" ?   A Or another way.. imagine if in the near future instead of buying  C a specifc numbers of processor we could buy a matrix of 4 x 4 cores-: for example ! Or a photonic or nanotechnology processor ?   0 This model of licensing will not work anymore !      RegardsT   FC   ------------------------------  % Date: Wed, 12 May 2004 10:33:21 -0400b# From: "John Smith" <a@nonymous.com>7* Subject: Re: Processor Based License Model, Message-ID: <ytudnZ9RNPYuqD_d4p2dnA@igs.net>  ; "Fabio Cardoso" <fabiopenvms@yahoo.com.br> wrote in message 7 news:f30679fb.0405120616.3dc21a31@posting.google.com...oK > Do you know what can change in the license model if the dual / multi-coreh, > processors begin to ship in high volumes ? >e< > For example: the super-Expensive Oracle RDB license model. >)F > If we buy a machine with 4 processors but each of them with 2 cores.% > Are we buying 4 or 8 "processors" ?p >rB > Or another way.. imagine if in the near future instead of buyingE > a specifc numbers of processor we could buy a matrix of 4 x 4 coresn; > for example ! Or a photonic or nanotechnology processor ?e >t1 > This model of licensing will not work anymore !     @ Don't worry about it - Oracle will find a way to charge you moreI irrespective of how many processor cards or processors per core you have.o   ------------------------------  % Date: Wed, 12 May 2004 08:58:28 -0700,+ From: "Barry Treahy, Jr." <Treahy@MMaz.com>n* Subject: Re: Processor Based License Model' Message-ID: <40A249A4.4090002@MMaz.com>e   Fabio Cardoso wrote:  J >Do you know what can change in the license model if the dual / multi-core+ >processors begin to ship in high volumes ?  > ; >For example: the super-Expensive Oracle RDB license model.  >mE >If we buy a machine with 4 processors but each of them with 2 cores.o% >Are we buying 4 or 8 "processors" ?   >yB >Or another way.. imagine if in the near future instead of buying D >a specifc numbers of processor we could buy a matrix of 4 x 4 cores; >for example ! Or a photonic or nanotechnology processor ? h > 1 >This model of licensing will not work anymore !   >h >  r > E I do know that for PC software, we're already had problems.  We have tH engineering design software that is ONLY available for Windows and they A crucify you with a major license increase if you run on multiple tI processors.  Well guess what the Intel P4 with HT looks like to Windows, p4 and to the software running on it?  You guessed it!   F I agree with you that the licensing models must be rethought, but I'm I sure that in the eyes of the CA's, Oracle's, HP's, etc., that would mean  A they are being asked to 'leave money on the table' no matter how 1G short-sighted their perspective might actually be...  I can't see them o( making that choice quickly, or at all...   Barrye   -- e  > Barry Treahy, Jr                       E-mail: Treahy@MMaz.com> Midwest Microwave                          Phone: 480/314-1320> Vice President & CIO                         FAX: 480/661-7028                            ------------------------------  # Date: Wed, 12 May 2004 17:14:41 GMTe% From: "John Vottero" <John@mvpsi.com> * Subject: Re: Processor Based License Model? Message-ID: <5Ysoc.6762$eH1.3385927@newssvr28.news.prodigy.com>D  7 "Chuck Chopp" <ChuckChopp@rtfmcsi.com> wrote in messagea1 news:Nxsoc.1523$U4.1279@bignews6.bellsouth.net...g > Barry Treahy, Jr. wrote: > H > > I do know that for PC software, we're already had problems.  We haveK > > engineering design software that is ONLY available for Windows and theysD > > crucify you with a major license increase if you run on multipleL > > processors.  Well guess what the Intel P4 with HT looks like to Windows,7 > > and to the software running on it?  You guessed it! I > > I agree with you that the licensing models must be rethought, but I'mcL > > sure that in the eyes of the CA's, Oracle's, HP's, etc., that would meanD > > they are being asked to 'leave money on the table' no matter howJ > > short-sighted their perspective might actually be...  I can't see them, > > making that choice quickly, or at all... >lI > And the really irritating aspect of a hyperthreading CPU appearing as a  dualI > CPU SMP system under Windows is that you don't necessarily get the samefJ > performance increase as compared to having 2 separate CPUs with one coreK > each.  If the on-board cache isn't increased and some of the other issues(K > associated with memory access aren't resolved, you'll get a mult-core CPUDK > that can't compete performance-wise with separate CPUs.  Things like NUMAaH > come to mind, along with instruction pre-fetching, pipelining, etc.... >vL > Maybe for databases what we really need is a license that permits X numberJ > of transactions per seccond to be processed by the database.  If you run itF > on hardware that is insufficient to process the licensed TPS volume, that'sI > fine, you didn't under buy on your licensing and can perform a hardwareaL > upgrade w/o increasing your licensing costs for the database.  But, if youG > do upgrade your hardware and the hardware could now have the database-J > executing at a TPS rating higher than you are licensed for, the database? > would simply throttle itself back to your licensed TPS limit.D >BJ > Wouldn't this make more sense for database software licensing?  It wouldC > certainly allow you to have a large SMP server and put a databaselL > application on it that doesn't monopolize the server's resources while notG > also paying through the nose to run that database application on that  largen	 > server.o  L It makes a lot of sense.  Microsoft has already started doing something likeI this.  MSDE is actually SQL Server with a throttle and a db size limit of J 2GB.  The throttle kicks in at around 5 connections.  You can redistributeK MSDE with your application (for free).  Lots of applications will work justfI fine with the throttle.  If performance becomes a problem, you just buy a  SQL Server license.a  K Man would I like to have something like this for Rdb.  Digital was close tohH that when they bundled runtime Rdb with VMS.  Selling Rdb was one of the! biggest mistakes Bob Palmer made.    >mK > And, there's even other ways around the multiple CPU issue on Windows and J > Linux systems, too.  Youd could implement server virtualization softwareJ > like VMware [most likely GSX/ESX server] where each virtual machine only hasdJ > a single CPU regardless of how many CPUs the host system has.  Then, youK > could distribute your database across those multiple virtual machines and E > run them in a "clustered" or "distributed" mode and thus bypass the0 license10 > fees associated with multiple CPUs per system. >i   ------------------------------  % Date: Wed, 12 May 2004 10:04:07 +0100e* From: "Richard Brodie" <R.Brodie@rl.ac.uk>* Subject: Re: Reserved TCP/IP Port Numbers?, Message-ID: <c7spa8$16lu@newton.cc.rl.ac.uk>  : "JF Mezei" <jfmezei.spamnot@teksavvy.com> wrote in message: news:3086a2ee2d10e3a2e962d8b2e335fb14@news.teranews.com...  P > Note however that it seems that microsoft has abused this list by reusing someN > ports for its own purposes, as demonstrated by recent faults in widnows that > allow viri to proliferate.  * Er, what exactly are you talking about JF?   ------------------------------    Date: 12 May 2004 06:17:38 -0600B From: clubley@remove_me.eisner.decus.org-Earth.UFP (Simon Clubley). Subject: RZ2DD-LS performance specifications ?3 Message-ID: <aOvGNirx28hs@eisner.encompasserve.org>2  I Does anyone have the average access time for the RZ2DD-LS, as well as anyR other performance information ??  L I've been able to find out that it's a 10,000 rpm drive and that's about it.   Thanks,d   Simon.   -- oB Simon Clubley, clubley@remove_me.eisner.decus.org-Earth.UFP       P SCO: Proudly pushing Microsoft down to #2 on the list of most disliked companies   ------------------------------  + Date: Wed, 12 May 2004 08:15:51 +0000 (UTC)nP From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply) Subject: Re: SHOW ROUTE $ Message-ID: <c7smfm$drs$1@online.de>  @ In article <40A18B77.B6D8A338@NeOaSrPtAhMlNiOnWk.net>, "David J.9 Dachtera" <djesys.nospam@NeOaSrPtAhMlNiOnWk.net> writes: w  1 > Phillip Helbig---remove CLOTHES to reply wrote: 
 > > [snip]6 > > TCPIP$CONFIG asks you if you want dynamic routing. > J > Based on experience, I don't recommend dynamic routing, unless it solvesD > a particularly sticky problem - it can cause more problems than it > solves, in my experience.I > I > "Permanent" (manually determied default) routes seem to serve best in a / > good number of cases, based on my experience.   F I agree.  I have one permanent route (apart from internal addresses), G that to my DSL router/switch/bridge.  (Unfortunately, this is a single   point of failure.)   ------------------------------  % Date: Wed, 12 May 2004 09:06:47 -0700A# From: "Tom Linden" <tom@kednos.com>  Subject: SPAM, etc9 Message-ID: <NDEMLKKEBOIFBMJLCECIOEMODDAA.tom@kednos.com>   8 In my never ending efforts to get rid of spam and virii,: I come across a lot I don't follow.  I am running MX5.3 on7 7.3-1 and I use Outlook as a pop client.  What shows up ) on the subject line appears intelligible,i  ; Stuart, You can delegate authority, but not responsibility.   8 Now, I am not stupid enough to open this type of mail in5 other than VMS mail, and looking at the 822 header it ; appears as gibbersih.  Is this perhaps a character encodings that Outlook understands  G Here is the header and then there is a .gif attachment, likely a virus.a= are =?  ?= delimiters that can be used to effectively filter?c  7 From:   SMTP%"dugzzfeabrrzwy@attbi.com"  "Clint Jepson"D To:     SMTP%"tom@kednos.com"I CC:g Subj: H =?utf-8?B?U3R1YXJ0LCBZb3UgY2FuIGRlbGVnYXRlIGF1dGhvcml0eSwgYnV0IG5vdCByZX Nwb25zaWJpbGl0eS4=?=  ' Return-Path: <dugzzfeabrrzwy@attbi.com>PG Received: from c-24-98-145-242.atl.client2.attbi.com (24.98.145.242) by.I           FREJA.KEDNOS.COM (MX V5.3 An1f) with SMTP for <tom@kednos.com>;N)           Wed, 12 May 2004 08:51:23 -0700o' Return-Path: <dugzzfeabrrzwy@attbi.com> ? Received: from gxuhpuyau [14.165.243.176] ([14.165.243.176]) bysK           c-24-98-145-242.atl.client2.attbi.com (3.6.3/attbi.com) with SMTPm idH           47740FE2 for <tom@kednos.com>; Wed, 12 May 2004 08:51:16 -0700( Message-ID: <A4786$5$21635d51@attbi.com>/ From: "Clint Jepson" <dugzzfeabrrzwy@attbi.com>i" To: "Romy Warner" <tom@kednos.com>   Press RETURN for more...   MAIL>u'     #10         12-MAY-2004 08:51:24.22r MAIL Subject:G =?utf-8?B?U3R1YXJ0LCBZb3UgY2FuIGRlbGVnYXRlIGF1dGhvcml0eSwgYnV0IG5vdCByZ  XNwb25zaWJpbGl0eS4=?=h% Date: Wed, 12 May 2004 08:51:16 -0700d MIME-Version: 1.0aA Content-Type: multipart/related; boundary="dn221635d51-attbi.com"n   --dn221635d51-attbi.comi Content-Type: text/html;         charset="iso-8859-1"+ Content-Transfer-Encoding: quoted-printable   > <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN"> <HTML><HEAD> ---i& Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).@ Version: 6.0.667 / Virus Database: 429 - Release Date: 4/23/2004   ------------------------------  % Date: Wed, 12 May 2004 12:00:01 +0100t9 From: Andrew Harrison <andrew_._remove_harrison@su_n.com>v* Subject: Re: SUN fails to advertise VMS...0 Message-ID: <c7t03h$bsb$1@new-usenet.uk.sun.com>   David Svensson wrote:  > Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> wrote in message news:<c7qp4d$jbd$1@new-usenet.uk.sun.com>.... >  >>John Smith wrote:e >> >>>d >>>JF, >>>iG >>>Perhaps their ad campaign is a reaction to Customer experiences likea >>>this: >>>t >>@ >>Sadly you would be incorrect, Sun started the HP Away campaign= >>before HP started their Solaris migarion campaign so unlessc; >>you have developed a working time machine the reaction isu >>all the other way arround. >>< >>>http://www.hp.com/hpinfo/newsroom/press/2004/040211a.html4 >>>"Belkin Plugs in HP Superdome System; Unplugs Sun >>>/ >  > F > If I understand english correctly I don't see that a time machine is? > needed in this case. He said "experiences like this: ", and IrF > interpret that as the reaction does not have to do directly with theH > press release mentioned, but implied a more general view on the issue.  E How can an add campaign from Sun that started before HP started their ? current round of attempts to get Sun customers to migrate to HPs  be a reaction to HP's campaign ?  @ The only way as I pointed out would be for Sun to have developed: a working time machine, gone forward 6 months etc etc etc.   Regardsr Andrew Harrisonm   ------------------------------    Date: 12 May 2004 04:11:33 -07006 From: andrew.rycroft@intrinsitech.com (Andrew Rycroft)8 Subject: TCP/IP routing - which is the correct behaviour= Message-ID: <58ba0101.0405120311.1c479de4@posting.google.com>t   Hi,s  C I have a PC connected via an extended network to an openVMS server,t and othre servers.  C The OpenVMS server and the other servers each have 2 network cards.sC One card connected to the internet, and the other the the corporatee network.  C All the servers have static routes, to route to the internet and toh the internal network.-  E From my PC I can connect to the internal network address as one would0B expect. I CANNOT connect to the external interface. I believe thisE also to be correct since the static route to the PC subnet is defined2? as being via the internal network interface. Dynamic routing is@	 disabled.o  D On the other servers I can connect to both the internal and external7 network cards, but the routing database looks the same.   7 Please can someone clarify which behaviour is correct ?e  C Is there anyway I can get the OpenVMS server to behave similarly tor the other servers ?S   Thanks Andrew   ------------------------------    Date: 12 May 2004 10:09:23 -0700. From: fabiopenvms@yahoo.com.br (Fabio Cardoso)& Subject: Re: TNSPING for OpenVMS Alpha= Message-ID: <f30679fb.0405120909.34e21e82@posting.google.com>t  s fabiopenvms@yahoo.com.br (Fabio Cardoso) wrote in message news:<f30679fb.0405071025.3b073d88@posting.google.com>...s) > Is there a TNSPING for OpenVMS Alpha ? e > K > It would be good/useful utility for db-links tests (Oracle x Oracle RDB).i >  > 	 > Regardse >  > FC       Anyone from Oracle here ?   J By th way my db-links connections from Oracle to Oracle RDB still freezing8 sometimes ! I will open an case after the DBAs analysis.   Regards    FC   ------------------------------  % Date: Wed, 12 May 2004 19:22:44 +0200e% From: "Jakob Erber" <erberj@yahoo.de>u7 Subject: VMS Linker Question - multiple defined symbolsn) Message-ID: <c7tmj4$hot$1@newshispeed.ch>r   Hello,  1 could somebody please help me with this question:   I I need to link a process which is using DCE and SSL at the same time. ThesK problem is, that the shareables of both products define the function CRYPT,o: which results in a multiple defined symbol linker warning.  F Knowing, that we need no DCE SSL features, this would not be much of aK problem, an interactive program works fine despite of this warning (we link D so that the Symbol CRYPT is first found in the SSL shared lib by the linker).  E But if we want to build our program in a shareable itself and load ituL dynamicly to another programm, it results in a dynamic load problem (key not found in tree).h  J So we strongly suspect, that we have to get rid of the linker warning. HowJ can we direkt the linker to use the symbol in the SSL library and not warn* about the CRYPT symbol in the DCE library?  & I hope one can understand what I mean.   best regards   Jakob    ------------------------------    Date: 12 May 2004 00:38:45 -0700. From: martinkirby12@yahoo.co.uk (Martin Kirby) Subject: Re: XDMP and TCPIP 5.4s< Message-ID: <224291b.0405112338.35905d05@posting.google.com>  D Given that you get the login box I think this is not a TCP/IP / XDMPF issue but something in the change of DECwindows from V1.2-? to V1.3-1.  9 > Warning: sm.num_hosts is 5 but sm.host_list has 1 hostsS2 > Xlib: connection to "_WSA445:" refused by server  C There are two resources used to hold the set of authorised users asoD set by the security dialog. "host_list" is a comma separated list ofE transport/node/user triplets and "num_hosts" is a count of the number C of entries in the list. In V1.3 I added that warning if the two didnA not match - which should only happen if they have been changed inn# dt.resources for the saved session.   > Actually, the resources are numHosts and hostList, for the newF desktop, but it looks like the error message was incorrectly cut&paste from the traditional desktop.d   > Xlib: No protocol specifiedh  @ When a connection is refused by the server the server provides aE string in the response. This string "No protocol specified" is cominghD from the server. I assume it is saying that rather than that a magicE cookie, kerberos or... connection the attempt had been made without asE protocol specified so it would only allow access on the basis of hosth name.t   --------  C The mechanism that DECwindows uses to pass the security settings to < the server is a DECwindows extension. If the server is not aA DECwindows server then none of this works and the security should C remain as that you had set for the server before starting the loginoA process. This shouldn't have changed between DECwindows versions.c  D If your display server can produce a log file that might be useful.    Also, you may like to add:  3     $ decw$loginlog == "sys$manager:decw$login.log"t  E to SYS$MANAGER:DECW$PRIVATE_APPS_SETUP.COM (copy from TEMPLATE if you D don't have it) and restart DECwindows. This will get a login log and1 it seems to be login that is causing the problem.,  D Another thing to try is the traditional desktop login instead of the new desktop. I.e., add  '     $ decw$start_new_desktop == "false"$  ) to the apps setup and restart DECwindows.o   ------  D Had a look at the code and I think I can see an error in the way theD "don't make the change if this is not DECwindows" code is formatted.@ The way it is implemented is by establishing an X error handler,E trying the change and rolling back if an error comes in. However, theyE place where the error handler was removed has been moved to later andrE that also moved the XSync call. The consequence may be that the error>@ is coming in after the check and the roll-back is not happening.  = If you have a support contract you may want to raise a case. r   Martin Kirby   ------------------------------    Date: 12 May 2004 01:07:00 -0700. From: martinkirby12@yahoo.co.uk (Martin Kirby) Subject: Re: XDMP and TCPIP 5.4r< Message-ID: <224291b.0405120007.253d116d@posting.google.com>  @ I had another look at the code. If my theory is correct then theE Traditional desktop login should work as it does not include the samea
 coding error.k   Martin Kirby   ------------------------------  % Date: Wed, 12 May 2004 07:37:30 +0200n* From: "Karsten Nyblad" <nospam@nospam.com>2 Subject: Re: You'll never guess what HP advertised- Message-ID: <c7sd6u$31bf$1@news.cybercity.dk>o  + <david20@alpha2.mdx.ac.uk> wrote in messagey# news:c7rr3c$fng$1@news.mdx.ac.uk...uD > The point was made that VMS clustering had no place because it was@ > not fail-over "clustering" but didn't provide fault tolerance.E > Whether HP had 3 or 10 operating systems spanning that range is not  relevent! > to the point that I was making.rJ > If VMS's solution has no place why would SUN and other Unix vendors have been4 > pursuing VMS style clustering for the last decade.  G The question is not whether or not VMS is a good operating system.  The J problem is that VMS is difficult to market without doing any harm to HP-UX and NonStop.   Karsten Nyblad ibpit1202 at sneakemail dot comi   ------------------------------  % Date: Wed, 12 May 2004 07:49:25 +0200,* From: "Karsten Nyblad" <nospam@nospam.com>2 Subject: Re: You'll never guess what HP advertised+ Message-ID: <c7sdtd$bn$1@news.cybercity.dk>   H "Bob Koehler" <koehler@eisner.nospam.encompasserve.org> wrote in message- news:tyVVmX2cI6y3@eisner.encompasserve.org...i@ > In article <c7qnkg$17ck$1@news.cybercity.dk>, "Karsten Nyblad" <nospam@nospam.com> writes:  > >sI > > IBM has used 50 years to build customer trust in that if something isnH > > supported by IBM, then they will not leave you while you are in deeb shit. 6 > > HP has already canceled a couple of product lines. >s2 >    IBM has canceled the 360 in favor of the 370.  K I think the 370 could run programs from the 360.  As far as I know, this isb@ a bit like saying that Digital canceled VMS in favor of OpenVMS.  8 >   They haven't supported 026 card punches for decades.  H Last time I checked, MVS still using 80 columns fixed length records for many things.   Karsten Nyblad ibpit1202 at sneakemail dot com    ------------------------------  % Date: Wed, 12 May 2004 08:19:52 +0200$* From: "Karsten Nyblad" <nospam@nospam.com>2 Subject: Re: You'll never guess what HP advertised, Message-ID: <c7sfm9$1su$1@news.cybercity.dk>  H "Bob Koehler" <koehler@eisner.nospam.encompasserve.org> wrote in message- news:yIuQmjIaUqkz@eisner.encompasserve.org...t@ > In article <c7qm9e$15ot$1@news.cybercity.dk>, "Karsten Nyblad" <nospam@nospam.com> writes:i >aI > > Please do not mix up how people make decision on buying products thatg aregJ > > mainly sold on image with products like operating systems where people arecA > > much more likely to base their decision on technical reasons.m > E >    Considering the market share of Windows, I don't see how you can:) >    even start to type that last phrase.e  F OK, First I think some people in Usenet group are pro VMS anti WindowsJ zealots.  Yes, Windows could do with a redesign for better security.  Yes,J the documentation sucks.  Yes, it is impossible to correct problems on theJ platform, because you do not get informative error messages, Yes, Yes, ...) But it is reliable enough for a desk top.E  L Secondly, Windows is also a cheap platform that is suitable for running deskF top applications.  The only other cheap platform is Linux and Linux is barely ready for the desk top.   Karsten Nyblad ibpit1202 at sneakemail dot coms   ------------------------------  % Date: Wed, 12 May 2004 08:53:03 +02000* From: "Karsten Nyblad" <nospam@nospam.com>2 Subject: Re: You'll never guess what HP advertised, Message-ID: <c7shkk$402$1@news.cybercity.dk>  + "Dr. Dweeb" <dr@dweeb.com> wrote in message ' news:c7rje1$2cuf$1@news.cybercity.dk...? > Karsten Nyblad wrote:hE > > The market for luxury items is about showing how much money have.R >aE > That is true for some, but certainly not true for all the items youv considerH > luxury.  You probably think a Mercedes Benz is a luxury car, yet it is@ > probably the best value proposition new car purchase there is. > K > We I inclined, I could prove to you that the best value proposition for arH > car owner is to buy a Mercedes Benz, rather than your Lada or skoda orK > whatever it is you thinkn we should all drive.  Anyone who has done fleetrE > management understands the sums - for Joe Familyman and the correcte > preconditions - buy a Merc.   L Feel free to think that your decission on which car to drive is solely based on rational reasons.  L > > It is not about the quality or true value of these products.  Thus, richI > > people buy Maybachs, Ferraris, etc. to show of their money.  The moreaB > > expensive the car is, the more valuable it is to its customers3 > > regardless of what it costs to produce the car.e >gJ > You must be a f*****g Dane.  That country is full of envious people with > opinions just like yours.w  L LOL.  You should watch your language.  It is you who claim that I envy thoseH people.  They can use their money as they want as far as I am concerned.L People use an amasingly large part of their wages for cars, and that is alsoJ true for blue collar workers, students, people on social welfare, etc.  AnF important reason for using all that money is to send an image to other persons.  B > Just what is a MayBach ?? - I will need to resaerch this one :-)  = It is Daimler-Chryslers car in the Bentley/R&R class of cars.m   Karsten Nyblad ibpit1202 at sneakemail dot comc   ------------------------------  % Date: Wed, 12 May 2004 09:47:44 +0200c* From: Paul Sture <nospam@sture.homeip.net>2 Subject: Re: You'll never guess what HP advertised* Message-ID: <2ge352F1oagfU1@uni-berlin.de>   Karsten Nyblad wrote:   E  > How many really need to transport big things more than a couple ofiD  > times a year?  How many need 4-wheel drive (except for driving inG  > snow a few times a year)?  How many need sports cars considering thesE  > speed limits?  Most men would be better of with a sedan, but sedantG  > does not send the right image.  A company like VW sells the same carrF  > with small modifications using three brands, typically Audi, VW andE  > Skoda, and at very different price.  Why? Because that allows them <  > to sell the same car to three different groups of people.  >K > Please do not mix up how people make decision on buying products that are L > mainly sold on image with products like operating systems where people are? > much more likely to base their decision on technical reasons.  >   G I cannot agree with you there. IBM for one definitely use the prestige oA factor when selling their solutions, and its effect is not to be uI underestimated. Microsoft certainly used the prestige factor in the late lE 1990s. I lost count of the number of times I heard the phrase "We're rI going Microsoft, of course" back then; many with purchasing power wanted c* to be associated with Bill Gates' success.  D Prestige is very important to customers, whether it be at a company F level or divisional / departmental levels within a company. Of course H they'll choose a short list of products according to cost and technical > merits, but prestige is an important part of the final choice.   ------------------------------  % Date: Wed, 12 May 2004 10:14:07 +0200 * From: Paul Sture <nospam@sture.homeip.net>2 Subject: Re: You'll never guess what HP advertised* Message-ID: <2ge4mhF1jcsdU1@uni-berlin.de>   Karsten Nyblad wrote:t4 > "Main, Kerry" <kerry.main@hp.com> wrote in message   > M > Windows might be crap, but it is on that platform all the cool desktop usertM > applications are.  And then it is cheap compared to other platforms.  ApplehJ > might have some cool computers, but you end up paying considerably more.  G I'm just looking at laptop prices here, but every time I see an advert  C in a glossy magazine for say a Dell, and then compare prices for a eI roughly equivalent Apple, I just don't come up with "considerably more", nC especially when I don't need to add Office to get word processing, T+ spreadsheet, and presentation applications.n   ------------------------------  % Date: Wed, 12 May 2004 10:06:29 +0200r  From: "Dr. Dweeb" <dr@dweeb.com>2 Subject: Re: You'll never guess what HP advertised, Message-ID: <c7slu2$80d$1@news.cybercity.dk>   Karsten Nyblad wrote:a- > "Dr. Dweeb" <dr@dweeb.com> wrote in messagee) > news:c7rh15$2a74$1@news.cybercity.dk...b >> Karsten Nyblad wrote:D >>> "Bob Koehler" <koehler@eisner.nospam.encompasserve.org> wrote in9 >>> message news:GeHCYM$fmhMj@eisner.encompasserve.org...rC >>>> In article <c7opdj$2a11$1@news.cybercity.dk>, "Karsten Nyblad"   >>>> <nospam@nospam.com> writes: >>>>G >>> When you chose an operating system you want to be sure that you canbF >>> buy new systems for many years.  Thus you want to be sure that theD >>> vendor actually wants to continue active development and sellingH >>> systems at competitive prices.  It is no easy decission to move to aD >>> new operating system if you have a million lines of code as many >>> companies have.e >>>  >>@ >> A million - You sound like Dr. Evil !  He had no clue either!C >> You are a few zeroes short for a very large number of companies.e >h  F Dr. Evil is a character from the Austin Powers films.  In one film heL attempts to hold the world to ransom for 1 million dollars.  At the time, heI owns a multi-billion dollar company.  Dr. Evil, having been in cryo-sleepIL for a few decades just had no clue that a million was not a very big number.   Are you with it ? :-)-   > Gee.  What a language. >0F >>> Chosing an operating system is also chosing a business partner forD >>> years to come.  When you chose business partners, you try to getF >>> partner that want to take the partnership in the same direction as >>> you do.l >>B >> Explain the paradox of boosting Linux (as you do) and using theD >> "business partner" argument against proprietary OSs as you do in >> this paragraph. >>H > OK, Change "choosing an operating system" into "choosing a proprietaryB > operating system."  Linux is not ready for the high availabilityC > market yet. I am not trying to boost Linux at the expense of VMS.!C > Beside, nobody would use Linux for 24/7 systems without getting anE > support contract.  Then the supporting company, e.g., RedHat, SuSE,t" > or SUN is your business partner. >>> H >>>>    Funny how these multiple product arguments only get used againstF >>>>    VMS. IBM sells MVS, AIX, Linux, ..., and nobody calls them not  >>>>    commited to all of them. >>> F >>> IBM has used 50 years to build customer trust in that if somethingF >>> is supported by IBM, then they will not leave you while you are inA >>> deeb shit. HP has already canceled a couple of product lines.  >>>e >>F >> Would you like to know about the cancelled IBM product lines ??   IC >> suspect you are too young to know too much.  I am sure there are E >> others with long memories who can tell you about different machineoE >> architectures that have come and gone - created by IBM.  There are E >> plenty of them.  Some of them lasted a long time, others less, buti@ >> they died and IBM helped customers to another product line orF >> customers went somewhere else.  This product life-cycle is true forG >> all vendors - IBM is not immune and makes no such claims.  They kill G >> products too and leave customers in exactly the amount of shit whicht2 >> they can legally and financially get away with. >uC > Yes and so?  It does not change the fact that customers trust IBMo! > more than they trust, e.g., HP.T >h  I While that is, I believe true, one can scarecly draw that conclusion from I your statement, which suggests that IBM does not cancel product lines - a  factually incorrect statement.  F >>> And how many of these are actively sold?  The first version of the@ >>> 2.6 kernel was only for x86.  Besides, there is an importantB >>> difference between open source and closed source software.  InG >>> later case you never know if the vendor decides to stop the productaB >>> and drop support.  That can get you in real troubles.  On openG >>> source you know that if nobody else will support the code, then yout >>> can do it yourself.  >>B >> The number of sites with the skills in house to maintain OSs isF >> ridiculously small compared to the number of sites using computers.D >> This is a specious argument that serious OpenSource boosters have) >> long since relegated to the scrap bin.i >iE > True.  But most bugs can be fixed without understanding more than a  > few hundred lines of code. >t  J I leave it to OS engineers to comment upon the veracity of this statement.   > Karsten Nyblad! > ibpit1202 at sneakemail dot com    ------------------------------  % Date: Wed, 12 May 2004 10:11:38 +0200e  From: "Dr. Dweeb" <dr@dweeb.com>2 Subject: Re: You'll never guess what HP advertised, Message-ID: <c7sm7m$8do$1@news.cybercity.dk>   Karsten Nyblad wrote:*B > "Bob Koehler" <koehler@eisner.nospam.encompasserve.org> wrote in7 > message news:yIuQmjIaUqkz@eisner.encompasserve.org...rA >> In article <c7qm9e$15ot$1@news.cybercity.dk>, "Karsten Nyblad"b >> <nospam@nospam.com> writes: >>D >>> Please do not mix up how people make decision on buying productsF >>> that are mainly sold on image with products like operating systems? >>> where people are much more likely to base their decision on  >>> technical reasons. >>F >>    Considering the market share of Windows, I don't see how you can* >>    even start to type that last phrase. >eH > OK, First I think some people in Usenet group are pro VMS anti WindowsF > zealots.  Yes, Windows could do with a redesign for better security.A > Yes, the documentation sucks.  Yes, it is impossible to correctrD > problems on the platform, because you do not get informative errorC > messages, Yes, Yes, ... But it is reliable enough for a desk top.  >l  7 There are zealots everywhere, including here.  However;   : Just because you are a zealot does not mean you are wrong.J Just because you are paranoid does not mean someone is not out to get you. ...e   You get the picture ?2  A > Secondly, Windows is also a cheap platform that is suitable forrB > running desk top applications.  The only other cheap platform is3 > Linux and Linux is barely ready for the desk top.  >s  G Getting better though.  I have Xandros version running and it is prettyoH good.  It does not support a primary requirement of mine, so it is still# just a plaything running in VMWare.    > Karsten Nyblad! > ibpit1202 at sneakemail dot coms   ------------------------------  % Date: Wed, 12 May 2004 07:09:55 -0400 ' From: "Main, Kerry" <kerry.main@hp.com>o2 Subject: RE: You'll never guess what HP advertisedR Message-ID: <FD827B33AB0D9C4E92EACEEFEE2BA2FB3145A9@tayexc19.americas.cpqcorp.net>   > -----Original Message-----4 > From: Karsten Nyblad [mailto:nospam@nospam.com]=20 > Sent: May 12, 2004 1:29 AM > To: Info-VAX@Mvb.Saic.Como4 > Subject: Re: You'll never guess what HP advertised >=20 >=20	 [snip ..]  >=20A > True.  But most bugs can be fixed without understanding more=20- > than a few > hundred lines of code. >=20 > Karsten Nyblad! > ibpit1202 at sneakemail dot comm >=20 >=20    A Yep, and all a good mechanic needs is a good screwdriver, hammer,cE crescent wrench and maybe a bit of duct tape to properly maintain anye car.  C Please, this is getting silly - you do not fix OS kernel, security, H cluster and driver problems by only understanding a few hundred lines ofH code. Especially ones which deal with loading, cluster interaction, raceE problems caused by faster servers, intermittent crashes (hw?/os?/isv?i software issues?) etc ...f   Regardsa  
 Kerry Main Senior Consultantc HP Services Canada Voice: 613-592-4660a Fax: 613-591-4477c Email: kerryDOTmainAThpDOTcomh. (remove the DOT's and AT for email address)=20   ------------------------------  + Date: Wed, 12 May 2004 11:54:27 +0000 (UTC)s From: david20@alpha2.mdx.ac.uk2 Subject: Re: You'll never guess what HP advertised) Message-ID: <c7t39j$rlo$1@news.mdx.ac.uk>s  Z In article <c7sd6u$31bf$1@news.cybercity.dk>, "Karsten Nyblad" <nospam@nospam.com> writes: >w, ><david20@alpha2.mdx.ac.uk> wrote in message$ >news:c7rr3c$fng$1@news.mdx.ac.uk...E >> The point was made that VMS clustering had no place because it was A >> not fail-over "clustering" but didn't provide fault tolerance.gF >> Whether HP had 3 or 10 operating systems spanning that range is not	 >releventv" >> to the point that I was making.K >> If VMS's solution has no place why would SUN and other Unix vendors haver >beenr5 >> pursuing VMS style clustering for the last decade.  >rH >The question is not whether or not VMS is a good operating system.  TheK >problem is that VMS is difficult to market without doing any harm to HP-UX-
 >and NonStop.- >-  D I don't see that. HP-UX has its strengths , VMS has it's strengths, A Tandem has it's strengths - hell even Windows has it's strengths. G By not pushing all its operating systems HP loses sales to competitors.uJ Sure some people who might have bought HP-UX or Nonstop might purchase VMSE but that is a lot better for HP than those users going to IBM or SUN.PM Indeed in the long run customers buying VMS are probably a lot more likely torL remain customers than those going to HP-UX. Traditionally a move to one UnixO platform this year has often led to a move to a different vendors Unix offeringlO in the future - a trend likely to increase as more and more Unix users consideru Linux.   ( F I doubt if you meant this but one reading of your statement above is :  M HP shouldn't market it's best product because it will hurt sales of its other  not so good products.tI That would seem to be rather stupid strategy since it would hamstring anyt7 company following that policy against it's competitors.a   )i  
 David Webb VMS and Unix team leader CCSS Middlesex University      t     >Karsten Nybladg  >ibpit1202 at sneakemail dot com >W >o >i   ------------------------------    Date: 12 May 2004 07:09:39 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) 2 Subject: Re: You'll never guess what HP advertised3 Message-ID: <H8S1f2KuRfAN@eisner.encompasserve.org>r  p In article <242b39f4397c01ef5e7522346b4ececc@news.teranews.com>, JF Mezei <jfmezei.spamnot@teksavvy.com> writes: > Bob Koehler wrote: >hE >>    IBM has canceled the 360 in favor of the 370.  They canceld theeC >>    original AIX/PC2 combo, and radically changed AIX before theycK >>    tried it again.  They haven't supported 026 card punches for decades.w >  > N > Sorry, but the 360 is still alive today. 370->370->390->whatever the name isI > this week is the same architecture, just like the pentium 4 is the samer@ > architecture as the first pentium or the 8086 for that matter.  F    No its not.  P4 uses the same instruction set as the first Pentium.@    Its a binary compatable superset of the 8086 under restrictedG    circumstances.  The architecure changed from 8086 to 80286 and again A    from 80286 to 80386.  Since 386 the architecture hasn't reallyt$    changed, just the implementation.  F    370 is still alive today, but is not using the same instruction set?    as 360.  370 is based on 360 and early 370 systems had a 360 B    compatability mode, but it is not 360.  The was an architectureA    change to go from 360 to 370, since 370 things have not really %    changed at the architecture level.    ------------------------------    Date: 12 May 2004 07:14:25 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)E2 Subject: Re: You'll never guess what HP advertised3 Message-ID: <xes7s3BXf$iQ@eisner.encompasserve.org>f  Z In article <c7scnd$30tn$1@news.cybercity.dk>, "Karsten Nyblad" <nospam@nospam.com> writes: > I > True.  But most bugs can be fixed without understanding more than a fewi > hundred lines of code.  
    ROTFLOL  9    You've been doing maintenance programing for how long?s   ------------------------------    Date: 12 May 2004 07:18:17 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)o2 Subject: Re: You'll never guess what HP advertised3 Message-ID: <SVn3yu+GLWXc@eisner.encompasserve.org>r  X In article <c7sdtd$bn$1@news.cybercity.dk>, "Karsten Nyblad" <nospam@nospam.com> writes:3 >>    IBM has canceled the 360 in favor of the 370.h > M > I think the 370 could run programs from the 360.  As far as I know, this issB > a bit like saying that Digital canceled VMS in favor of OpenVMS.  E    No.  The 360 had a compatability mode under early 370s.  It's more"D    like the VAX-11 series having a compatability mode for PDP-11 andF    saying digitial canceled PDP-11 in favor of VAX, except they didn't!    get around to it for 15 years.s  A    360 compatability mode on 370 is probably "easier" than PDP-11v8    compatability mode on VAX-11, but it's the same idea.  J > Last time I checked, MVS still using 80 columns fixed length records for > many things.  G    Yep.  But irrelevant.  I can do 80 column fixed length on Linux if I     just want to write the code.    ------------------------------    Date: 12 May 2004 07:23:20 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) 2 Subject: Re: You'll never guess what HP advertised3 Message-ID: <j15TNNINSJ+$@eisner.encompasserve.org>c  Y In article <c7sfm9$1su$1@news.cybercity.dk>, "Karsten Nyblad" <nospam@nospam.com> writes:e > H > OK, First I think some people in Usenet group are pro VMS anti Windows
 > zealots.  A    What doyou expect to find in comp.os.vms?  That's where you're     posting.y   >  Yes, Windowsd  ...+ > But it is reliable enough for a desk top.   >    That depends on what you're trying to do with your desktop.  N > Secondly, Windows is also a cheap platform that is suitable for running deskH > top applications.  The only other cheap platform is Linux and Linux is  > barely ready for the desk top.  H    Nonsense.  I use Linux desktops all the time.  I use Solaris desktopsG    sometimes.  I use Mac desktops.  OK, Mac and Solaris aren't quite as ?    cheap as a Wall-Mart PC, but Macs aren't all that expensive.n  G    And I'll keep my VMS desktops for which I've paid less than $200.  I A    don't happen to have Lotus/123 loaded right now, but I do haves    DECwrite and Latex.   ------------------------------  % Date: Wed, 12 May 2004 14:35:42 +0200t* From: "Karsten Nyblad" <nospam@nospam.com>2 Subject: Re: You'll never guess what HP advertised, Message-ID: <c7t5n1$rrh$1@news.cybercity.dk>  + <david20@alpha2.mdx.ac.uk> wrote in messageu# news:c7t39j$rlo$1@news.mdx.ac.uk...fH > I doubt if you meant this but one reading of your statement above is : > I > HP shouldn't market it's best product because it will hurt sales of itsD otherL > not so good products.oK > That would seem to be rather stupid strategy since it would hamstring any 9 > company following that policy against it's competitors..  J It is you who call it HP's best product.  HP-UX generates more sales so itA is a better product when it comes to getting a black bottom line.    Karsten Nyblad ibpit1202 at sneakemail dot como   ------------------------------  % Date: Wed, 12 May 2004 06:52:26 -0400r# From: "John Smith" <a@nonymous.com>p( Subject: [OT]: What's Sun really saying?, Message-ID: <DrCdnd8WXMF2nD_dRVn-uw@igs.net>  K Also, let me [be] really clear about our Linux strategy. We don't have one.sI We don't at all. We do not believe that Linux plays a role on the server.nJ Period. If you want to buy it, we will sell it to you, but we believe thatK Solaris is a better alternative, that is safer, more robust, higher quality 2 and dramatically less expensive in purchase price.K [Jonathan Schwartz, Executive VP - Software, Sun Microsystems, 19 Sep 2003]R3 http://www.eweek.com/article2/0,4149,1274614,00.aspo    F What's most disappointing is there appears to be no real commitment toI Linux. Are you, or aren't you, moving your desktops? I guess I just don't1 understand your Linux strategy.4I [Jonathan Schwartz, Executive VP - Software, Sun Microsystems, in an open2 memo to IBM, 21 Jan 2004]    ------------------------------   End of INFO-VAX 2004.263 ************************ <dugzzfeabrrzwy@attbi.com>i" To: "Romy Warner" <tom@kednos.com>   Press RETURN for more...   MAIL>u'     #10         12-MAY-2004 08:51:24.22r MAIL Subject:G =?utf-8?B?U3R1YXJ0LCBZb3UgY2FuIGRlbGVnYXRlIGF1dGhvcml0eSwgYnV0IG5vdCBy