1 INFO-VAX	Wed, 02 Jul 2003	Volume 2003 : Issue 362       Contents: Re: Another ? on GNV& Re: CNXMGRERR with V7.3-1 DSSI cluster6 Re: Compaq Solutions Aliance - Do it still existing  ?6 Re: Compaq Solutions Aliance - Do it still existing  ?6 Re: Compaq Solutions Aliance - Do it still existing  ?6 Re: Compaq Solutions Aliance - Do it still existing  ?6 Re: Compaq Solutions Aliance - Do it still existing  ?6 Re: Compaq Solutions Aliance - Do it still existing  ?6 Re: Compaq Solutions Aliance - Do it still existing  ?6 Re: Compaq Solutions Aliance - Do it still existing  ?6 Re: Compaq Solutions Aliance - Do it still existing  ? Re: cxx performance  Re: cxx performance  Re: cxx performance  Re: cxx performance  Re: cxx performance  Re: cxx performance  Re: cxx performance (IA64) Re: cxx performance (IA64) Enumerating logicals Re: Enumerating logicals( Re: HP Powers More TOP500 SupercomputersI Re: HP Webcast this morning on Next-Generation Intel Itanium 2 processors I Re: HP Webcast this morning on Next-Generation Intel Itanium 2 processors I Re: HP Webcast this morning on Next-Generation Intel Itanium 2 processors I Re: HP Webcast this morning on Next-Generation Intel Itanium 2 processors < More than VMS is required, but VMS is a great starting point Re: Mount a backup /image  Mount a backup /image  Re: New OpenVMS-to-Itanium FAQ RE: New OpenVMS-to-Itanium FAQ. Re: Open VMS Alpha upgrade from 7.2-1 to 7.2-2E Re: OpenVMS Host-Based MiniMerge Statement and Schedule for Customers ' Re: Porting C Code VAX -> Alpha -> IA64 ' Re: Porting C Code VAX -> Alpha -> IA64 ' Re: Porting C Code VAX -> Alpha -> IA64  Re: Rethinking  V.M.S  Re: Rethinking  V.M.S  Re: Rethinking  V.M.S  Re: Running VMS off CD- Standard Digital Markup Language (SDML) DTD ? J Re: Tri-architecture cluster demonstrated at DECUS Ottawa Technical Update Re: VMS backup to NFS shares Re: VMS backup to NFS shares  F ----------------------------------------------------------------------  % Date: Wed, 02 Jul 2003 16:32:10 +0200 ) From: Bart Zorn <B.Zorn@xs4all.nospam.nl>  Subject: Re: Another ? on GNV 6 Message-ID: <3f02eceb$0$49098$e4fe514c@news.xs4all.nl>  H BASH gets probably installed as a DCL command verb. PRODUCT REMOVE does  not undo that.  
 You could do:   . $ SET COMMAND/INPUT=SYS$LIBRARY:DCLTABLES.EXE-7    /OUTPUT=SYS$COMMON:[SYSLIB]DCLTABLES.EXE BASH/DELETE   @ Then do a INSTALL/REPLACE SYS$LIBRARY:DCLTABLES.EXE or a reboot.  	 Bart Zorn    John E. Malmberg wrote:  > Bradford J. Hamilton wrote:  > 3 >> In article <dzZJa.3260$Qj.166@news.cpqcorp.net>, A >> "John E. Malmberg" <Malmberg@dskwld.zko.dec.compaq.hp> writes:  >>K >> Thanks, John.  I successfully removed the product, first applying Thomas - >> Pauli's suggestion of $set proc/parse=ext.  >>J >>> Of course there is a matter of finding any command files that are now  >>> using GNV utilities. >> >>  I >> $bash now raises an error message, but I'm assuming this will go away   >> upon 7 >> reboot; if not, I'll have to do some looking around.  >  > I > I kind of doubt that rebooting will fix that since bash is a component   > supplied by GNV :-)  >  > -John # > malmberg@dskwld.zko.dec.compaq.hp  > Personal Opinion Only  >    ------------------------------  % Date: Wed, 02 Jul 2003 11:35:49 +0100 * From: Nic Clews <sendspamhere@[127.0.0.1]>/ Subject: Re: CNXMGRERR with V7.3-1 DSSI cluster ' Message-ID: <bducni$46g$1@lore.csc.com>    Martin Vorlaender wrote: >  > Keith Parris wrote: C > > "Martin Vorlaender" <martin.vorlaender@pdv-systeme.de> wrote...  > [...snip...] > > I > > You might learn more from having someone analyze the crash dumps, but  > B > I don't have anyone at hand, so I tried CCAT, but the CD versionC > (4.0-2.0) doesn't have it. Looking into "Ask OpenVMS", I found an H > article referring to VMS v5.3-2 (saying the problem is fixed in v5.4),G > and one referring to a  VAX 6000 CI cluster (that one at least saying J > the cause for CNXMGRERR was a "Sequence number mismatch between expected( > and received cluster message packets".  H Keith didn't mention that the actual point that CNXMGRERR drops out, theF comment in the source code has the reason, there are other reasons forG CNXMGRERR. I saw this, and the net effect was the systems were crashing 7 due to a symptom, which was actually a hardware issue.       --  ? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer Sciences  nclews at csc dot com    ------------------------------  + Date: Wed, 02 Jul 2003 10:42:48 +0200 (MET) 9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> ? Subject: Re: Compaq Solutions Aliance - Do it still existing  ? ; Message-ID: <01KXS3OYHP4YAPKEWT@sysdev.deutsche-boerse.com>   C > > Netscape is slowly tilting in the wind as a viable alternative; H > > Mozilla on VMS is  not current enough. Soon VMS users become 'lockedH > > out' of sites simply because those sites have less and less interestF > > in adhering to W3C standards as opposed to implementing the latestJ > > Microsoft "let's lock everyone into using our browser" by implementing > > a server-side 'feature'. > H >    So are you saying the Borg wins, somebody has to make Mozilla speakJ >    Borg, and Apple will have to make Safari speak Borg?  (Microsoft justG >    announced it will do no more Explorer for Mac since Apple is doing 
 >    Safari.)   F I agree.  While we can and should demand the availability of browsers F which support the latest HTML standard, we should NOT demand browsers I which support proprietary extensions.  Doing so is EXACTLY what the Borg   wants us to do.   I If the site is not readable due to proprietary extensions, send an email  G saying that still another customer has been lost due to this fact, and  F hint that the number of customers lost who don't write such emails is  probably much larger.    ------------------------------  $ Date: Wed, 2 Jul 2003 13:05:35 +0100* From: "John Travell" <john@travell.uk.net>? Subject: Re: Compaq Solutions Aliance - Do it still existing  ? 6 Message-ID: <bduhqh$10ia8n$1@ID-120847.news.dfncis.de>  : "Larry Kilgallen" <Kilgallen@SpamCop.net> wrote in message- news:L8EPt9HVZkUi@eisner.encompasserve.org... L > In article <01KXS3OYHP4YAPKEWT@sysdev.deutsche-boerse.com>, Phillip Helbig, <HELBPHI@sysdev.deutsche-boerse.com> writes: > K > >>    So are you saying the Borg wins, somebody has to make Mozilla speak H > >>    Borg, and Apple will have to make Safari speak Borg?  (Microsoft justJ > >>    announced it will do no more Explorer for Mac since Apple is doing > >>    Safari.) > > I > > I agree.  While we can and should demand the availability of browsers I > > which support the latest HTML standard, we should NOT demand browsers L > > which support proprietary extensions.  Doing so is EXACTLY what the Borg > > wants us to do.  > > L > > If the site is not readable due to proprietary extensions, send an emailJ > > saying that still another customer has been lost due to this fact, andI > > hint that the number of customers lost who don't write such emails is  > > probably much larger.  > 5 > I gather you have not actually tried this yourself.  > J > The DSPP folks who run this web site have a nasty habit of sending email8 > that is incompatible with VMSmail -- even to VMS ISVs.   Concur. D I lost a whole bundle of mail when Mozilla got confused about a mail- containing microsoft proprietary attachments.      -- John Travell  VMS crashdump expertise for hire john@jomatech.com  +44-(0)23-92552229 http://www.jomatech.com/       --- & Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).A Version: 6.0.493 / Virus Database: 292 - Release Date: 25/06/2003    ------------------------------   Date: 2 Jul 2003 06:50:02 -0500 - From: Kilgallen@SpamCop.net (Larry Kilgallen) ? Subject: Re: Compaq Solutions Aliance - Do it still existing  ? 3 Message-ID: <L8EPt9HVZkUi@eisner.encompasserve.org>   w In article <01KXS3OYHP4YAPKEWT@sysdev.deutsche-boerse.com>, Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> writes:   I >>    So are you saying the Borg wins, somebody has to make Mozilla speak K >>    Borg, and Apple will have to make Safari speak Borg?  (Microsoft just H >>    announced it will do no more Explorer for Mac since Apple is doing >>    Safari.) > H > I agree.  While we can and should demand the availability of browsers H > which support the latest HTML standard, we should NOT demand browsers K > which support proprietary extensions.  Doing so is EXACTLY what the Borg   > wants us to do.  > K > If the site is not readable due to proprietary extensions, send an email  I > saying that still another customer has been lost due to this fact, and  H > hint that the number of customers lost who don't write such emails is  > probably much larger.   3 I gather you have not actually tried this yourself.   H The DSPP folks who run this web site have a nasty habit of sending email6 that is incompatible with VMSmail -- even to VMS ISVs.   ------------------------------  + Date: Wed, 02 Jul 2003 14:52:12 +0200 (MET) 9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> ? Subject: Re: Compaq Solutions Aliance - Do it still existing  ? ; Message-ID: <01KXSCFXJMUAAM7Y4A@sysdev.deutsche-boerse.com>   L > > If the site is not readable due to proprietary extensions, send an emailJ > > saying that still another customer has been lost due to this fact, andI > > hint that the number of customers lost who don't write such emails is  > > probably much larger.  > 5 > I gather you have not actually tried this yourself.   D This was a general remark about my reaction to unreadable websites, G usually those which are unreadable for no reason other than "trying to  H be a k001 webprogrammer"---for example, using JavaScript for links when & HTML HREF links would do the same job.   ------------------------------   Date: 2 Jul 2003 08:11:28 -0500 - From: Kilgallen@SpamCop.net (Larry Kilgallen) ? Subject: Re: Compaq Solutions Aliance - Do it still existing  ? 3 Message-ID: <zGk7+AWR9JRt@eisner.encompasserve.org>   w In article <01KXSCFXJMUAAM7Y4A@sysdev.deutsche-boerse.com>, Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> writes: M >> > If the site is not readable due to proprietary extensions, send an email K >> > saying that still another customer has been lost due to this fact, and J >> > hint that the number of customers lost who don't write such emails is >> > probably much larger. >>6 >> I gather you have not actually tried this yourself. > F > This was a general remark about my reaction to unreadable websites, I > usually those which are unreadable for no reason other than "trying to  J > be a k001 webprogrammer"---for example, using JavaScript for links when ( > HTML HREF links would do the same job.  F Certainly.  My general reaction to such sites is to go to another thatE provides the same content.  In this particular case, however, we have C the designated site for VMS ISVs choosing to email in a VMS-hostile  environment.   ------------------------------  # Date: Wed, 02 Jul 2003 13:22:46 GMT # From: "John Smith" <a@nonymous.com> ? Subject: Re: Compaq Solutions Aliance - Do it still existing  ? G Message-ID: <G0BMa.27625$a51.2310@news02.bloor.is.net.cable.rogers.com>   3 "John Reagan" <john.reagan@hp.com> wrote in message ) news:f4jMa.3693$RV.62@news.cpqcorp.net...  > John Smith wrote:  > * > > Mozilla on VMS is  not current enough. > B > Lets see Mozilla 1.4 was announced on June 30th and right now on JulyC > 1st, I'm downloading the PCSI kit for OpenVMS.  Or are you saying  that8 > Mozilla.org isn't focusing on the correct feature set?    C pardon me...I've been involved with a wedding for the past week and E have only just made it back to the mounds of correspondence and  work  on my desk.   F My point was that with more and more web site developers *choosing* toF support IE-specific features which are not necessarily W3C compatible,E developers of other web browsers have to make a concious choice as to B whether they will be 'standards based' thereby potentially runningD into some of the same problems non-IE users hit when they browse theE HP web site (as an example), or to follow MS and attempt to implement A IE 'extensions' which may or may not peacefully co-exist with W3C 
 standards.   ------------------------------  + Date: Wed, 02 Jul 2003 16:12:55 +0200 (MET) 9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> ? Subject: Re: Compaq Solutions Aliance - Do it still existing  ? ; Message-ID: <01KXSEF1OR2CAM7Y4A@sysdev.deutsche-boerse.com>   H > My point was that with more and more web site developers *choosing* toH > support IE-specific features which are not necessarily W3C compatible,G > developers of other web browsers have to make a concious choice as to D > whether they will be 'standards based' thereby potentially runningF > into some of the same problems non-IE users hit when they browse theG > HP web site (as an example), or to follow MS and attempt to implement C > IE 'extensions' which may or may not peacefully co-exist with W3C  > standards.  G There is only one valid option here: stick to standards and complain to C and about those who don't.  Microsoft is well known for introducing H their own "extensions", often for functionality which is ALREADY COVEREDE by some standard.  The reason is clear: to lock users, developers etc G into their products.  Superior quality they don't have, so they have to G resort to tactics like this.  The more people who accept this, the more B Microsoft comes to dominate the world, not because of high qualityH (which they obviously don't have) but because of such dishonest tactics.  H Of course, a lot of this stuff is not documented.  Microsoft's point is I not to have people use its "standards", but have people buy its software  I (and perhaps continually upgrade their hardware to cope with the bloated  I software).  Thus, these "standards" will always remain poorly documented   and a "moving target".  See   -    http://www.hevanet.com/peace/microsoft.htm   E Personally, I don't care about the legal angle.  If they produce bad  H stuff, even bad stuff which doesn't allow people to replace parts of it G with better parts, they should be allowed to do so.  "Just say no!"  I  H think it is sad that some judges have ruled that Microsoft has to allow F for more modularity to allow competing products to work with Windows, F since the justification is that Microsoft has a practical monopoly in H the PC market.  One should combat the cause, not they symptom.  But, as B long as Microsoft isn't doing something patently illegal (such as G willful misrepresentation etc), the way to fight the cause is to stick  / to standards.  That is what they are there for.   D The recent publicity of open-source software has introduced a false D dichotomy: proprietary is bad, open-source is good.  In the case of E Microsoft, of course, the quality is objectively low.  However, some  > folks have gone so far as to say that proprietary software is  intrinsically bad.  H I think this comes mainly from the false assumption that Microsoft is a G typical example of proprietary software.  Obviously, I have no problem    using GOOD proprietary software.  C They key point is that the software itself should be able to be as  G proprietary and closed as it wants to be.  However, it should strictly  G adhere to standards when file formats, data-exchange protocols etc are  G involved, i.e. when interaction with other systems is possible.  It is  G the latter, not the former, area (in addition to bad quality) which is   Microsoft's problem.   ------------------------------  + Date: Wed, 02 Jul 2003 16:15:48 +0200 (MET) 9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> ? Subject: Re: Compaq Solutions Aliance - Do it still existing  ? ; Message-ID: <01KXSFBDSHM2AM7Y4A@sysdev.deutsche-boerse.com>   8 > >> I gather you have not actually tried this yourself. > > H > > This was a general remark about my reaction to unreadable websites, K > > usually those which are unreadable for no reason other than "trying to  L > > be a k001 webprogrammer"---for example, using JavaScript for links when * > > HTML HREF links would do the same job. > H > Certainly.  My general reaction to such sites is to go to another thatG > provides the same content.  In this particular case, however, we have E > the designated site for VMS ISVs choosing to email in a VMS-hostile  > environment.  G This (HP's decision) is just plain stupid.  I think people are missing  F the point when they say that we don't like WORD attachments in emails G etc.  I don't send people files which only make sense on a VMS machine  H unless I know that they have a VMS machine.  The equivalent would be HP G corresponding with people who develop PC applications and insisting on  H doing the correspondence with DEC Document files or whatever.  It's not E an issue of what one likes or not, it's an issue of stupidity or the  
 lack thereof.    ------------------------------  # Date: Wed, 02 Jul 2003 15:18:54 GMT # From: "John Smith" <a@nonymous.com> ? Subject: Re: Compaq Solutions Aliance - Do it still existing  ? H Message-ID: <yJCMa.43198$x4o.36711@news04.bloor.is.net.cable.rogers.com>  F "Phillip Helbig" <HELBPHI@sysdev.deutsche-boerse.com> wrote in message5 news:01KXSEF1OR2CAM7Y4A@sysdev.deutsche-boerse.com... < > > My point was that with more and more web site developers
 *choosing* to > > > support IE-specific features which are not necessarily W3C compatible, F > > developers of other web browsers have to make a concious choice as toF > > whether they will be 'standards based' thereby potentially runningD > > into some of the same problems non-IE users hit when they browse the ? > > HP web site (as an example), or to follow MS and attempt to 	 implement E > > IE 'extensions' which may or may not peacefully co-exist with W3C  > > standards. > F > There is only one valid option here: stick to standards and complain toE > and about those who don't.  Microsoft is well known for introducing B > their own "extensions", often for functionality which is ALREADY COVERED C > by some standard.  The reason is clear: to lock users, developers  etc F > into their products.  Superior quality they don't have, so they have toD > resort to tactics like this.  The more people who accept this, the moreD > Microsoft comes to dominate the world, not because of high qualityA > (which they obviously don't have) but because of such dishonest  tactics. > F > Of course, a lot of this stuff is not documented.  Microsoft's point isA > not to have people use its "standards", but have people buy its  softwareB > (and perhaps continually upgrade their hardware to cope with the bloated ? > software).  Thus, these "standards" will always remain poorly 
 documented > and a "moving target".  @ Included in this problem is that with the explosion of so-calledD 'programmers' coming on the scene over the past decade, and the easyD and cheap availability of software from Microsoft (via the net or at@ the corner store), it is easy to understand see how ignorance is bliss.  C The Microsoft manuals are easy to read (some may not think so), and B there is a cadre of support people at Microsoft who are willing toD tell you how to use their product in the Micrsoft-approved manner at $39.95 (sort of) per call.  B For many companies it's the 'no decision decision' - get MicrosoftC products and support as opposed to trying to decipher W3C standards B and implement them with only newsgroups as support mechanisms (notF true, but often percieved that way). Besides, "everyone uses MicrosoftC except for those Mac types....and unix/linux/everything else is for B propellor heads and we're focused on getting things done quick and/ dirty as opposed to building empires....right?"          > See  > / >    http://www.hevanet.com/peace/microsoft.htm  > F > Personally, I don't care about the legal angle.  If they produce badF > stuff, even bad stuff which doesn't allow people to replace parts of itE > with better parts, they should be allowed to do so.  "Just say no!"  I C > think it is sad that some judges have ruled that Microsoft has to  allow > > for more modularity to allow competing products to work with Windows,D > since the justification is that Microsoft has a practical monopoly inF > the PC market.  One should combat the cause, not they symptom.  But, asC > long as Microsoft isn't doing something patently illegal (such as B > willful misrepresentation etc), the way to fight the cause is to stick 1 > to standards.  That is what they are there for.   = The old adage is that "War is too important to be left to the E generals" and what the legal system is not doing with Microsoft makes < us feel somewhat like Churchill at the results of the MunichC conference between Chamberlain and Hitler - sickened and disgusted, + knowing what the inevitable result will be.     E > The recent publicity of open-source software has introduced a false E > dichotomy: proprietary is bad, open-source is good.  In the case of F > Microsoft, of course, the quality is objectively low.  However, some? > folks have gone so far as to say that proprietary software is  > intrinsically bad. > D > I think this comes mainly from the false assumption that Microsoft is a@ > typical example of proprietary software.  Obviously, I have no problem " > using GOOD proprietary software. > D > They key point is that the software itself should be able to be as? > proprietary and closed as it wants to be.  However, it should  strictlyD > adhere to standards when file formats, data-exchange protocols etc are E > involved, i.e. when interaction with other systems is possible.  It  isE > the latter, not the former, area (in addition to bad quality) which  is > Microsoft's problem.    C But VMS also gets painted with the same brush if only because it is E cast as 'that evil proprietary stuff'. Of course the cure for that is F advertising VMS. And while VMS doesn't have the wealth of applicationsB Windows has, it has one thing that in this day and age ought to beE winning selling points - security and reliability. But people have to  be told.   ------------------------------  # Date: Wed, 02 Jul 2003 07:09:49 GMT 6 From: "Craig A. Berry" <craigberry@mac.com.spamfooler> Subject: Re: cxx performance@ Message-ID: <3dca076f1205e24c35c900231124532c@free.teranews.com>  = In article <477e0934.0307012141.2d8865ba@posting.google.com>, 3  usenet_vms@lehrerfamily.com (Joshua Lehrer) wrote:   D > "Craig A. Berry" <craigberry@mac.com.spamfooler> wrote in message > > news:<1e99b2b7f574092d1ade34986ec50dc0@free.teranews.com>...A > > In article <9d337b47.0306302201.35036222@posting.google.com>, # > >  ohm62@hotmail.com (OHM) wrote:    > > G > > Forgive me if I've missed it, but I haven't noticed anyone talking  K > > about the demangler database in this thread.  That's one obvious place  J > > where a C++ compiler on VMS has to do a lot more work than a compiler J > > on other platforms, and very disk-intensive work at that.  I don't do  > H > Why does the VMS based compiler need to do more work with regards to a? > demangler database as compared to a compiler based on another  > platform?   F Because on VMS, linker symbols have a maximum length of 31 characters C (and for a long time had to be all upper case, though I'm not sure  C that's still true).  Mixed-case, potentially long symbol names are  H emulated by "mangling" the real name.  This default behavior of the C++ H compiler is (as far as I can tell) identical to and compatible with the H /NAMES=(AS_IS,SHORTENED) option on the C compiler.  I think the mangled D name is the first 24 characters of the real name plus a 7-character # checksum of the name.  For example:   : $ cxxdemangle/mangle "__xmlDoValidityCheckingDefaultValue" __XMLDOVALIDITYCHECKING3NQL295$   H The compiler has to maintain this database mapping the mangled names to B the real names.  By default that map is in an indexed RMS file in B [.CXX_REPOSITORY]CXX$DEMANGLER_DB.  Obviously the compiler has to @ access this file many, many times in the process of compiling a H program.  As I've already suggested, you can control where that file is B located, and it seems like a RAM disk might be worth trying.  For E optimum performance, though, the demangler table should perhaps be a  H hash in memory, but that could be quite a pain since the compiler still B needs to work correctly on older systems with very limited memory.  C I should reemphasize what I said before, which is that I've always  G *assumed* the name repository was the cause of slow C++ compilations.   G Anyone actually affected by this should do their own analysis and find  G out how busy that file is and whether it's really a major contributing   factor to resource consumption.    ------------------------------   Date: 2 Jul 2003 00:22:29 -0700 2 From: johan.nilsson@esrange.ssc.se (Johan Nilsson) Subject: Re: cxx performance= Message-ID: <c3c6388a.0307012322.65ab7a2f@posting.google.com>   n ejohnson@factset.com (Eric Johnson) wrote in message news:<ef79676b.0307010756.7699e18e@posting.google.com>...d > ohm62@hotmail.com (OHM) wrote in message news:<9d337b47.0306302201.35036222@posting.google.com>... > E > > My PC based collegues (using uniprocessor PentiumIV's 1Ghz 512Mb) E > > compile their 10 man year project (also using templates, although J > > maybe in slighly lesser extent) in about a half the time I need to getJ > > my 1 man year project rebuilt.  This is particularly offensive, when I6 > > think about the relative costs of these systems... > E > I'm glad to hear someone else say this too.  So I'll beat it around  > some more. > ? > VMS systems are expensive in total cost.  We all know that.   @ > Yes, I've heard all about the reduced total cost of ownership B > and how wonderful the total return of investment is.  That looks > great on paper.  > F > But when push comes to shove, and you've got a PC with a semi-modernG > C++ compiler than can outstrip an alpha system that costs two orders  D > of magnitude more, its mildly frustrating.  Heck, it makes me want > to compile on the PC.   < True. And that's what I normally do. Our current project wasF cross-platform from the beginning and we tried to stick with pure ANSIE C++/C, so most of the stuff can be initially developed on my PC. When B it comes to threading, timer support and other stuff requiring theC platform specific API it becomes harder (e.g. Win32 vs POSIX vs VMS C system services). Sigh. Imagine changing a small piece of code in a B single function and wait several minutes before the compilation isC complete and the unit tests can be run - it makes we want to scream A (especially when the same thing takes only some seconds on my 4-5  years old PC hardware).   C Why aren't precompiled headers and incremental linking supported? I C believe precompiled headers are even supported on the same compiler  (DEC CXX 6.5) on the Tru64.    > C > If you want VMS to continue to live, the developers tools have to I > be treated with first class love, and this includes overall performance ! > of that tool for the developer.    I'll second to that.   > F > Perhaps the problem is in the EDG front end, perhaps its in the GEMMF > code base.  I don't know.  I don't care.  I just want it an order ofF > magnitude faster.  Especially given the extra zero on the price tag.   Definitely. Agreed.   E A bit off-topic, but: is there a version of GNU C++ for OpenVMS Alpha 8 and, if so - anyone got experience from using it? If theD compilation/linking speeds were an order of magnitude faster I couldC go for using that for mainline development and then use DEC C++ for  the release compilations.    // Johan   ------------------------------  % Date: Wed, 02 Jul 2003 09:26:48 +0100 * From: Nic Clews <sendspamhere@[127.0.0.1]> Subject: Re: cxx performance' Message-ID: <bdu55m$2qk$1@lore.csc.com>    "David J. Dachtera" wrote: >  > Joshua Lehrer wrote: > > ] > > Nic Clews <sendspamhere@[127.0.0.1]> wrote in message news:<bdp5lo$d20$1@lore.csc.com>...   N > > > lookups), have generous CHANNELCNT and corresponding FILLM (not sure howL > > > the compilers and linkers for C fare here YMMV). Have generous working > > > sets.  > > 
 > > we do. > I > How 'bout page file quota? (It's a user process parameter in AUTHORIZE: > > pgflquota. See also the system parameters pql_dpgflquota and > pql_mpgflquota.)  / Thank you for responding about the other items.   F While you're essentially correct about the pagefile quota, ideally I'dB like to avoid page faulting, and having to use the page file, so ID didn't really see it as a concern. I'd be encouraged to have a smallE MAXPROCESSCNT and BALSETCNT, and based on what AUTOGEN sets aside for @ the nonpageable portions of VMS, divide up the memory evenly and: generously between the processes that need to be resident.   --  ? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer Sciences  nclews at csc dot com    ------------------------------  % Date: Wed, 02 Jul 2003 10:24:35 +0100 * From: Nic Clews <sendspamhere@[127.0.0.1]> Subject: Re: cxx performance' Message-ID: <bdu8hu$375$1@lore.csc.com>    Eric Johnson wrote:  >   E > I ran monitor file_system_cache for a bit, and we average over 95%.   H OK that is potentially pretty good, getting into the detail, look at theH number of IO's and reckon up the number of hard IOs that come from cacheF misses. i.e. if you're attempt rate was say 500 per second, with a 95%F hit rate, 25 IOs per second are still 'hard' (go to disk). This may or5 may not be cause for concern in the scheme of things.     > Our ACP_* values are...  > SYSGEN>  SHOW ACP_*  >   C > ACP_WINDOW                      7          7         1         -1  > Pointers   D  @ If you're suffering from file fragmentation, increasing this (or1 per-disk with /WINDOW and mount time) could help.   K > > How 'bout page file quota? (It's a user process parameter in AUTHORIZE: @ > > pgflquota. See also the system parameters pql_dpgflquota and > > pql_mpgflquota.) > ? > This is pretty representative of what a developer would have.  > ; > Maxjobs:         0  Fillm:       225  Bytlm:        65000 ; > Maxacctjobs:     0  Shrfillm:      0  Pbytlm:           0 ; > Maxdetach:       0  BIOlm:       100  JTquota:     200000 ; > Prclm:          10  DIOlm:       100  WSdef:         2048 ; > Prio:            4  ASTlm:       100  WSquo:        16400 ; > Queprio:         0  TQElm:       100  WSextent:     60000 ; > CPU:        (none)  Enqlm:      5000  Pgflquo:    1000000  >   H My observations is that FILLM seems pretty low, check CHANNELCNT which IF would have thought on modern systems with adequate memory should be at* least 512, and a FILLM of 497 per process.  F WSDEF and WSQUO, do they have to be so different? For a process to useH its quota and to grow to it, the process needs to pagefault above PFRATHA every AWSTIME each QUANTUM to have a single WSINC which means the F swapper may have to consume operating system overhead, add to that theD real time which passes while the process gets itself "up to size". AD WSQUO = WSDEF means that the process does not have to go through theH working set growth phase, only if there is 'free memory' to grow towards	 WSEXTENT.   G e.g. Default WSINC of 150 and default PFRATH of 120, means your process C needs to exceed 120 faults/sec for 95 CPU seconds, probably about 2 F minutes to you and me before it gets to WSQUOTA. This may not be a lotG on a single build taking upwards of 30 minutes, but if you are invoking B the compiler multiple times, the process starts back down again at
 WSDEFAULT.  > Having said that, here's the other edge to the sword. Say yourH developers sit in TPU, this is one sure fire memory hog, and you'll findE your developers wandering off for coffee, scratching their heads, you F know the typical things that developers do while code is in an editor,E all using that 16400 pages of memory. Expensive! EDT's not as greedy. C The TPU police will probably have a shout at me now, but a fact's a  fact.   ; You may wish to consider BATCH building queues or accounts.   0 Also do consider the PQL_M parameters in SYSGEN.  ; > If you invoke the CXX v6.5 on a non-trivial C++ file with H > /list/show=stat, you'll see that a considerable amount of time is lost@ > in the "Final" section, particularly "peepholing", and "object > scheduling". > G > Someone may need to smack around the EDG folks, because there's a ton G > of time blown in the lex/parse/lower too which is probably beyond the2 > direct control of HP.n  H Outside my scope, but I know you're suggesting areas to look for a cure,A as the "business" end of the compiler is being revamped for EPIC,oD concentrate on fixing the symptoms, i.e. tackling system performance# during the unfavourable statistics.l   -- A? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer SciencesD nclews at csc dot com    ------------------------------  % Date: Wed, 02 Jul 2003 10:36:45 +0100s* From: Nic Clews <sendspamhere@[127.0.0.1]> Subject: Re: cxx performance' Message-ID: <bdu98r$395$1@lore.csc.com>i   "Craig A. Berry" wrote:  >    ... I > The compiler has to maintain this database mapping the mangled names tocC > the real names.  By default that map is in an indexed RMS file inIC > [.CXX_REPOSITORY]CXX$DEMANGLER_DB.  Obviously the compiler has tonA > access this file many, many times in the process of compiling atI > program.  As I've already suggested, you can control where that file istC > located, and it seems like a RAM disk might be worth trying.  ForWF > optimum performance, though, the demangler table should perhaps be aI > hash in memory, but that could be quite a pain since the compiler stilleD > needs to work correctly on older systems with very limited memory.  1 And a candidate for file tuning, ANAL/RMS/FDL andM> EDIT/FDL/NOINTER/SCRIPT=OPT or whatever to suit, followed by a CONVERT/FDL.   D > I should reemphasize what I said before, which is that I've alwaysG > *assumed* the name repository was the cause of slow C++ compilations. H > Anyone actually affected by this should do their own analysis and findH > out how busy that file is and whether it's really a major contributing! > factor to resource consumption.M   Good advice. --  ? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer Sciences  nclews at csc dot comM   ------------------------------  $ Date: Wed, 2 Jul 2003 12:57:27 +0100* From: "John Travell" <john@travell.uk.net> Subject: Re: cxx performance6 Message-ID: <bduhph$10v0uk$1@ID-120847.news.dfncis.de>  7 "Nic Clews" <sendspamhere@[127.0.0.1]> wrote in messagee! news:bdu55m$2qk$1@lore.csc.com...> > "David J. Dachtera" wrote: > >  > > Joshua Lehrer wrote: > > >E; > > > Nic Clews <sendspamhere@[127.0.0.1]> wrote in messageU# news:<bdp5lo$d20$1@lore.csc.com>...P >SL > > > > lookups), have generous CHANNELCNT and corresponding FILLM (not sure howRF > > > > the compilers and linkers for C fare here YMMV). Have generous workingd
 > > > > sets.o > > >  > > > we do. > >ZK > > How 'bout page file quota? (It's a user process parameter in AUTHORIZE:o@ > > pgflquota. See also the system parameters pql_dpgflquota and > > pql_mpgflquota.) >p1 > Thank you for responding about the other items.p >eH > While you're essentially correct about the pagefile quota, ideally I'dD > like to avoid page faulting, and having to use the page file, so IF > didn't really see it as a concern. I'd be encouraged to have a smallG > MAXPROCESSCNT and BALSETCNT, and based on what AUTOGEN sets aside foroB > the nonpageable portions of VMS, divide up the memory evenly and< > generously between the processes that need to be resident. >h > --  F If you are running a recent version of OpenVMSAlpha the concerns about3 MAXPROCESSCNT and BALSETCNT are no-longer relevant.aK On VAX and pre-V7 Alpha all of the process page tables were resident within-L the PDH, itself resident in S0 space. The problem was that a large number ofG processes, each with a large VIRTUALPAGECNT, could exceed the availableo5 virtual address space to map all of them in S0 space.fE Current Alpha versions keep the page tables in S2 space, along with ayK shedload of other things that used to be in S0 space. Since S2 space can be.L expanded to silly sizes if the need arises the address space issue no-longer exists. L The only issue around page tables these days is when you have a memory tightH system, setting BALSETCNT to the smallest useful value will minimise oneJ aspect of how much memory is permanently reserved for VMS. Since there areK many other things to adjust to maximise the physical memory available to annE application a LOT of research into system tuning is called for... :-).     -- John Travell  VMS crashdump expertise for hire john@jomatech.comm +44-(0)23-92552229 http://www.jomatech.com/       ---e& Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).A Version: 6.0.493 / Virus Database: 292 - Release Date: 25/06/2003    ------------------------------  $ Date: Wed, 2 Jul 2003 09:23:25 +0100* From: "Richard Brodie" <R.Brodie@rl.ac.uk># Subject: Re: cxx performance (IA64)K+ Message-ID: <bdu4pu$sq0@newton.cc.rl.ac.uk>a  A "Craig A. Berry" <craigberry@mac.com.spamfooler> wrote in messagel: news:3dca076f1205e24c35c900231124532c@free.teranews.com...  G > Because on VMS, linker symbols have a maximum length of 31 characterstD > (and for a long time had to be all upper case, though I'm not sure > that's still true).s  D They can be mixed on Alpha but the standard libraries have uppercaseE names. Does anyone know if the symbol length restrictions are relaxed = in IA64? I understand there will be new object/image formats.r   ------------------------------  # Date: Wed, 02 Jul 2003 17:23:50 GMTr& From: John Reagan <john.reagan@hp.com># Subject: Re: cxx performance (IA64)s2 Message-ID: <GyEMa.3791$Wb2.2343@news.cpqcorp.net>   Richard Brodie wrote:nC > "Craig A. Berry" <craigberry@mac.com.spamfooler> wrote in messages< > news:3dca076f1205e24c35c900231124532c@free.teranews.com... >  > G >>Because on VMS, linker symbols have a maximum length of 31 charactersiD >>(and for a long time had to be all upper case, though I'm not sure >>that's still true).  >  > F > They can be mixed on Alpha but the standard libraries have uppercaseG > names. Does anyone know if the symbol length restrictions are relaxedJ? > in IA64? I understand there will be new object/image formats.  >  >   C The linker/librarian have always been case-sensitive (yes, VAX and .F Alpha).  It is the compilers that traditionally upcased global names. B You can get several of the compilers to use mixed-cased names via C various language features or command line qualifiers.  You can use sI non-printable characters for all they care, but .MAPs, etc. might look a w
 little funny.a  / The VAX linker/librarian has a max length of 31i  E The Alpha linker/librarian has a max length of 63 although most (C++ tG might use all 63 chars worth??) compilers still enforce max lengths of  H 31 for global names.  The extra space was to allow compilers (Pascal in / particular) to create global names of the form dE 'module-name'.'symbol-name' to create module-local symbols.  The VAX oF linker/object language has a concept called module-local symbols that G didn't get directly ported to Alpha.  The extra name length allows you   to 'do it yourself'.  G The IA64 linker doesn't seem to have a length limit for names.  From a bA quick glance, I think the librarian is currently at 128 (I did a mA LIBR/LIST and looked at the 'key length').  However, none of the eD compilers (that I know of) have been modified to allow global names @ longer than 31.  C++ might be one of that will use longer names.   --   John Reagan ' Compaq Pascal/{A|I}MACRO Project Leaderg Hewlett-Packard Company    ------------------------------  # Date: Wed, 02 Jul 2003 09:19:30 GMT 4 From: "Johan Nilsson" <johan.nilsson@esrange.ssc.se> Subject: Enumerating logicalst@ Message-ID: <cf1e4c34b5e5257f11fbd437841229b8@free.teranews.com>   Hi,A  D is there a way to programmatically enumerate all logical definitionsJ contained in a specific logical name table, without resorting to parse the output of show log/table=xxx .   // Johan   ------------------------------  $ Date: Wed, 2 Jul 2003 12:33:43 +0200< From: "Martin Vorlaender" <martin.vorlaender@pdv-systeme.de>! Subject: Re: Enumerating logicalsc5 Message-ID: <bducf2$11c49j$1@ID-56200.news.dfncis.de>l   Johan Nilsson wrote:F > is there a way to programmatically enumerate all logical definitionsB > contained in a specific logical name table, without resorting to* > parse the output of show log/table=xxx .  = There's a package written by Ferry Bolhar. You can find it atp9 ftp://ftp.process.com/vms-freeware/fileserv/lnmlookup.zipa: It's a user-written system service, so requires privileges  (at least for the installation).   cu,O   Martin -- rF   OpenVMS:                | Martin Vorlaender  |  VMS & WNT programmer3    The operating system   | work: mv@pdv-systeme.detF    God runs the           |   http://www.pdv-systeme.de/users/martinv/:    earth simulation on.   | home: martin@radiogaga.harz.de   ------------------------------  # Date: Wed, 02 Jul 2003 16:11:47 GMT 3 From: wallacethinmintr@eircom.net (Russell Wallace)c1 Subject: Re: HP Powers More TOP500 Supercomputersr0 Message-ID: <3f03030c.756436326@news.eircom.net>  A On 1 Jul 2003 16:04:00 -0700, keithparris_NOSPAM@yahoo.com (Keith- Parris) wrote:  F >But Intel has a pretty good track record of backing the architectures) >which eventually win in the marketplace.u   Do they?  A >They backed the Ethernet standard, along with Digital and Xerox.S  C Lots of people backed Ethernet. In fact, Token Ring versus EthernetaB was pretty nearly IBM versus the world; if you're going to draw an8 analogy from that, seems to me Intel are in IBM's shoes.  F But for CPU architectures, Intel produced x86, iAPX432, i860 and i960.E One success and three failures. If we're going by history, that's noti good betting odds.  E >As with the Z80, some folks will use X86-64 and like it, but withoutsG >Intel's backing of the ISA extensions, it has a tough road ahead of ite+ >for acceptance in the general marketplace.a  F The Z80 was accepted just fine in the general marketplace. It was onlyC abandoned when the world moved on to the next "bitness generation".i@ x86-64 is 64 bits just like Itanium, so that doesn't apply here.   -- a "Sore wa himitsu desu."g To reply by email, remove9 the small snack from address. ! http://www.esatclear.ie/~rwallaces   ------------------------------   Date: 2 Jul 2003 05:45:25 -0700r) From: daniel@mimer.se (Daniel Gustafsson)eR Subject: Re: HP Webcast this morning on Next-Generation Intel Itanium 2 processors= Message-ID: <de4cfd03.0307020445.541ed943@posting.google.com>t  d "Bill Todd" <billtodd@metrocast.net> wrote in message news:<teicneuOq_8B35-iXTWJgA@metrocast.net>...5 > "Rick Jones" <foo@bar.baz.invalid> wrote in messagew. > news:cDkMa.3696$K41.2289@news.cpqcorp.net... > >eH > > I guess we aren't worrying too much about cache these days, or wouldD > > have mentioned that the 1.7 GHz POWER4+ has a 128 MB L3 cache :) > F > It would be interesting to see how it would do without that, and theN > soon-to-arrive PC970 (and the Apple versions of it, up to 2 GHz) that has noM > off-chip L3 and only 512 KB of on-chip L2 should provide at least a clue inSN > that regard (my impression is that they will be available in MP servers, but6 > who knows whether they'll appear in this benchmark).  A Going from 8MB L3 to 32MB L3 give a pretty significant speedup in A SPECfp2000 on POWER4+ 1450Mhz. 8MB L3 gives 984 and 32MB L3 givesb 1091..  O http://www.specbench.org:/cpu2000/results/res2003q2/cpu2000-20030407-02073.htmlsO http://www.specbench.org:/cpu2000/results/res2003q2/cpu2000-20030407-02065.htmlV  D This machine probably have other changes but POWER 1.5Ghz with 128MB
 L3 gets 1398.iO http://www.specbench.org:/cpu2000/results/res2003q2/cpu2000-20030505-02142.htmly   -- Daniel Gustafsson > http://www.mimer.se - DBMS with optimistic concurrency control   ------------------------------  # Date: Wed, 02 Jul 2003 13:53:57 GMT 9 From: "Fred Kleinsorge" <my-last-name@stardotzko.dec.com>eR Subject: Re: HP Webcast this morning on Next-Generation Intel Itanium 2 processors1 Message-ID: <VtBMa.3757$DS1.932@news.cpqcorp.net>W  @ You might as well give up with Bill.  He will continue to try toH pathetically reach for any crumb he can to prove his point.  Eventually,H when all else fails he'll fall back to something along the lines of "EV8. would have been better/faster/cheaper/cooler".  I BTW his definition of a "smidge" would be a "vast gulf" if Ooopstreon andu Madison were reversed.    3 "Rick Jones" <foo@bar.baz.invalid> wrote in message,, news:cDkMa.3696$K41.2289@news.cpqcorp.net...+ > Bill Todd <billtodd@metrocast.net> wrote:mH > > Of course another question is whether that added SPECint performance willK > > actually translate into anything worthwhile in real-world applications.h IrJ > > notice that your 4-processor SPECweb99_SSL result has not yet improvedJ > > (despite being run on an HP-UX base and hence presumably being able toG > > benefit from the same compiler that one might suspect generated thedH > > SPECint_base score you refer to above:  SGI only managed to get 1077 with< > > presumably the best compiler they had available to them) >oF > I wonder if there is a correlation between SPECint and SPECweb99_SSL	 > scores.u > H > > - and has now dropped to third place behind not only the 4-processorD > > 1.8 GHz Opteron but now the 4-processor 1.7 GHz POWER4+ as well.C > > And the dual-Madison results remain only a smidge ahead of both  > > Opteron and P4.s >R4 > You didn't look deeply enough at the HP website :) >l* > http://www.hp.com/go/hpintegritypresskit >./ > which has a link to a performance writeup at:t >  >oL http://www.hp.com/products1/servers/press/kits/integrity/Performance_Fact_Sh eet.pdfo >m= > which gives a bunch of benchmark results, among them a 3702uG > SPECweb99_SSL for the four-CPU 1.5 GHz rx5670, and 1930 SPECweb99_SSLt > for the 2 CPU 1.5 GHz rx2600.  >cF > I guess we aren't worrying too much about cache these days, or wouldF > have mentioned that the 1.7 GHz POWER4+ has a 128 MB L3 cache :) andA > you forgot to mention the 1.7GHz POWER4+ results from IBM use al% > dedicated SSL hardware accelerator.  >mG > And speaking of P4 and x86... If you go to IBM's xSeries benchmarking G > pages, you will see where they are touting two new 4 CPU Xeon resultstF > at 2.8 GHz for the x360 and the x255.  They come-in at 2174 and 2110 > SPECweb99_SSL respectively.b >l >eL ftp://ftp.pc.ibm.com/pub/special/serverperformance/pb_x360_specweb99_ssl_jun e03.pdfo >t >sL ftp://ftp.pc.ibm.com/pub/special/serverperformance/pb_x255_specweb99_ssl_jun e03.pdfc >hD > Those went online at IBM yesterday (June 30th 2003).  They haven'tE > gone up on http://www.spec.org yet but IBM mentions SPEC review andgI > such in those PDF files so ostensibly, baring something untoward duringrH > review those should appear there at some point to take their place forF > comparison with all the other results - the latest rx5670 and rx2600* > results likely among them at that point. >: > rick jones >a? > SPECweb is a trademark of SPEC, results as of July 1, 2003 onh. > www.hp.com, www.spec.org and ibm.com etc etc >k > --J > Wisdom Teeth are impacted, people are affected by the effects of events.H > these opinions are mine, all mine; HP might not want them anyway... :)C > feel free to post, OR email to raj in cup.hp.com  but NOT BOTH...m   ------------------------------  # Date: Wed, 02 Jul 2003 13:56:34 GMTr9 From: "Fred Kleinsorge" <my-last-name@stardotzko.dec.com>aR Subject: Re: HP Webcast this morning on Next-Generation Intel Itanium 2 processors1 Message-ID: <mwBMa.3758$cS1.162@news.cpqcorp.net>.  5 "Bill Todd" <billtodd@metrocast.net> wrote in messages, news:0BudncjFdqTCk5-iXTWJkA@metrocast.net... > J > Nope.  I suspect that if/when the Alpha team finally rescues Itanic from its-K > current core in 3 - 4 more years I'll have nothing left to disparage hereH; > except cHumPaq's ethics (and the occasional dimwit post).  >a  D You would reduce the occasional part by at least half if you stopped posting.   ------------------------------   Date: 2 Jul 2003 09:55:12 -0500.- From: Kilgallen@SpamCop.net (Larry Kilgallen)rR Subject: Re: HP Webcast this morning on Next-Generation Intel Itanium 2 processors3 Message-ID: <qecJwZyM9uaV@eisner.encompasserve.org>e  m In article <mwBMa.3758$cS1.162@news.cpqcorp.net>, "Fred Kleinsorge" <my-last-name@stardotzko.dec.com> writes:  > 7 > "Bill Todd" <billtodd@metrocast.net> wrote in message . > news:0BudncjFdqTCk5-iXTWJkA@metrocast.net... >>K >> Nope.  I suspect that if/when the Alpha team finally rescues Itanic from  > itstL >> current core in 3 - 4 more years I'll have nothing left to disparage here< >> except cHumPaq's ethics (and the occasional dimwit post). >> > F > You would reduce the occasional part by at least half if you stopped
 > posting.  H And Fred, you would reduce the other half if you stopped quoting him :-)   ------------------------------  # Date: Wed, 02 Jul 2003 16:33:57 GMTt# From: "John Smith" <a@nonymous.com>tE Subject: More than VMS is required, but VMS is a great starting pointsH Message-ID: <VPDMa.43846$x4o.27799@news04.bloor.is.net.cable.rogers.com>  $ California's New Rules Of Disclosure  C State law will force companies nationwide to make security breaches: public   By George V. Hulme InformationWeekg
 June 23, 2003D URL:F http://www.informationweek.com/story/showArticle.jhtml?articleID=10700 733<      E A California law that takes effect July 1 will force companies inside@E and outside the state to do what they historically have been loath toe8 do: disclose embarrassing information-security breaches.  @ If companies believe that their California customers' personally? identifiable financial information may have been accessed by an C unauthorized party, they must inform those customers of the breach.g? Disclosure may be delayed if law-enforcement officials deem ther- disclosure could jeopardize an investigation.   : Of 376 organizations polled for the 2003 Computer SecurityE Institute/FBI Computer Crime and Security Survey, each experiencing a @ breach in the past year, half say they kept the incidents quiet.; Thirty percent of those surveyed reported the breach to laww: enforcement, while 21% sought legal counsel. A majority ofF organizations say negative publicity was the reason for not disclosing% security breaches to law enforcement.   A Now organizations that do business with Californians won't have aaA choice--and the same may soon be true for companies that don't. At@ federal law similar to California's is in the works: Sen. DianneD Feinstein, D-Calif., is readying a federal bill based largely on the California law.   A Security experts say most companies, especially those outside thecD already heavily regulated health-care and financial industries, haveD done little to prepare for the new law. "Companies underestimate theF impact of the law," says Ryan McGee, director of product marketing forD Internet security company Network Associates Inc.'s McAfee division.F He says his company has received few inquires from customers about howA to comply. "It will take lawsuits and serious damages before many 2 businesses become concerned about it," McGee says.  F The law, which passed last year, is only now grabbing the attention ofC companies outside California. "We found out that it applies to us at9 few weeks ago," says a security specialist at a Northeasts@ consumer-goods manufacturer. "We're looking at how we can betterE encrypt customer information and possibly segment customer names froma@ their financial information so a hacker would have to breach two databases."p   ......etc.....   ------------------------------   Date: 2 Jul 2003 12:14:03 -0500s; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)B" Subject: Re: Mount a backup /image3 Message-ID: <VrMXJNFCeO8+@eisner.encompasserve.org>   Q In article <3f030c3a$0$29623$626a54ce@news.free.fr>, "vn" <zool@zool.com> writes:t > Hi,p > M > Is it possible to mount a saveset on a tape, saved with backup /image, likeo
 > a device ? f  0    No, unless you intend to write a lot of code.  ? > Or is there any solution to recover a single file in a backup'( > /image without restore an entire disk?  ?    Yes.  You can recover any single file from any backup except G    /physical by using /select.  A save set created /image is fine with  (    restore using /select and not /image.  D    The only situation you need to worry about is the VAX Standalone F    BACKUP.  That software always does /image.  But it would be absurd D    to boot Standalone BACKUP to do a single file restore even if it G    could do it.  There's no problem using the online VMS BACKUP commandoB    to /select even if the /image was made using Standalone BACKUP.   ------------------------------  $ Date: Wed, 2 Jul 2003 18:45:53 +0200 From: "vn" <zool@zool.com> Subject: Mount a backup /image4 Message-ID: <3f030c3a$0$29623$626a54ce@news.free.fr>   Hi,h  K Is it possible to mount a saveset on a tape, saved with backup /image, like H a device ? Or is there any solution to recover a single file in a backup& /image without restore an entire disk?   Thanks   Victor   ------------------------------  # Date: Wed, 02 Jul 2003 15:20:14 GMT 3 From: wallacethinmintr@eircom.net (Russell Wallace)o' Subject: Re: New OpenVMS-to-Itanium FAQa0 Message-ID: <3f02f80a.753618195@news.eircom.net>  @ On Tue, 1 Jul 2003 07:19:24 -0700, "Tom Linden" <tom@kednos.com> wrote:   >I doubt anyG >clear thinking person would seriously consider translating PL/I to C++hH >as even an option.  So you may wish to remove that as well as an optionI >since it questions your credibility.  Reengineering yes, Translating no.s  F I'm curious, what features does PL/1 have that would require a program< written in it to be completely redesigned for a port to C++?   -- . "Sore wa himitsu desu."o To reply by email, removeI the small snack from address.o! http://www.esatclear.ie/~rwallacei   ------------------------------  $ Date: Wed, 2 Jul 2003 08:33:36 -0700# From: "Tom Linden" <tom@kednos.com> ' Subject: RE: New OpenVMS-to-Itanium FAQO9 Message-ID: <CIEJLCMNHNNDLLOOGNJICEEGHIAA.tom@kednos.com>V   >-----Original Message-----p; >From: Russell Wallace [mailto:wallacethinmintr@eircom.net]d' >Sent: Wednesday, July 02, 2003 8:20 AM- >To: Info-VAX@Mvb.Saic.Com( >Subject: Re: New OpenVMS-to-Itanium FAQ >- >-A >On Tue, 1 Jul 2003 07:19:24 -0700, "Tom Linden" <tom@kednos.com>A >wrote:< >i
 >>I doubt anysH >>clear thinking person would seriously consider translating PL/I to C++I >>as even an option.  So you may wish to remove that as well as an optionaJ >>since it questions your credibility.  Reengineering yes, Translating no. > G >I'm curious, what features does PL/1 have that would require a programe= >written in it to be completely redesigned for a port to C++? 9 How about ON conditions, and lexical scoping for starters> >  >--  >"Sore wa himitsu desu." >To reply by email, remove >the small snack from address." >http://www.esatclear.ie/~rwallace >h >---' >Incoming mail is certified Virus Free.e; >Checked by AVG anti-virus system (http://www.grisoft.com). A >Version: 6.0.489 / Virus Database: 288 - Release Date: 6/10/2003s >n ---v& Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).@ Version: 6.0.489 / Virus Database: 288 - Release Date: 6/10/2003   ------------------------------  % Date: Wed, 02 Jul 2003 11:22:28 -0400h& From: Jilly <jilly@clarityconnect.com>7 Subject: Re: Open VMS Alpha upgrade from 7.2-1 to 7.2-2s2 Message-ID: <3F02F8B4.C7005828@clarityconnect.com>  C Please refer to ftp://ftp.compaq.com/pub/services/software/ovms.pdfo  D V7.2-2 ended Standard Support on Dec. 31, 2002 and is only supportedB under Prior Version Support contracts at this time.  Prior Version/ Support is an additional cost support contract.r   Beach, Runner@nospam.com wrote:e >  > Keith Parris wrote:F > L > > Rob.Buxton@wcc.govt.nz wrote in message news:<3eff8459.13509046@news>...F > > > If you're running a fully patched 7.2-1 then there's very littleJ > > > difference between taht and 7.2-2. My understanding is that 7.2-2 is< > > > just a roll together of patches into a stable release. > >l@ > > Actually, 7.2-2 also included some of the features from 7.3. > L > There is a rename patch that should be applied before the upgrade. Some of > thei% > ECOs put patches i the wrong place.M > F > 7.2-1 is also prior version support, while 7.2-2 still is supported. >  > Beach Runner   -- aC Jilly	- Working from Home in the Chemung River Valley - Waverly, NY 0 	- jilly@clarityconnect.com			- Brett Bodine fan+ 	- Mark.Jilson@hp.com				- since 1975 or soi 	- http://www.jilly.baka.com   ------------------------------  $ Date: Wed, 2 Jul 2003 09:16:38 -0400 From: norm.raphael@metso.comN Subject: Re: OpenVMS Host-Based MiniMerge Statement and Schedule for Customers? Message-ID: <OFC4E3AB62.BAA9C5B1-ON85256D57.0048DEA7@metso.com>r  0 And your point is...? [Now that wasn't so hard.]  + Host-Based Minimerge Schedule as of 6/30/03r  B Task Name                               Start           CompletionB ---------                               -----           ---------- HBMM Features (code and debug)B   Minimerge from a bitmap (essentials)  06/20/2003      08/15/2003B   Bitmap Reset                          06/19/2003      09/25/2003B   Merge Priority                        06/24/2003      09/25/2003B   Master Bitmap Policy                  06/19/2003      10/01/2003B HBMM code and debug phase complete      06/17/2003      10/20/2003B Performance characterization            06/17/2003      10/09/2003B User documentation                      07/14/2003      11/20/2003B Internal testing and Qualification      10/10/2003      12/29/2003B External Field Test                     12/31/2003      03/24/20042 Production Kit available                03/25/2004    I From:  susan_skonetski@hotmail.com (Sue Skonetski) on 07/01/2003 02:45 PMe  = Please respond to susan_skonetski@hotmail.com (Sue Skonetski)o   To:    Info-VAX@Mvb.Saic.Com cc:   C Subject:    OpenVMS Host-Based MiniMerge Statement and Schedule forh        Customers    F There was a word schedule attached but I know how you folks feel about that.u   sue,  4 ----------------------------------------------------  + Host-Based MiniMerge Schedule Now Availableo  C The OpenVMS Engineering and Business Management group would like tonF announce the schedule for Host-Based MiniMerge capability.  As most ofB you know, the teams have been spending virtually all of their timeD over the past 3 months doing design, "proof of concept" and scheduleF development for this important feature.  We are now ready to make this4 information available to the OpenVMS user community.  F Overview - The purpose of the Host-Based MiniMerge feature is to allow= a "minimum" MERGE of only shadow set changes (rather than theeD currently-required "full" MERGE) when a member of an OpenVMS clusterB is removed and re-introduced into a cluster.  A similar capability< exists today for CI-based storage, and this new "host-based"E implementation will allow a MiniMerge to occur on any type of storageoC that is utilized in a Host-Based Volume Shadowing shadow set.  ThisiA implementation utilizes the "write bitmap" technology in use withoA MiniCopy today, in order to keep track of shadow set data changes C while the cluster member is out of the cluster.  When the member is.C brought back in, shadow sets will be able to MERGE only the changeddA areas of the disk, thus resulting in a much faster restoration ofi shadow sets into the cluster.o   ------------------------------  + Date: Wed, 02 Jul 2003 10:35:15 +0200 (MET)o9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com>e0 Subject: Re: Porting C Code VAX -> Alpha -> IA64; Message-ID: <01KXS3F6SXTUAPKEWT@sysdev.deutsche-boerse.com>o  > >     CC/STANDARD=VAXC/EXTERN_MODEL=COMMON_BLOCK/SHARE_GLOBALS  ) Standard=VAXC is almost an oxymoron.  :-)   D Even if this option will be supported on Itanium, I would recommend 3 fixing everything which breaks if you don't use it.   H To be fair, the VAX compiler predates the '89 C standard.  Somewhere on F the net you can find an article by Dennis Ritchie (probably there's a H link from his home page) in which he praises the quality of the (first)  VAX C compiler.    ------------------------------  # Date: Wed, 02 Jul 2003 14:17:26 GMTi- From: "Ed Vogel" <edward.vogel_nospam@hp.com>b0 Subject: Re: Porting C Code VAX -> Alpha -> IA642 Message-ID: <WPBMa.3760$HX1.3023@news.cpqcorp.net>  , "Code Monkey" <me@here.com> wrote in message* news:bdt2fb$nc7$1@sparta.btinternet.com...K | | Will the Alpha Compaq C compiler generate code as efficiently as it cany withG | this symbol ? Or to unlock it's true potential should it be changed ?t |a  E     No.  There are some optimizations that the compiler would do when  /STAND=RELAXEDC     is specified that it does not do when /STAND=VAXC is specified.   F     However, sometimes turning on these optimizations breaks code that relied on a VAX CnL     behavior (this is why the optimization is turned off when /STAND=VAXC is specified).   I | | Also does anybody know if the OpenVMS C Compiler on Itanium will have  the   | /STANDARD=VAXC option in it ??  F     The goal for the C compiler on Itanium is "recompile and go".  The /STAND=VAXC qualifiern     will be supported there.       Ed Vogel      HP C for OpenVMS Engineering   ------------------------------   Date: 2 Jul 2003 14:10:19 GMTi/ From: Thomas Dickey <dickey@saltmine.radix.net>n0 Subject: Re: Porting C Code VAX -> Alpha -> IA64* Message-ID: <bdup4b$n3b$1@news1.radix.net>  : Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> wrote:? >>     CC/STANDARD=VAXC/EXTERN_MODEL=COMMON_BLOCK/SHARE_GLOBALSc  + > Standard=VAXC is almost an oxymoron.  :-)   F > Even if this option will be supported on Itanium, I would recommend 5 > fixing everything which breaks if you don't use it.   J > To be fair, the VAX compiler predates the '89 C standard.  Somewhere on H > the net you can find an article by Dennis Ritchie (probably there's a J > link from his home page) in which he praises the quality of the (first)  > VAX C compiler.a  = otoh, it did not care if one added spurious &'s before arraysfA (that was fixed in a later version, but caused me some confusion)    --  = Thomas E. Dickey <dickey@radix.net> <dickey@herndon4.his.com>b http://dickey.his.coma ftp://dickey.his.com   ------------------------------  $ Date: Wed, 2 Jul 2003 18:33:52 +1000: From: "David McKenzie" <david.mckenzie@paradigm-shift.biz> Subject: Re: Rethinking  V.M.SC Message-ID: <3f029987$0$59950$c30e37c6@lon-reader.news.telstra.net>    Hmmm,n  C going past the semantics, and the words of the question, there is aLH different point of view. Why not just look at the concept of addressableJ storage. VM implies that there is a process that swaps or pages, one couldK take the view that it is all storage and performance enhancemant would deal L with the fact that some is quicker than others. This may well be implementedF as VM is now, but ot doesn't have to be. After all look at the logicalG absurdity of placing a page file on a ram disk. Because it looks like aeG disk, it needs page file, because it is fast, it acts like memory. I do H think we are stuck in a paradigm where mechanism, and implementation areI seen as wedded together. Surely the reality is that address space is just 8 that, and where and how it is stored is a sub construct.    ; "Fabio Cardoso" <fabiopenvms@yahoo.com.br> wrote in messaged: news:20030629211410.38560.qmail@web20204.mail.yahoo.com... > People >i >a4 > The VMS was developed 25 years ago when memory was) > expensive and Virtual Memory an option.p3 > But do we need a Virtual Memory System nowadays ?i3 > May be we should think in another architecture asr9 > the memories are becoming cheaper. May be the VM Systems: > can be improved or rethinked... What are the performance< > issues related in having the VM (paging, swapping, related5 > to disk I/O delay, etc ...) What do u think about ?n >i	 > Regardss >  > FC >0 >g >o > =====  > ========================== > Fbio dos Santos Cardoso > OpenVMS System Manager > Rio de Janeiro - Brazil  > fabiopenvms@yahoo.com.br > ========================== > $ > __________________________________ > Do you Yahoo!?- > SBC Yahoo! DSL - Now only $29.95 per month!+ > http://sbc.yahoo.com   ------------------------------  + Date: Wed, 02 Jul 2003 11:07:07 +0200 (MET)s9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com>g Subject: Re: Rethinking  V.M.S; Message-ID: <01KXS4JWMQZGAM7Y4A@sysdev.deutsche-boerse.com>l  I > After all look at the logical absurdity of placing a page file on a ramlG > disk. Because it looks like a disk, it needs page file, because it isfH > fast, it acts like memory. I do think we are stuck in a paradigm whereG > mechanism, and implementation are seen as wedded together. Surely theeE > reality is that address space is just that, and where and how it isa > stored is a sub construct. S  E Interesting point.  However, it is nice to know what is a "real disk"eA (or, at least, which shadow sets have at least one real disk as a1E member!) so that it is clear what data will remain when the power is e
 switched off.C  B It is also nice to know, for performance reasons, what is on what H physical disk, which is one reason why I dislike the unix idea of mount G points etc.  One can always hide this information via logical names if .
 necessary.   ------------------------------   Date: 2 Jul 2003 08:12:38 -0500e; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)u Subject: Re: Rethinking  V.M.S3 Message-ID: <5YxQygT113Yb@eisner.encompasserve.org>r  <  "Fabio Cardoso" <fabiopenvms@yahoo.com.br> wrote in message;  news:20030629211410.38560.qmail@web20204.mail.yahoo.com...  > People >s >e4 > The VMS was developed 25 years ago when memory was) > expensive and Virtual Memory an option. 3 > But do we need a Virtual Memory System nowadays ?F3 > May be we should think in another architecture asy9 > the memories are becoming cheaper. May be the VM Systemf: > can be improved or rethinked... What are the performance< > issues related in having the VM (paging, swapping, related5 > to disk I/O delay, etc ...) What do u think about ?d  D   Even embedded realtime systems are moving in the direction of trueC   virtual memory support because of its many adavantages other than,D   saving RAM.  This is true even though the file system is often RAMB   based (you're not saving RAM if you page to RAM) and the kernelsC   are often bigger and therefor require more RAM than those kernels:   which don't do VM.   ------------------------------   Date: 2 Jul 03 08:10:35 +0200s) From: p_sture@elias.decus.ch (Paul Sture)g Subject: Re: Running VMS off CDm) Message-ID: <Kypc5703mHcZ@elias.decus.ch>   m In article <20030701094611.22461.qmail@nym.alias.net>, Doc.Cypher <Use-Author-Address-Header@[127.1]> writes:lI > On Mon, 30 Jun 2003, "David J. Dachtera" <djesys.nospam@fsi.net> wrote:p >>"Doc.Cypher" wrote:  >>> P >>> For those that have noticed I have an interest in remailers (duh!). The goodO >>> news is that the new Type-III remailer (Mixminion) is written in Python anda% >>> will therefor be runnable on VMS.  >>> N >>> Any system such as that should be as secure as possible, and I'm wondering8 >>> about the possibility of booting and running off CD. >>> P >>> Has anyone done this?  If so, is XFC enough of a performance booster to makeO >>> this a viable proposition?  What would be a reasonable amount of memory fort >>> the system?d >>>  >>> Thanks for any advice. >>H >>I'm sure it can be done, provided you are prepared to hack up your ownI >>bootable image. I've done it. It's not easy, and the time investment isb >>substantial. > O > It would be an interesting exercise, I'd probably be a lot more familiar witheF > the boot process - and have a lot of coasters - by the end of it. :) >e  F A drive with a write protect button would be a handy thing to have for> a project like this. It could save a lot of time and coasters.  hH >>I'd have to question the reason why, however. What does a write-lockedB >>system disk do for you? ...and are you prepared to deal with the >>ramifications of that? > O > It should mean that physical access would be required - or intimate knowledge 9 > of an undiscovered exploit - to compromise the machine.  >    ------------------------------  $ Date: Wed, 2 Jul 2003 09:42:46 -0700* From: "Alder" <MUNDDGNTDYTV@spammotel.com>6 Subject: Standard Digital Markup Language (SDML) DTD ?+ Message-ID: <3f030b8a$1@obsidian.gov.bc.ca>o  J Anyone know if there is a public one out there somewhere and where to find it ?   Thanks,g   Alders   ------------------------------   Date: 2 Jul 2003 08:00:38 -0500 ; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) S Subject: Re: Tri-architecture cluster demonstrated at DECUS Ottawa Technical Update93 Message-ID: <FHVOH5Onu6UD@eisner.encompasserve.org>   W In article <uIkMa.3698$K51.152@news.cpqcorp.net>, hoff@hp.nospam (Hoff Hoffman) writes:os > In article <qLVVcLA1TSNw@eisner.encompasserve.org>, koehler@eisner.nospam.encompasserve.org (Bob Koehler) writes:0  D > :  Anyone notice anything that didn't work (other than the support > :  contract)?e >   ; >   I certainly know of clustering stuff that doesn't work    K    That's all I wanted to know.  If folks do bump into things that actually3G    don't work in any unsupported configuration it would be nice if theyhH    dropped a note to the rest of us.  I don't expedct VMS engineering to    spend time on it.   ------------------------------   Date: 2 Jul 2003 08:06:00 -0500t; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)o% Subject: Re: VMS backup to NFS shares 3 Message-ID: <2mpv14NB9lgL@eisner.encompasserve.org>e  p In article <7a78a31c.0307011157.2efe5e9e@posting.google.com>, jason.mccormick@lexi.com (Jason McCormick) writes: > E >  According to the vendors of the unit, if the VAX isn't pushing out3G > enough data to fill the stream for continuous writing, the tape driverD > back-catches and stops and starts, etc. over and over.  They claimE > that this puts an enormous amount of wear and tear on the tapes and C > heads inside the unit and the repeated failures I'm seeing is thel > result of this.a  F    On the one hand the data flow issue is nothing new.  It first pokedC    it's head up when streaming 9-tracks were attached to VAX 8200. aG    BACKUP got a lot of work in the area of performance just after.  YounC    might want to look at the recommended account parameters for the C    account being used for BACKUP since a lot of the solution is in .    memory consumption.  G    On the other hand I wouldn't expect a tape drive vendor to sell me aeA    tape drive that wasn't robust enough for the environment I was C    planning to use it in.  Does this vendor list VAXen as supportedf    systems for this drive?   ------------------------------  % Date: Wed, 02 Jul 2003 09:34:19 -0400l8 From: Jim Agnew - VCU/MCV Neurosurgery <jpagnew@vcu.edu>% Subject: Re: VMS backup to NFS shares ' Message-ID: <3F02DF5B.2B72AE79@vcu.edu>o  E Actually, the 8mm units if you have the correct microcode/firmware incF them, they last us for years...  'course we tuned backup to the limit,5 with an entire vaxstation dedicated just to backup...o   jime   Andrew Balaam wrote: > J > We've had repeated problems with both DAT and 8mm Exabyte units over theG > years - the VAXstations we have are just not fast enough to get these J > drive streaming, so they have to keep starting, stopping and going back.G > 8mm Exabyte are worse - the system was really designed for video, notSF > data - too much of the tape is wrapped around the head. Writing fileI > markers is very slow - and with the VMS using ANSI labelling, the labelC > itself is a file as I recall.l > . > tape position lost errors were quite common. > 8 > >>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<< > I > On 01/07/03, 20:57:55, jason.mccormick@lexi.com (Jason McCormick) wrotel) > regarding Re: VMS backup to NFS shares:c > N > > > > I was hoping that a VMS guru could answer a question for me.  I have aN > > > > VAXStation 4000 90A and I'm attempting to get a usable backup solutionK > > > > for it.  The tape unit I've been using on it requires a data streameM > > > > speed that the VAX simply can't keep pushing out to the unit and it's-( > > > > casuing physical damage.  [snip] > > >  > > > Whoa, there!!! > > >o > > > Care to elucidate??? > > >RJ > > > How can BACKUP be causing physical damage to a tape, drive, etc. ??? > G > >  According to the vendors of the unit, if the VAX isn't pushing outaI > > enough data to fill the stream for continuous writing, the tape driveuF > > back-catches and stops and starts, etc. over and over.  They claimG > > that this puts an enormous amount of wear and tear on the tapes andoE > > heads inside the unit and the repeated failures I'm seeing is theiJ > > result of this.  I have an identical unit on a large, fast server withJ > > an LVD card and the unit has performed flawlessly.  The combination ofI > > a slower unit and the SCSI-1 is appearantly not enough to support the I > > unit. They've actually gone as far as to replace the entire unit so I H > > know it's not a lemon unit.  I've been dealing with this issue for 3B > > years so I'm trying to remove the tape unit from the equation.   -- RF "4,000 years ago I made a mistake."  Elrond Half-Elven, in "Fellowship of the Ring"  F "I try not to be right any more than necessary". -- Larry Wall, author of the Perl Language   ------------------------------   End of INFO-VAX 2003.362 ************************