1 INFO-VAX	Fri, 30 Aug 2002	Volume 2002 : Issue 477       Contents: Re: "inview" Article Re: "inview" Article Re: "inview" Article- A big thank you to Carnegie Mellon for CMU_IP ; am/pm  (was Re: Is the HP/Compaq merger really going well?) ? Re: am/pm  (was Re: Is the HP/Compaq merger really going well?) ? Re: am/pm  (was Re: Is the HP/Compaq merger really going well?) ? Re: am/pm  (was Re: Is the HP/Compaq merger really going well?) - Announcing Availability Manager Version 2.2-1 - Announcing Availability Manager Version 2.2-1 P Re: Bad quality of OVMS software engineering was (Re: High quality of HP Softwar Re: banner displayed Re: banner displayed RE: banner displayed2 Re: Catalyst 4006 - no link with alpha server 1000P Re: Current ways to restrict FTP access to user's login directory tree? tree?tre Re: DECW$CALENDAR file format.( Re: EDIT - TPU . Configuration File help( Re: EDIT - TPU . Configuration File help+ Re: Editing the MAIL Undeliverable Response  File open (URGENT!)  Re: File open (URGENT!)  Re: File open (URGENT!) / FORTRAN: Structure constructors & hex constants A From the Peanut Gallery: (was Re: HP-Compaq Merger Went Smoothly) E Re: From the Peanut Gallery: (was Re: HP-Compaq Merger Went Smoothly) E Re: From the Peanut Gallery: (was Re: HP-Compaq Merger Went Smoothly) E Re: From the Peanut Gallery: (was Re: HP-Compaq Merger Went Smoothly) E Re: From the Peanut Gallery: (was Re: HP-Compaq Merger Went Smoothly) E Re: From the Peanut Gallery: (was Re: HP-Compaq Merger Went Smoothly) E Re: From the Peanut Gallery: (was Re: HP-Compaq Merger Went Smoothly) E Re: From the Peanut Gallery: (was Re: HP-Compaq Merger Went Smoothly) ) FYI: IBM's AIX wins DII COE certification   Re: How to configure DECNet-plus  Re: How to configure DECNet-plus  Re: How to configure DECNet-plus Re: How to configure TCP IP $ Re: HP and Dell dump Microsoft Works$ Re: HP and Dell dump Microsoft Works. Re: HP earnings - how did the VMS business do?. Re: HP earnings - how did the VMS business do?. Re: HP earnings - how did the VMS business do?. Re: HP earnings - how did the VMS business do?" Re: HP-Compaq Merger Went Smoothly" Re: HP-Compaq Merger Went Smoothly: HP-UX and Tru64 - Side by Side Comparison for Cluster & DT. Re: Is the HP/Compaq merger really going well?. Re: Is the HP/Compaq merger really going well?. Re: Is the HP/Compaq merger really going well?. Re: Is the HP/Compaq merger really going well?; Re: Is the HP/Compaq merger really going well?  YOU BET!!!!  Lint tool for HPQ CXX  Re: Lint tool for HPQ CXX  Re: Lint tool for HPQ CXX  Re: Lint tool for HPQ CXX  Re: Lint tool for HPQ CXX  Re: Lint tool for HPQ CXX  Re: Old games for VAX system.  Re: Old games for VAX system.   OT - re: From the Peanut Gallery! PRODUCT (PCSI) and TCPIP services  Re: re white boxen Re: re white boxen Re: Reverse LAT  Re: Reverse LAT " Re: silliest thing heard this week" Re: silliest thing heard this week" Re: silliest thing heard this week" Re: silliest thing heard this week" Re: silliest thing heard this week" Re: silliest thing heard this week" Re: silliest thing heard this week" Re: silliest thing heard this week" Re: silliest thing heard this week Re: Simple cluster Re: Simple cluster Re: Simple cluster Support for MPE (OT)- Re: The "Gold" Standard in Disaster Tolerance - Re: The "Gold" Standard in Disaster Tolerance - Re: The "Gold" Standard in Disaster Tolerance - Re: The "Gold" Standard in Disaster Tolerance - Re: The "Gold" Standard in Disaster Tolerance ! tru64.org and openvms.org changes % Re: tru64.org and openvms.org changes % Re: tru64.org and openvms.org changes = Re: White Paper on DT clustering - VMS as the "Gold Standard"   F ----------------------------------------------------------------------   Date: 29 Aug 2002 17:29 CDT ' From: carl@gerg.tamu.edu (Carl Perkins)  Subject: Re: "inview" Article - Message-ID: <29AUG200217294574@gerg.tamu.edu>   / koehler@eisner.aspm.encompasserve.org writes... o }In article <tSab9.15687$GK2.10375@news02.bloor.is.net.cable.rogers.com>, "John Smith" <a@nonymous.com> writes:  }>  9 }> Maybe I should have asked for a Czech instead. Blonde.  }>   } * }   He might not have been to your liking. }   = "Blonde" is the feminine form. "Blond" is the masculine form.   G Well, it would be if modern English use different gendered word forms - B this was the case in the French from which the words were stolen.    --- Carl   ------------------------------  # Date: Fri, 30 Aug 2002 00:48:04 GMT 1 From: "Terry C. Shannon" <terryshannon@attbi.com>  Subject: Re: "inview" Article . Message-ID: <8hzb9.226839$me6.30671@sccrnsc01>  3 "Jan-Erik Sderholm" <aaa@aaa.com> wrote in message  news:3D6CD96B.171227@aaa.com... ; > Well, got it today. I'm running my own consulting company ? > with no VMS support contract in my own name. But I usualy get 4 > whatever is sent out to "VMS customers" in Sweden. > ; > I saw your advertising sudgestions. You know, some people 4 > does a lot of cash writing this kind if things :-)  K Not from the New HP, they don't. Budgets are tight and peecee ads on TV are 
 expensive.   ------------------------------  # Date: Fri, 30 Aug 2002 03:52:08 GMT 1 From: "David J. Dachtera" <djesys.nospam@fsi.net>  Subject: Re: "inview" Article ' Message-ID: <3D6EF122.C9B505C4@fsi.net>    "Terry C. Shannon" wrote:  > 5 > "Jan-Erik Sderholm" <aaa@aaa.com> wrote in message ! > news:3D6CD96B.171227@aaa.com... = > > Well, got it today. I'm running my own consulting company A > > with no VMS support contract in my own name. But I usualy get 6 > > whatever is sent out to "VMS customers" in Sweden. > > = > > I saw your advertising sudgestions. You know, some people 6 > > does a lot of cash writing this kind if things :-) > M > Not from the New HP, they don't. Budgets are tight and peecee ads on TV are  > expensive.  D IMHO, all the more reason why those dollars would be better spent on VMS.   --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  % Date: Thu, 29 Aug 2002 23:09:09 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca> 6 Subject: A big thank you to Carnegie Mellon for CMU_IP, Message-ID: <3D6EE1AB.C92B1272@videotron.ca>  L For years, I have been using the CMU TCPIP package on my all mighty MicrovaxJ II, and had also been using a dial out PPP plugin from Mr Wolfgang Moeller@ until I got a router and cable connection. (thanks to him too !)  N Tonight, I decided to finally put on my microvax II the same TCP software thatM I have on my other VMS nodes (TCPIP services). So I did a final IPNCP NETEXIT D of the CMU software, but preserved the files in their own directory.    M The folks of CMU did a most excellent job for the documentation and software. E I was able to run their software with VAX VMS 7.2 ! Yes, it had a few L reliability problems that required a restart now and then, but for somethingI that was publicly available, I would like to thank all those involved for J having given the VMS community a TCP stack that has been of use to so manyM (and still in use, I would think at shops that don't benefit from big budgets  or hobbyist licences).   ------------------------------    Date: 29 Aug 2002 12:44:11 -0700. From: SPAMSINK2001@YAHOO.COM (Alan E. Feldman)D Subject: am/pm  (was Re: Is the HP/Compaq merger really going well?)= Message-ID: <343f30ae.0208291144.63e9646c@posting.google.com>   q "Bill Todd" <billtodd@metrocast.net> wrote in message news:<NNmb9.16408$On.785648@bin3.nnrp.aus1.giganews.com>... B > "Carl Nelson" <carl.nelson@mcmail.maricopa.edu> wrote in message/ > news:3D6DF1C0.A3294F97@mcmail.maricopa.edu...  >  > ...  >  > >   Why 12:00 m?    @ 12:00 m? Are you kidding? For the entire 60-second interval from# 12:00:00 to 12:00:59.99999...? Why?     K > Silly question, obvious answer.  The more interesting question is why you K > think calling midnight '12 a.m.' makes any more sense than calling it '12 N > p.m.', since that gets back to the root of the entire discussion (which thenJ > becomes rather similar to the discussion between those who like to start: > their loops at 0 and those who like to start them at 1).     = The notation 12 am for midnight is much better than 12 pm for  midnight. Here's why.     Short version:   < With midnight being 12 am, pm changes to am at 00:00. If you< instead make midnight be 12 pm, then pm changes to am at the: smallest possible non-zero amount of time after 00:00. The; former is the simplest, most natural, most consistent, most < convenient, and most-used convention. With the former, am/pm4 functions like a column the same way the numbers do. Therefore, it is better.   
 Long version:    5 AM/PM in a time value simply functions as column. For 5 the sake of argument, let's put the columns in order:     2002-08-29 pm 01:34:45   ; Evidently, we have the following columns: year, month, day,  am/pm, hours, minutes, seconds.    > As time marches on, the value in the seconds column increases.= When it rolls over from 59 to 00, the next-bigger column, the 9 "minutes" column, is incremented by one. When the minutes < column rolls over from 59 to 00, the next-bigger column, the hours, is incremented by one.    = You can see a pattern developing here. Consequently, when the ? hours column rolls over from 11 to 12 (12 being the point where ; the day is incremented by one), then it is most natural and ; consistent to increment the am/pm column. And when pm rolls ? over to am, the day is incremented by one. AM/PM fits in nicely 	 that way.    9 So, when it is "pm 11:59:59", a second later it naturally ? becomes "am 12:00:00". So, with 12 midnight being classified as > am, the am/pm slot simply becomes another column which behaves? just like the other columns during rollovers. That would not be   true if you made midnight 12 pm.   < (Yes, I know that am or pm typically follows the time, which8 puts it "out of order", but we do even more out-of-order> shenanigans with dates but dates still follow the incrementing3 rules as if they were written in decreasing order.)    9 This also simplifies programming. Would you really rather 7 bother to make an exception in your program, actually 2 : exceptions, just for this? No, of course not. Making 12 am< midnight and 12 pm noon is the most elegant solution. And if: you did program it into a clock, the 12:00 pm displayed at7 midnight would only display for that instant and become > essentially invisible. So you'd end up with am for the rest of< the minute that 12:00 is displayed anyway. Or would you have9 your clock display 12:00 PM for sixty seconds immediately > followed by 12:01 AM for the next sixty seconds? I don't think so.    = It is also why all electronic devices use the convention that ; 12 midnight is 12 am. [Well, I certainly hope there are no  # exceptions -- I've never seen one.]    ? Another reason is that making 12 midnight be pm makes life more > complicated. Now we all agree that during the first hour after7 midnight, 12:01, 12:02, ..., 12:59 are AM. So is it not ? simplest and most convenient to have 12:00 midnight also be AM? ; This way, all the times starting with 12 in that first hour ? have the same am/pm value, namely am. What sense would it be to  have    PM 11:59 PM 12:00 AM 12:01	 AM 12:02     be the convention?   > And what about 12:00:01? Would that not be AM? What sense does? it make to have the first instant of the interval of times that ; start with 12 be AM and the rest PM (not including the hour & from 12:00 to 13:00, of course)? None.   4 Not only do all electronic devices use this sensible5 convention [AFAIK], but even the New York City Subway = schedules!!! Yes, they actually publish their schedules. You   can look at them at    - http://www.mta.info/nyct/service/schemain.htm    = You'll see that as times increase down any given column, they < change to bold to indicate PM. And whenever an 11:xx time is= followed by 12:00, the 12:00 is in bold. This is probably due < in part to a computer being used to make such tables. (Check9 the schedule for the no. 2 train and look at Saturday and ; Sunday Northbound times. There you'll find both a 12:00 am   time and a 12:00 pm time.)   ? So with midnight being 12 am, pm changes to am at 00:00. If you < instead make midnight be 12 pm, then pm changes to am at the; smallest non-zero amount of time after 00:00. The former is  better.     It simply makes sense.   > If you need to avoid possible misinterpretation of times, then> yes, go ahead and use times like 11:59PM, or 12:01AM, asumming? you don't need better than a minute accuracy. And don't use the 9 term midnight! I think most people would interpret Friday < midnight to be the "border" between Friday and Saturday, but some might not.     Anyway, there you have it.    3 > Calling midnight 00:00 neatly avoids the issue.       B Agreed, but people but clocks, computers, and VCR's and expect theD am/pm system. That being the case, making midnight 12 am and noon 12 pm is the better convention.   And while people who like M > to start their loops at 1 rather than 0 might suggest that is simply raises E > another (why not 24:00?), they'd have to change the way minutes are M > presented to make that consistent (e.g., 01:00 becomes 24:60, 02:00 becomes  > 01:60, ...). >  > - bill    * And thanks for helping me kill the thread.     Disclaimer: JMHO Alan E. Feldman  spamsink2001 at yahoo dot com    ------------------------------  % Date: Thu, 29 Aug 2002 23:24:05 +0200 9 From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com> H Subject: Re: am/pm  (was Re: Is the HP/Compaq merger really going well?)' Message-ID: <3D6E90F5.D3EF7C16@aaa.com>    About AM/PM...  < Just skip am/pm all together and use 00:00:00  --  23::59.59  as in most parts of the world...   Jan-Erik Sderholm.    ------------------------------    Date: 29 Aug 2002 20:49:46 -0700. From: spamsink2001@yahoo.com (Alan E. Feldman)H Subject: Re: am/pm  (was Re: Is the HP/Compaq merger really going well?)= Message-ID: <b096a4ee.0208291949.5c135598@posting.google.com>   U Jan-Erik Sderholm <aaa@aaa.com> wrote in message news:<3D6E90F5.D3EF7C16@aaa.com>...  > About AM/PM... > > > Just skip am/pm all together and use 00:00:00  --  23::59.59" > as in most parts of the world... >  > Jan-Erik Sderholm.     C Well, until the whole world uses the 24-hour system, many must deal D with am/pm. And if you use am/pm electronic devices, be advised thatC 12:00 am is midnight and 12:00 pm is noon. If one has to use am/pm, C that is the simplest, most convenient, most logical, most sensible, A most mathematical, most-used convention. It is what should be the - convention and it is the de facto convention.    It's really very simple.     Disclaimer: JMHO Alan E. Feldman    ------------------------------  % Date: Fri, 30 Aug 2002 00:34:53 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca> H Subject: Re: am/pm  (was Re: Is the HP/Compaq merger really going well?)+ Message-ID: <3D6EF5BE.1525FC7@videotron.ca>    "Alan E. Feldman" wrote:E > Well, until the whole world uses the 24-hour system, many must deal 
 > with am/pm.   L Seems that the USA is the one that sticks with the am-pm thing.  Rest of theD world (as well as VMS) doesn't have any problems with the real time.   ------------------------------  # Date: Thu, 29 Aug 2002 20:32:40 GMT 0 From: Douglas Hoeger <Douglas.Hoeger@compaq.com>6 Subject: Announcing Availability Manager Version 2.2-18 Message-ID: <pf0tmu4hdeavc8mo5k6e29uie4s9ksutar@4ax.com>  9 Availability Manager Version 2.2-1 may be downloaded from   : <http://www.openvms.compaq.com/openvms/products/availman/>  , Availability Manager Version 2.2-1 provides:  C 	Support for the latest versions of OpenVMS, including the version / that supports the Common Operating Environment.    	New cluster display features.   ? 	Improved scalability for monitoring large numbers of disks on  OpenVMS clusters.   ! 	Minor performance improvements.    	Bug fixes.  4 Complete details are available in the documentation.  C Thank you for your continued patronage and the opportunity to serve  your business needs.  
 Sincerely,   The Availability Manager Team  OpenVMS EngineeringN Hewlett-Packardg   ------------------------------    Date: 29 Aug 2002 15:17:37 -07001 From: susan_skonetski@hotmail.com (Sue Skonetski)h6 Subject: Announcing Availability Manager Version 2.2-1= Message-ID: <857e9e41.0208291417.17724416@posting.google.com>e  9 Availability Manager Version 2.2-1 may be downloaded froma  : <http://www.openvms.compaq.com/openvms/products/availman/>  , Availability Manager Version 2.2-1 provides:  C 	Support for the latest versions of OpenVMS, including the versionl/ that supports the Common Operating Environment.f 	New cluster display features.e? 	Improved scalability for monitoring large numbers of disks ons OpenVMS clusters.d! 	Minor performance improvements.n 	Bug fixes.  4 Complete details are available in the documentation.  C Thank you for your continued patronage and the opportunity to servee your business needs.  
 Sincerely,   The Availability Manager Teamn OpenVMS Engineeringn Hewlett-Packardv   ------------------------------  % Date: Thu, 29 Aug 2002 12:18:06 -0700o. From: JMK <jeffrey.m.klopotic@lmco.nospam.com>Y Subject: Re: Bad quality of OVMS software engineering was (Re: High quality of HP Softwarr/ Message-ID: <3D6E736E.9FB59A76@lmco.nospam.com>W   Fabio Cardoso wrote:  5 > May be HP/Oracle shoul create an OpenVMS Task Forceh0 > to port all the software. Everybody know about2 > the bad quality of developers nowadays. So it is5 > becoming much more rare to find a good profesional,m > with OSes background.    Fabio,  G At one time, configuration management was a job in itself and addressed E exactly the problems you discuss.  They had their own computers, test D software, budget, staff, etc., and their purpose was to test any newE software before putting it on the main system to avoid that which youo	 describe.u  H This problem isn't new, and it appears it hasn't gone away.  Perhaps itsC time to reevaluate expectations which seem to have come with peopleiD being brought up on PC's where you slap in a disk, install, and thenH compute away, not thinking that any problems can arise.  What used to beA a glitch on one's gaming screen is now an error in the eigenvalueC, solution to a billion dollar design project.  H For right now, however, many companies *are* relying on blind faith thatE commericial off the shelf (COTS) software does not need configurationeF management.  For many, that may be true, but for others, they may need1 to consciously decide whether they should or not.i  B The bottom line is that no matter how good one programmer is, theyH cannot address all issues which arise in a software system environment. F There still needs to be a check on how all the software interacts with4 each other, and this is on a system by system basis.  D So, it appears the work is still out there.  Are there companies who have the willingness to do it?   Food for thought,   
 Jeff Klopotici   ------------------------------  % Date: Thu, 29 Aug 2002 12:27:09 -0700h+ From: "xenman" <xenman@sprynet.nospaam.com>  Subject: Re: banner displayedt2 Message-ID: <aklsiv$1dm$1@slb6.atl.mindspring.net>  8 I prefer that they "press" the key rather than "hit" it.  6 Sort of reminds me of the RSTS days where to login the6 user first needed to type in "hello", but the computer8 prompt was "Please say hello".  Saying "hello" would get4 you nowhere as the PDP 11/70 didn't understand voice	 commands.a   ------------------------------  % Date: Thu, 29 Aug 2002 17:00:40 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca>  Subject: Re: banner displayedn, Message-ID: <3D6E8B60.58B49A7B@videotron.ca>  
 xenman wrote:  > : > I prefer that they "press" the key rather than "hit" it.  N I prefer "press any key". It generates support calls of people wanting to knowK where the "any" key is located. (You'd think that Compaq/HP woudl have seenP. fit to add an "ANY" key on theyr keyboards :-)   ------------------------------  % Date: Thu, 29 Aug 2002 17:38:40 -0400v; From: "Webb, William W Raleigh, NC" <wwebb1@email.usps.gov>l Subject: RE: banner displayedJK Message-ID: <BA52530E3149734A9BAABDBBFA808E49028BEA59@rlghncst964.usps.gov>a  : In the southern US, pressing of things, usually other than7 keypads (elevator controls, e.g.), is often referred tol "mashing that button".   (Mash any button to continue)r   WWWebb   -----Original Message-----6 From: "JF Mezei" [mailto:jfmezei.spamnot@videotron.ca]' Sent: Thursday, August 29, 2002 5:00 PM- To: "Info-VAX@Mvb.Saic.Com"  Subject: RE: banner displayed     
 xenman wrote:h >o: > I prefer that they "press" the key rather than "hit" it.  I I prefer "press any key". It generates support calls of people wanting to@ knowK where the "any" key is located. (You'd think that Compaq/HP woudl have seen . fit to add an "ANY" key on theyr keyboards :-)   ------------------------------    Date: 29 Aug 2002 13:10:32 -0700" From: pg1221@hotmail.com (ratheza); Subject: Re: Catalyst 4006 - no link with alpha server 1000e= Message-ID: <75d523ee.0208291210.4be8eed5@posting.google.com>    Alan  C Thanks for responding.  The company that installed the equipment is B coming by next week on an unrelated job, but they said they take aC look at it.  I haven't been able to pull out all of the information C you requested, as I've got a few other high-priority projects goingt on.  Thanks again!   Patrick   e Alan Greig <a.greig@virgin.net> wrote in message news:<ut2pmuob6tssf6g9ktbokbt7ujtb3q28a4@4ax.com>...lD > On 27 Aug 2002 21:04:01 -0700, pg1221@hotmail.com (ratheza) wrote: > I > >I have an issue with my new network equipment and our older machines. tH > >We have 3 Alpha Server 1000s, running Meditech on MAGIC.  We had someI > >downtime scheduled a few weeks ago, so i took the oppportunity to plugsI > >them into my Cisco Catalyst 4006.  But seeing as it was a good time todI > >do this, of course it didn't work - i couldn't even get a link light. oG > >I learned later these servers will only communicate at 10/half, so i G > >set the ports on the 4006 correspondingly, but still no luck.  ThesepE > >servers are mission-critical (hosptial), so unfortunately I cannot F > >take them offline to test.  Any ideas?  I imagine there is an issue > H > There is absolutely no reason why this shouldn't work whether tru64 ot> > VMS is running on the machines.. What were they plugged into
 > previously?p > @ > Also you can stick DE-500 ethernet cards into them if you need > 100Mb/sec  > G > >with broadcast, but would that affect the physical link notificationtH > >on the cisco?  I can get more detailed on the equip info when i am inG > >the hospital, don't know if any of you have worked with this kind of $ > >set up?  hope not =)....  thanks! > @ > Please provide any further info you can including VMS or tru64
 > version. > 
 > >patrick   ------------------------------  % Date: Thu, 29 Aug 2002 23:29:35 -0500a( From: Rich Jordan <duodec@speakeasy.net>Y Subject: Re: Current ways to restrict FTP access to user's login directory tree? tree?trei, Message-ID: <3D6EF4AF.3080104@speakeasy.net>   Keith A. Lewis wrote:v > jordan@ccs4vms.com (Rich Jordan) writes in article <cc5619f2.0208290710.43e1ea58@posting.google.com> dated 29 Aug 2002 08:10:05 -0700: > H >>The Q's TCPIP apparently doesn't have a way to restrict the CD commandE >>so that the user is confined to their login directory tree, meaning F >>they can attempt to move anywhere in the filesystem (and in fact twoC >>users have to do just that to move files to special import/export F >>directories).  The remaining users need access to their home tree soG >>the anonymous aliases wouldn't work (there are too many users for the.E >>10 alias limit anyway).  The TCPIP V5.3 docs seem to be the same asaB >>the V5.1 in this area, and the release notes don't show anything# >>useful so no joy with an upgrade.r >  > L > This really isn't a FTP/TCPIP issue; it's one of general file protection. K > You don't really want users to have one set of files they can access when'J > they are logged in interactively vs. a more restricted set when they areM > connected to the FTP server, do you?  They could circumvent that by logging K > in interactively and copying files between their home directory and thoselL > other places they aren't supposed to mess with, and then ftp them (or vice	 > versa).  > M > To set all (regular) files on the volume "USER" so that only the owner (and 9 > no other normal users) can access them, the command is:  > E >     $ SET FILE DISK$USER:[*...]*.*;* /PROT=(OWNER:RWED,GROUP,WORLD)  > D > For more complex needs I refer you to the _OpenVMS Guide to System > Security_ manual.  > - > --Keith Lewis              klewis$mitre.orgi@ > The above may not (yet) represent the opinions of my employer.  G Interactive access is more restricted.  When they log in interactively oF they are held in a captive menu that can run specific programs and do D specific tasks; they can't get to the DCL prompt and flit about the A system.  Since FTP access bypasses the menu, and all its layered aH security, they have greater access (from a certain point of view).  And H since some users have to change directories (to defined logical names), H all the other users know about those logicals, (you know how users are) 2 and we have the usual percentage of experimenters.  G System security is reasonably tight, and we're fixing up the remainder uF now that its necessary (it wasn't with only interactive and pathworks H access going on, given the captive accounts) but it wouldn't seem to be G such an issue to restrict FTP to only working in the directory tree of n the users home directory.h  I It turns out that Hunter's FTP should be able to do what we need.  We're o going to test it and make sure.d   Rich Jordans   ------------------------------  % Date: Thu, 29 Aug 2002 17:23:20 -0400 ; From: "Brian Tillman" <tillman_brian@notnoone.notnohow.com>b' Subject: Re: DECW$CALENDAR file format. $ Message-ID: <3d6e90fc$1@news.si.com>  K >Since HP insists on keeping the "open"  tagged to VMS, how about releasingr thenH >file format of the DEW$CALENDAR.DWC file so that we can start using theI >application more intelligently ? (or publish an API that can make use ofg it).  K I played around with that for some time, with the idea that I would write a H character-cell interface.  I had much of the format doped out, but there> were some things I could make sense of, so I never pursued it. --A Brian Tillman                   Internet: tillman_brian at si.com A Smiths Aerospace                          tillman at swdev.si.comp= 3290 Patterson Ave. SE, MS      Addresses modified to preventy< Grand Rapids, MI 49512-1991     SPAM.  Replace "at" with "@"8        This opinion doesn't represent that of my company   ------------------------------  % Date: Thu, 29 Aug 2002 17:35:21 -0400 ; From: "Brian Tillman" <tillman_brian@notnoone.notnohow.com>a1 Subject: Re: EDIT - TPU . Configuration File helps$ Message-ID: <3d6e9514$1@news.si.com>  L >         I use to have a configuration file that would allow me to do basicG >task with the pf keys. It would kind of emulate the edt editor . (edit  /edt)oJ >commands. Does any one have a configuration file that has good shortcuts.  K The change mode keypad commands, the nokeypad commands, or the command lineuI commands?  Obviously, the first of these is best emulated with SET KEYPAD1 EDT. --A Brian Tillman                   Internet: tillman_brian at si.comaA Smiths Aerospace                          tillman at swdev.si.comt= 3290 Patterson Ave. SE, MS      Addresses modified to prevent < Grand Rapids, MI 49512-1991     SPAM.  Replace "at" with "@"8        This opinion doesn't represent that of my company   ------------------------------  % Date: Thu, 29 Aug 2002 23:14:46 -0400s, From: "Francisco Ortega" <fortega@iblues.cc>1 Subject: Re: EDIT - TPU . Configuration File helptA Message-ID: <6kBb9.52326$Hd4.8093553@e3500-atl1.usenetserver.com>a   THANKS to everybody.      7 "Francisco Ortega" <fortega@iblues.cc> wrote in messageo; news:BKcb9.35238$Hd4.7623558@e3500-atl1.usenetserver.com...  > Hi, G >          I use to have a configuration file that would allow me to doe basictH > task with the pf keys. It would kind of emulate the edt editor . (edit /edt) K > commands. Does any one have a configuration file that has good shortcuts.c > Thanks >a > Franciscoo >  >t >  >e   ------------------------------  % Date: Thu, 29 Aug 2002 15:17:25 -0400o- From: "Peter Weaver" <peter.weaver@stelco.ca> 4 Subject: Re: Editing the MAIL Undeliverable Response6 Message-ID: <akls0c$1k4vi0$1@ID-141708.news.dfncis.de>  7 "Jeff Barnes" <barnesjw@dfo-mpo.gc.ca> wrote in messagee& news:YmKa9.1106$i%.266208@localhost...L > How would one modify the"Undeliverable Mail" response you receive when the atK > users account has the DISMAIL flag set?  The stock undeliverable responsen > includes the following:    >...  H You might want to tell us what IP stack you're using on top of VMS since& that is what is returning the message.   -- Peter WeaverL Opinions are my own, and do not reflect the opinions of my employer, nor theK company that it sub-contracts to, nor the company that it sub-contracts to.    ------------------------------  # Date: Thu, 29 Aug 2002 20:33:46 GMTh( From: Carlo <carlo.pettirossi@libero.it> Subject: File open (URGENT!)) Message-ID: <3D6E8559.5FF7C19E@libero.it>    Hi everyone,  H I'm a beginner in Fortran. I wanted to have some informations concerningB the opening of file copied from a VMS machine onto a unix computer (probably an Alpha).E At the moment there is a fortran application running on unix, opens a?C file coming from the VMS machine (say vms.dat) previously convertedPB through a C application which doesn't do anything else but copying> string, integer, and byte values as they are onto another file/ (unix.dat) as they are, but converts the float.h> This is a solution that I don't like that much. Aside from its> inefficiency (2 programs doing what only one can do), what theE programmer did is converting only the VAXF (real*4) data. If the data @ inside the original file will change, the modifications affect 2@ programs in 2 different languages. I don't know why the originalA programmer chose this solution. As he is a programmer with a good C knowledge of C and Fortran - well...surely better than mine :-) - I E thought that he did like that, cause the CONVERT="...." placed in the H OPEN command doesn't work well. I tried just today for the first time toE use the conversion, overriding the C application. The results are not ? good. I opened the file vms.dat (copied on the unix machine) as H SEQUENTIAL, STREAM, UNFORMATTED. I'm pretty sure that this is the reason of the bad data reading. Anyway, my questions are: B 1) how does the CONVERT "behave" when it has to cope with integer,: character, byte and real values together in the same file?H 2) can someone tell me how it works? How does it recognize the format ofH data just by reading a file? Is there a tag within the record indicating1 that the data is an integer, a float or whatever? 2 I need to find a solution asap. Thanks in advance.   Carlo    ------------------------------    Date: 29 Aug 2002 16:42:41 -0600- From: koehler@encompasserve.org (Bob Koehler)n  Subject: Re: File open (URGENT!)3 Message-ID: <$T8aYJ5Pk4kF@eisner.encompasserve.org>n  T In article <3D6E8559.5FF7C19E@libero.it>, Carlo <carlo.pettirossi@libero.it> writes: > Hi everyone, > J > I'm a beginner in Fortran. I wanted to have some informations concerningD > the opening of file copied from a VMS machine onto a unix computer > (probably an Alpha).  G > At the moment there is a fortran application running on unix, opens a.E > file coming from the VMS machine (say vms.dat) previously converted2D > through a C application which doesn't do anything else but copying@ > string, integer, and byte values as they are onto another file1 > (unix.dat) as they are, but converts the float.s@ > This is a solution that I don't like that much. Aside from its@ > inefficiency (2 programs doing what only one can do), what theG > programmer did is converting only the VAXF (real*4) data. If the datacB > inside the original file will change, the modifications affect 2B > programs in 2 different languages. I don't know why the originalC > programmer chose this solution. As he is a programmer with a goodAE > knowledge of C and Fortran - well...surely better than mine :-) - IiG > thought that he did like that, cause the CONVERT="...." placed in the8J > OPEN command doesn't work well. I tried just today for the first time toG > use the conversion, overriding the C application. The results are notvA > good. I opened the file vms.dat (copied on the unix machine) aseJ > SEQUENTIAL, STREAM, UNFORMATTED. I'm pretty sure that this is the reason > of the bad data reading.  G    There's a vary good chance that the file type you stored into is notiF    compatable with the other qualifiers of the OPEN statement.  MostlyD    this is because the VMS file system tracks these for you and UNIXE    doesn't.  Also you must correctly choose ASCII or binary mode whenaG    you transfer the file from the UNIX to the VMS systems.  Reading andvF    writing in C may be altering the file attributes in addition to the    contents.  >    Your belief that the UNIX system is an Alpha means that theG    non-floating point representations are identical to those on the VMSfD    system.  Alpha also understands the VAX F floating point, but the=    UNIX Fortran compiler for Alpha won't do it.  That leaves:   K    1)  If the VMS system is also an Alpha, compile and operate the program e*       under VMS using IEEE floating point.  J    2)  Convert only the floating point (what your C program is doing now).   > Anyway, my questions are:dD > 1) how does the CONVERT "behave" when it has to cope with integer,< > character, byte and real values together in the same file?  C       CONVERT= in the OPEN converts unformatted data to the native pD       representation of the target machine, IF the type declaration C       is accurate.  For example, if you have some old Fortran whichnG       stores characters in INTEGER, READ, or DOUBLE PRECISION variablesaB       they will be mangled per numeric conversion, rather than theH       CHARACTER conversion.  If you EQUIVALENCE everything to a big byteF       (LOGICAL*1) array and write out that array, you'll realy hose up       the conversion.n  C       What value are you using for CONVERT= ?  Certainly there is aiD       value which will do the correct conversion if the code is good?       and you are correct about the UNIX system being an Alpha.o  0       OBTW, is the VMS system a VAX or an Alpha?  J > 2) can someone tell me how it works? How does it recognize the format ofJ > data just by reading a file? Is there a tag within the record indicating3 > that the data is an integer, a float or whatever?v4 > I need to find a solution asap. Thanks in advance.  H    It does not recognize the format from the file, it converts accordingE    to the variable type your reading to/from and the target system ofi    the CONVERT=.  F    And if your UNIX compiler supports the CONVERT= extension, you must7    be carefull to convert only on one system, not both.0   ------------------------------  % Date: Thu, 29 Aug 2002 14:49:18 -0700p. From: JMK <jeffrey.m.klopotic@lmco.nospam.com>  Subject: Re: File open (URGENT!)/ Message-ID: <3D6E96DE.EE0DB55E@lmco.nospam.com>e  D Be aware that in VMS, there is a Fortran carriage control format forE file output that is different from other file formats.  It places '0'e@ and '1' for line and page feeds respectively.  Depending on whatG programs you have written, this could cause trouble in reading what you2 believe to be ASCII data.M  F Also, in transferring files between different OS's, subtle line formatC issues may arise -- especially if you have linux tools around, liker? tar-gz.  They work, but need to be treated with care if you are"G converting anything to be used on VMS.  Open files up in the EVE editorr6 to see if you have any weird leftovers from detarring.  G If you have a Fortran file and want to change format to something else,-H look up Convert and FDL (file definition language) to see how to convert  files between different formats.   Hope this helps,   Jeff   ------------------------------   Date: 29 Aug 02 21:26:13 MDT" From: ivie@cc.usu.edu (Roger Ivie)8 Subject: FORTRAN: Structure constructors & hex constants% Message-ID: <LiibsDzFDA$w@cc.usu.edu>i  G I've recently been playing with new features of FORTRAN that I've nevereJ used before, and have stumbled across a stumper. When trying to initializeL a structure of derived type, I get warnings if I use hex constants that have. the sign bit set. For example, compiling this:   	program boogerh 	type t  	  integer x	 	end typet
 	type(t) y 	y=t(16#98765432)  	end   I get the following:   $ for booger           y=t(16#98765432) .............^J %F90-W-WARNING, The INTEGER(KIND=4) value is out-of-range.   [16#98765432]6 at line number 6 in file DISK$SK5:[I.IVIE]BOOGER.FOR;1 $-  J The best workaround I've been able to come up with is something like this:   	program boogerR 	type ts 	  integer x	 	end typeh
 	type(t) y 	integer bit31 	parameter(bit31=16#80000000)o 	y=t(16#18765432 .or. bit31) 	end  M Basically, baffling the compiler with b******t. Surely there must be a bettere way? -- eN -------------------------+----------------------------------------------------+ Roger Ivie               | Free the ROT-13! : ivie@cc.usu.edu          |    http://www.freesklyarov.org/ http://cc.usu.edu/~ivie/ |   ------------------------------  # Date: Thu, 29 Aug 2002 20:55:26 GMTk1 From: "Terry C. Shannon" <terryshannon@attbi.com>0J Subject: From the Peanut Gallery: (was Re: HP-Compaq Merger Went Smoothly). Message-ID: <2Tvb9.225169$me6.30213@sccrnsc01>  5 "Bill Todd" <billtodd@metrocast.net> wrote in messageG< news:beXa9.130290$Aw4.5478762@bin2.nnrp.aus1.giganews.com... >k> > "Terry C. Shannon" <terryshannon@attbi.com> wrote in message8 > news:nUUa9.38653$_91.1034@rwcrnsc51.ops.asp.att.net...K > > Well, seems to me that HPQ pretty much made its numbers, and it sure ast > hellH > > doesn't look as bad as Sun (which might be a bargain at four bucks). >rG > I guess if one considers a $2 billion loss for the quarter to be good  news,a	 > anyway.r >aI > To avoid a great deal of additional (and unnecessary) typing, I'll justiL > quote a note that I sent to Mike Magee tonight about various contributionsE > Terry has recently made to The Inquirer that sheds light on Terry'sn, > objectivity about this, and other, topics: >0 >t > Dear Mike, > L > Since I read The Inquirer for content rather than vendor PR, articles suchI > as August 27th's on the supposed Superdome 'records' are disappointing.nI > While Terry may get paid or otherwise derive compensation for producinge such5 > trash, I hope you don't get paid for publishing it.k > L > First of all, the TPC-C 'record' is of course rather carefully engineered:H > the top of the non-clustered TPC-C result column remains occupied by aJ > 128-processor Fujitsu PRIMEPOWER machine running (gasp!) SPARCs - but itL > happens to be running a SymfoWARE database rather than Oracle, hence fails% > to meet Terry's arbitrary criteria.s >eH > Now, one could bluster about having to use twice as many processors toK > achieve only a modest performance edge, except that this is precisely the0K > same strategy that Superdome (with 64 processors) used to squeak past the L > IBM Regatta (with 32 processors) very close on its heels in third place onA > the list.  And while Terry was careful to call out the marginaldJ > per-processor performance edge that Superdome enjoys over SunFire in hisF > TPC-H comparisons, he was conspicuously silent about this nearly 2:1K > per-processor performance edge that the IBM Regatta enjoys over Superdomeh in > TPC-C. >mK > He also neglected to mention how the SunFire compares with Superdome on aaB > TPC-H cost/performance basis (the SunFire is slightly superior - surprise!),eL > though he was careful to point out the marginal cost/performance advantage* > Superdome has over the Regatta in TPC-C. >vL > Terry's resemblance to a Carly&Curly sock-puppet is hardly confined to theC > above article.  On the same day, after a title and single leading,	 paragraphhJ > gratuitously deriding Sun and anyone else who might dare to question theF > wisdom of the cHumPaq merger, he then quotes HP's Q3 financial press releases@ > VERBATIM (with an occasional word and, four times, a brief andH > largely-uninformative sentence of his own thrown in, and elsewhere theG > occasional press release sentence or paragraph left unquoted) FOR THEDI > REMAINING 14 PARAGRAPHS OF HIS 'ARTICLE'.  If this is what you considerxJ > journalism, I'm afraid your future in the field may be limited - and theF > inclusion of a copyright symbol following such blatant plagiarism is > out-right farce. >oD > The press release that Terry quoted without attribution appears atD > http://www.hp.com/hpinfo/investor/financials/quarters/2002/q3.htm.J > Interesting aspects that he neglected to point out include the fact that HPH > lost over $2 billion in the quarter but chose to characterize somewhat moreI > than that amount as an unaudited pro-forma adjustment to make it appearn thatL > they would have made money had it not been for the costs of the merger andL > restructuring (though they seem happy to include as revenue and profit anyJ > benefits of the merger and restructuring, such as they may be), and that theaL > only profitable segments for the quarter (at least if one ignores that $2+? > billion in additional expenses that HP would like us to) wereaH > Imaging/Printing ($813 million) and Services ($275 million), while theK > Personal Systems Group lost $198 million and the Enterprise Systems Groupo > lost $422 million. >lL > Not one to miss an opportunity to tout his patron, Terry included an overtL > (and wholly irrelevant) plug for HP's (supposed) 'cryptographic expertise'K > at the end of his "Paranoia or Pragmatism?" article just one day earlier.cH > Since he has done this sort of thing for years in his own 'newsletter'E > (another rather dubious example of the art of journalism) one can'teI > characterize it as new behavior on his part, but at least in my limitedt0 > experience it is somewhat new to The Inquirer. >lE > Oh, dear.  After all the above my suspicions were aroused about the H > Superdome article as well, and - lo and behold! - it too seems to have beenK > lifted (only slightly less verbatim than the financial article) from this  HPK > press release:  http://www.hp.com/hpinfo/newsroom/press/26aug02b.htm.  At_L > least he had the decency to leave the copyright symbol off that 'article'. >oL > So, Mike:  When I first pointed out, in far less detail and without havingL > yet noticed the blatant plagiarism, the close relationship between Terry'sF > Inquirer articles and HP PR, you effectively replied "Well, you know Terry,J > but I'll copy him on this."  That may have been a marginally appropriateJ > (though I'd still say rather blinkered) response given the brevity of myF > initial comment, but I suggest that it's far from adequate given the > additional information above.c >tL > So I hope that you'll publish this rebuttal to help even things up, and be ac2 > bit more careful what you publish in the future. >s  K You will be pleased to note that Mister Magee did in fact publish this welltJ thought out and articulate rebuttal to the ravings of a patriotic AmericanL who has served his country, fought in a war, is an NRA life member, and will3 long remember September 11. And Veteran's Day, too.   2 Thanks for adding a bit of humour to Magee's site.  K As for "blatant plagiarism," be advised that my son will be admitted to the E bar in November. I can't help but notice that most trade press organsyJ leverage press releases, so I guess they're guilty of the same mortal sin, eh?0  F Hint: I would strongly urge a prudent individual to carefully read theJ statutes and findings on libel and slander. But in this case, just keep it. up. And ratchet it up a few more notches, too.  F Thank you, and have an absolutely ginger-peachy day today, and an even better one tomorrow!   ------------------------------  % Date: Thu, 29 Aug 2002 17:32:40 -0400f- From: JF Mezei <jfmezei.spamnot@videotron.ca>wN Subject: Re: From the Peanut Gallery: (was Re: HP-Compaq Merger Went Smoothly), Message-ID: <3D6E92DD.3687F684@videotron.ca>   "Terry C. Shannon" wrote:h- > who has ... is an NRA life member, and will-  H Admitting you are a devout follower of Charlton Heston's religion is notN something you should be doing here. You aren't going to be making friends, andC those who do not agree with that religion will think lesser of you.    ------------------------------  # Date: Thu, 29 Aug 2002 22:29:34 GMT01 From: "Terry C. Shannon" <terryshannon@attbi.com>eN Subject: Re: From the Peanut Gallery: (was Re: HP-Compaq Merger Went Smoothly). Message-ID: <ifxb9.225968$me6.30410@sccrnsc01>  : "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message& news:3D6E92DD.3687F684@videotron.ca... > "Terry C. Shannon" wrote:i/ > > who has ... is an NRA life member, and willn >sJ > Admitting you are a devout follower of Charlton Heston's religion is notL > something you should be doing here. You aren't going to be making friends, and-E > those who do not agree with that religion will think lesser of you.l  J Nope, I am a citizen of America (the greatest country in the world, by theH way) and I know a wee bit about Constitutional law, to wit, the right toD bear arms. I currently do not possess a firearm (at least here in myJ soon-to-be-vacated residence) but such possession is a right and it is oneK that I will defend whether you like it or not. As for my religion, which ismG actually none of your business whatsoever, I was baptised Episcopalian,tI attended a Methodist church, and a Catholic high school. I am ordained inaD another faith, but it has nothing to do with firearms and is, again,! absolutely none of your business.t  L Charlton Heston? Methinks he's a has-been actor with Altzheimer's Disease. IH sure as hell am not a devout follower of him, or whatever "religion" you seem to think he embodies.  I You, or any other customer of mine who has a legitimate beef with this is F entitled to a pro-rated refund of all the money you have sent me. Fair enough?n  @ Long Live America, and Remember September 11: Payback's a bitch!   ------------------------------  # Date: Fri, 30 Aug 2002 00:49:50 GMTu* From: "Jack Peacock" <peacock@simconv.com>N Subject: Re: From the Peanut Gallery: (was Re: HP-Compaq Merger Went Smoothly)@ Message-ID: <Oizb9.24938$On.1294089@bin3.nnrp.aus1.giganews.com>  : "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message& news:3D6E92DD.3687F684@videotron.ca... > "Terry C. Shannon" wrote:n/ > > who has ... is an NRA life member, and willaE > those who do not agree with that religion will think lesser of you., >hL Those of us who are fellow travellers see it as one more plus on the side ofB credibility.  There are parts of the US where an NRA membership isK considered to be an asset and often appear in political ads.  It's a factora! when I determine who to vote for.     Jack Peacockl   ------------------------------  # Date: Fri, 30 Aug 2002 03:06:32 GMT-1 From: "David J. Dachtera" <djesys.nospam@fsi.net>@N Subject: Re: From the Peanut Gallery: (was Re: HP-Compaq Merger Went Smoothly)' Message-ID: <3D6EE66F.88ED3551@fsi.net>    JF Mezei wrote:5 >  > "Terry C. Shannon" wrote:r/ > > who has ... is an NRA life member, and will  > J > Admitting you are a devout follower of Charlton Heston's religion is notP > something you should be doing here. You aren't going to be making friends, andE > those who do not agree with that religion will think lesser of you.    Two-edged sword, guy.m  F I don't hold with the NRA 100%, but I do support the right to keep andE bear arms, including Concealed Carry. I believe that the thought of aeC potential victim also firing in self-defense makes a more effectivehD deterrent than a death penalty in a court system like ours that's soF bogged down with frivolous or even ridiculous litigation it can't hopeD to deal justice fairly. There's a reason why Illinois' moratorium on- executions remains in place as you read this.a  E I understand you may have strong feelings in the other direction, but F like the man said: when guns are outlawed only outlaws will have guns.  H ...or as a another man said, I may disagree with what you say but I will( defend to my death your right to say it.   -- t David J. Dachterac dba DJE Systems" http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/s   ------------------------------  % Date: Fri, 30 Aug 2002 00:05:49 -0400d( From: David Froble <davef@tsoft-inc.com>N Subject: Re: From the Peanut Gallery: (was Re: HP-Compaq Merger Went Smoothly), Message-ID: <3D6EEF1D.4020209@tsoft-inc.com>   Jack Peacock wrote:c  < > "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message( > news:3D6E92DD.3687F684@videotron.ca... >  >>"Terry C. Shannon" wrote:e >>. >>>who has ... is an NRA life member, and will >>>eE >>those who do not agree with that religion will think lesser of you.u >> >>N > Those of us who are fellow travellers see it as one more plus on the side ofD > credibility.  There are parts of the US where an NRA membership isM > considered to be an asset and often appear in political ads.  It's a factoro# > when I determine who to vote for.t >    Jack Peacock       I Things are going downhill rapidly.  Now we have an NRA argument in c.o.v.a  O I do have to ask this question of those that think it's guns that kill and not oK people.  When a drunk driver kills someone, do we blame the driver, or ban t automobiles?   Dave   ------------------------------  # Date: Fri, 30 Aug 2002 04:32:24 GMTn* From: "Bill Todd" <billtodd@metrocast.net>N Subject: Re: From the Peanut Gallery: (was Re: HP-Compaq Merger Went Smoothly)@ Message-ID: <szCb9.28641$On.1492635@bin3.nnrp.aus1.giganews.com>  5 "David Froble" <davef@tsoft-inc.com> wrote in messagei& news:3D6EEF1D.4020209@tsoft-inc.com...   ...o  K > Things are going downhill rapidly.  Now we have an NRA argument in c.o.v.   F Not a good sign, unless a miracle occurs and it becomes the *one* such3 discussion in history to remain civil and sensible.h   >eL > I do have to ask this question of those that think it's guns that kill and notLL > people.  When a drunk driver kills someone, do we blame the driver, or ban > automobiles?  L And since you're one of the few people here with whom I can remain congenialL while having a conversation about something we may disagree on, even if said9 discussion becomes, shall we say, animated, I'll respond:s  K We blame the driver, *but* we also place significant controls (e.g., driverwG testing and in many cases education, driver licensing, vehicle testing, G vehicle registration) on the use of automobiles (without banning them),t because they *are* dangerous.d   - bill   ------------------------------  # Date: Fri, 30 Aug 2002 04:58:50 GMTe* From: "Bill Todd" <billtodd@metrocast.net>N Subject: Re: From the Peanut Gallery: (was Re: HP-Compaq Merger Went Smoothly)B Message-ID: <eYCb9.172182$Aw4.7360189@bin2.nnrp.aus1.giganews.com>  < "Terry C. Shannon" <terryshannon@attbi.com> wrote in message( news:2Tvb9.225169$me6.30213@sccrnsc01...   ...   H > You will be pleased to note that Mister Magee did in fact publish this well% > thought out and articulate rebuttal9  K I would indeed be pleased if I could find it:  could you provide a pointer?i   ...t  I > As for "blatant plagiarism," be advised that my son will be admitted toa the0G > bar in November. I can't help but notice that most trade press organsgL > leverage press releases, so I guess they're guilty of the same mortal sin, > eh?S  E I think the relevant distinction is between 'leveraging' and 'quotingl0 extensively, verbatim, and without attribution'.   >lH > Hint: I would strongly urge a prudent individual to carefully read the- > statutes and findings on libel and slander.d  / See above, plus the definition of 'plagiarism'.i   - bill   ------------------------------  # Date: Fri, 30 Aug 2002 04:38:22 GMTn1 From: LESLIE@JRLVAX.HOUSTON.RR.COM (Jerry Leslie)v2 Subject: FYI: IBM's AIX wins DII COE certification= Message-ID: <2FCb9.374636$q53.12342240@twister.austin.rr.com> # Keywords: aix,dii_coe,certification-  9    http://www.gcn.com/vol1_no1/daily-updates/19710-1.htmlg'    IBM's AIX wins DII COE certifications  H   "IBM Corp.'s current proprietary version of the Unix operating system,A    AIX 5L, has received certification for the Defense Informations1    Infrastructure's Common Operating Environment.   F    AIX 5L is the first platform to be certified under Version 4 of DIID    COE's Kernel Platform Compliance Program, said Greg Lefelar, Unix2    server sales manager for the IBM federal group.  	    [snip](  D    To get the certification, AIX 5L Version 5.1 was tested on an IBMG    7044-170 workstation, which Lefelar said is part of the same productIC    family as the p690. A list of validated platforms in the DII COEe2    Kernel Platform Compliance Program is online at)    diicoe.disa.mil/coe/kpc/KPCP_VPL.htm."e  ! Note: that link gets a 404 error.   2 --Jerry Leslie   (my opinions are strictly my own)9   Note: leslie@jrlvax.houston.rr.com is invalid for email.   ------------------------------  # Date: Thu, 29 Aug 2002 18:22:56 GMTy From: dittman@dittman.neta) Subject: Re: How to configure DECNet-plus-6 Message-ID: <4Etb9.4205$Wb1.4176@nwrddc01.gnilink.net>  / Peter LANGSTOEGER <peter@langstoeger.at> wrote: ^ : In article <3D6D7F69.ADF189B8@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes: :>Yong Liu wrote:hO :>> I installed DECNet-plus. run decnet-config. The problem is even a show host 0 :>> local file takes forever. What's wrong here? :>> > :>> Any pointers? Detailed direction will be most appreciated. :> :>PRODUCT REMOVE DECNET-PLUS( :>@VMSINSTAL the real DECNET  (decnet 4) :>@NETCONFIG :>N :>If you do not need the functionality of decnet-plus, the real decnet (decnet> :>4) is much smaller and much simpler to configure and manage.  O : Nope. I used both in the last 19 years. And DECnet-Plus is definitely better.eJ : DECnet Phase 5 was crap with DECnet-VAX Extensions (V5.4) but that was a
 : decade ago.   > I have to agree.  I didn't like DECnet-Plus, preferring DECnet> Phase IV, until I took a class on DECnet-Plus.  Once I had the= concepts down on the design of the OSI model, everything fell 9 into place and I find DECnet-Plus easier to configure and  manage.  -- t Eric Dittman dittman@dittman.netl= Check out the DEC Enthusiasts Club at http://www.dittman.net/    ------------------------------  % Date: Thu, 29 Aug 2002 20:34:12 -0400a( From: David Froble <davef@tsoft-inc.com>) Subject: Re: How to configure DECNet-pluse* Message-ID: <3D6EBD84.30803@tsoft-inc.com>   dittman@dittman.net wrote:  1 > Peter LANGSTOEGER <peter@langstoeger.at> wrote:e` > : In article <3D6D7F69.ADF189B8@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes: > :>Yong Liu wrote: Q > :>> I installed DECNet-plus. run decnet-config. The problem is even a show hosti2 > :>> local file takes forever. What's wrong here? > :>> @ > :>> Any pointers? Detailed direction will be most appreciated. > :> > :>PRODUCT REMOVE DECNET-PLUS* > :>@VMSINSTAL the real DECNET  (decnet 4) > :>@NETCONFIG > :>P > :>If you do not need the functionality of decnet-plus, the real decnet (decnet@ > :>4) is much smaller and much simpler to configure and manage. > Q > : Nope. I used both in the last 19 years. And DECnet-Plus is definitely better.sL > : DECnet Phase 5 was crap with DECnet-VAX Extensions (V5.4) but that was a > : decade ago.r > @ > I have to agree.  I didn't like DECnet-Plus, preferring DECnet@ > Phase IV, until I took a class on DECnet-Plus.  Once I had the? > concepts down on the design of the OSI model, everything fello; > into place and I find DECnet-Plus easier to configure andi	 > manage.S >   P I've never used DECnet-Plus.  I'm not knocking it, Can't, since I don't know it.  + But, for ease of configuration and running:   C @netconfig  (and answer a few questions, saying No to most of them)m	 @startnett  J Can't get much easier.  Yeah, there is setting up known nodes, another no P brainer,  SET NODE XXX, answer with a DECnet address, such as 1,1, and re-enter O the node name (don't know why).  Then doing the same for the perm database, or aQ copying the volitile database to the perm database, whatever suits you.  Tell me l& again how much simpler DECnet Plus is?   Dave   ------------------------------  # Date: Fri, 30 Aug 2002 03:55:22 GMTi From: dittman@dittman.nete) Subject: Re: How to configure DECNet-plus 5 Message-ID: <K0Cb9.1199$2M3.344@nwrddc04.gnilink.net>R  ) David Froble <davef@tsoft-inc.com> wrote:u : dittman@dittman.net wrote:  2 :> Peter LANGSTOEGER <peter@langstoeger.at> wrote:a :> : In article <3D6D7F69.ADF189B8@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes:  :> :>Yong Liu wrote:R :> :>> I installed DECNet-plus. run decnet-config. The problem is even a show host3 :> :>> local file takes forever. What's wrong here?t :> :>>  A :> :>> Any pointers? Detailed direction will be most appreciated.l :> :>  :> :>PRODUCT REMOVE DECNET-PLUSr+ :> :>@VMSINSTAL the real DECNET  (decnet 4)u :> :>@NETCONFIGa :> :>mQ :> :>If you do not need the functionality of decnet-plus, the real decnet (decnetcA :> :>4) is much smaller and much simpler to configure and manage.i :> eR :> : Nope. I used both in the last 19 years. And DECnet-Plus is definitely better.M :> : DECnet Phase 5 was crap with DECnet-VAX Extensions (V5.4) but that was ai :> : decade ago. :>  A :> I have to agree.  I didn't like DECnet-Plus, preferring DECnet A :> Phase IV, until I took a class on DECnet-Plus.  Once I had thet@ :> concepts down on the design of the OSI model, everything fell< :> into place and I find DECnet-Plus easier to configure and
 :> manage. :> o  R : I've never used DECnet-Plus.  I'm not knocking it, Can't, since I don't know it.  - : But, for ease of configuration and running:i  E : @netconfig  (and answer a few questions, saying No to most of them)s : @startnetr  L : Can't get much easier.  Yeah, there is setting up known nodes, another no R : brainer,  SET NODE XXX, answer with a DECnet address, such as 1,1, and re-enter Q : the node name (don't know why).  Then doing the same for the perm database, or  S : copying the volitile database to the perm database, whatever suits you.  Tell me r( : again how much simpler DECnet Plus is?  < DECnet-Plus requires @net$config, and defining nodes is done: with decnet_register.  If you understand the ISO model the; configuration of DECnet-Plus makes sense, too.  I find them 8 both the same level for simple configurations.  For more9 complex configurations, esp. trying to bridge DECnet over,: a WAN, DECnet-Plus is easier as you can use DECnet-over-IP< instead of trying to get your routers to encapsulate DECnet. -- r Eric Dittman dittman@dittman.nety= Check out the DEC Enthusiasts Club at http://www.dittman.net/    ------------------------------  % Date: Thu, 29 Aug 2002 17:40:49 -0400r; From: "Brian Tillman" <tillman_brian@notnoone.notnohow.com>n$ Subject: Re: How to configure TCP IP$ Message-ID: <3d6e9514$3@news.si.com>  $ >But If I do a show network, then itD >still say that tcpip not started. but I can telnet with no problem.  I That is because you have not used START/NETWORK TCP-IP to start UCX.  Seeh the OpenVMS docs.  --A Brian Tillman                   Internet: tillman_brian at si.comrA Smiths Aerospace                          tillman at swdev.si.comi= 3290 Patterson Ave. SE, MS      Addresses modified to prevente< Grand Rapids, MI 49512-1991     SPAM.  Replace "at" with "@"8        This opinion doesn't represent that of my company   ------------------------------    Date: 29 Aug 2002 13:08:28 -0700+ From: seanobanion@attbi.com (Sean O'Banion)y- Subject: Re: HP and Dell dump Microsoft Worksu= Message-ID: <f883d5a4.0208291208.23755c4d@posting.google.com>g  B There is no support for Wordperfect on Mac, though it can still be8 downloaded.  There have been no real changes since 1997.  B And I have to wonder about a quote of support from 1998: a lot has happened since then.  ? It does appear that the VMS version is activly being worked on,- though:-/ http://www.legacytech.net/Products/products.htm  Has anybody here used it?  e4 Can you use the mouse to get to the pull down menus?     Sean  e Alan Greig <a.greig@virgin.net> wrote in message news:<7a6smug8qhnav8b5pngh9s72run6nbnpj7@4ax.com>... B > On Thu, 29 Aug 2002 11:57:57 GMT, winston@SSRL.SLAC.STANFORD.EDU > wrote: >  > H > >Especially since most elements of Corel Office actually run on Linux. > * > And some of them on VMS for that matter:r > http://www3.corel.com/cgi-bin/gx.cgi/AppLogic+FTContentServer?pagename=Corel/PressRelease/Details&id=CC1ZYU9TI9C > H > "Support for Linux fits well with Corel's mission statement - to focusC > technology leadership to offer customers value, compatibility andtG > choice. Corel presently offers its award-winning Corel WordPerfect oneD > numerous operating systems including Linux, UNIX, Macintosh, DOS,, > Open VMS and various Windows platforms".  > 8 > Although a third party handles the VMS port for Corel. > G > Hmm, maybe HP could be persuaded to bundle WordPerfect with VMS. Botho? > the VT interface and X-Windows versions are available for VMSn >  >  > >> > >>Or not.  But one can hope. > >> > >k
 > >Indeed. > > 
 > >-- Alan   ------------------------------  # Date: Thu, 29 Aug 2002 21:07:15 GMTd1 From: "Terry C. Shannon" <terryshannon@attbi.com> - Subject: Re: HP and Dell dump Microsoft Workss. Message-ID: <72wb9.225282$me6.30158@sccrnsc01>  2 "Alan Greig" <a.greig@virgin.net> wrote in message2 news:mqrrmuk2bl6oau6l3crak8d3fulkangs9p@4ax.com... >.+ >  http://www.theinquirer.net/?article=5161  >y" > HP and Dell dump Microsoft Works > 9 > And who can blame them? Corel wins with big PC builders  >h  J And Microsoft Works would better be named "A Union OS that sometimes works when it's not on strike."e   ------------------------------  # Date: Thu, 29 Aug 2002 21:05:45 GMTc1 From: "Terry C. Shannon" <terryshannon@attbi.com>o7 Subject: Re: HP earnings - how did the VMS business do?e- Message-ID: <J0wb9.122634$aA.26674@sccrnsc02>u  / "John Nebel" <nebel@csdco.com> wrote in messageiA news:Pine.OSF.4.21.0208280859020.21298-100000@athena.csdco.com...e >aI > Does anyone know how the VMS part of HPs business is doing financially?i >!  L Not being an HP beancounter or director, hell no! But things look squatulentI in the BCS space, don't they? Not surprising, if for no other reason thanyA the fact that any HPQ competitor has an easy marketing job: visit8 comp.os.vms and cut and paste.  H That said, Sr VP/GM Brandling of the Australia/New Zealand branch of HPQL noted that Alpha sales in the two months post-merger were up 35 percent overL the corresponding period a year ago. (Haven't a clue about the period a year
 ago, though!)l   Cheers,u  & Terry "It's the MARKETING, Stupid!" S.   ------------------------------  # Date: Fri, 30 Aug 2002 02:50:26 GMT * From: "Bill Todd" <billtodd@metrocast.net>7 Subject: Re: HP earnings - how did the VMS business do?tC Message-ID: <S3Bb9.297398$2p2.12253406@bin4.nnrp.aus1.giganews.com>   < "Terry C. Shannon" <terryshannon@attbi.com> wrote in message' news:J0wb9.122634$aA.26674@sccrnsc02.... >p1 > "John Nebel" <nebel@csdco.com> wrote in message C > news:Pine.OSF.4.21.0208280859020.21298-100000@athena.csdco.com...5 > >iK > > Does anyone know how the VMS part of HPs business is doing financially?  > >  > C > Not being an HP beancounter or director, hell no! But things lookC
 squatulentK > in the BCS space, don't they? Not surprising, if for no other reason thantC > the fact that any HPQ competitor has an easy marketing job: visit   > comp.os.vms and cut and paste.  J You can thank Compaq for that - for the Alphacide, for the associated liesK and broken promises, and for that fact that, despite having had well over agI year now to make *some* kind of effort to try to repair things, it's donev absolute squat.n   >aJ > That said, Sr VP/GM Brandling of the Australia/New Zealand branch of HPQI > noted that Alpha sales in the two months post-merger were up 35 percent' overI > the corresponding period a year ago. (Haven't a clue about the period ah year > ago, though!)u  D That might depend rather strongly on whether the two-month period in? question included any significant amount of time after 6/25/01.s   - bill   ------------------------------  % Date: Thu, 29 Aug 2002 23:29:34 -0400y- From: JF Mezei <jfmezei.spamnot@videotron.ca>c7 Subject: Re: HP earnings - how did the VMS business do?A, Message-ID: <3D6EE672.CC3888E7@videotron.ca>   Bill Todd wrote:L > You can thank Compaq for that - for the Alphacide, for the associated liesM > and broken promises, and for that fact that, despite having had well over asK > year now to make *some* kind of effort to try to repair things, it's done. > absolute squat.k  M We can personally expect that the Alpha murder caused a great deal of damage.3F However, we don't know how much incentives Compaq gave to the few veryI important customers who generate most of the profits. It is possible thatrN Compaq has tructured incentives such that customers continue to buy Alphas andL get a break elsewhere. That would allow Compaq/HP to claim increase on Alpha sales etc etc.  L Lets face it, those with large Alpha investments (with VMS) won't be able toM migrate to something else overnighht and during that time, they will continuecK to use alpha and require additional hardware to handle growth requirements.m   ------------------------------  % Date: Fri, 30 Aug 2002 00:00:46 -0400u( From: David Froble <davef@tsoft-inc.com>7 Subject: Re: HP earnings - how did the VMS business do?t, Message-ID: <3D6EEDEE.5050104@tsoft-inc.com>   Bill Todd wrote:  > > "Terry C. Shannon" <terryshannon@attbi.com> wrote in message) > news:J0wb9.122634$aA.26674@sccrnsc02...e > 1 >>"John Nebel" <nebel@csdco.com> wrote in messageDC >>news:Pine.OSF.4.21.0208280859020.21298-100000@athena.csdco.com...l >>J >>>Does anyone know how the VMS part of HPs business is doing financially? >>>  >>> C >>Not being an HP beancounter or director, hell no! But things lookn >> > squatulent > K >>in the BCS space, don't they? Not surprising, if for no other reason thaniC >>the fact that any HPQ competitor has an easy marketing job: visitu  >>comp.os.vms and cut and paste. >> > L > You can thank Compaq for that - for the Alphacide, for the associated liesM > and broken promises, and for that fact that, despite having had well over aiK > year now to make *some* kind of effort to try to repair things, it's donee > absolute squat.v    L I've also got to jump in here with some comments.  My first question is, if M Compaq wasn't so worthy of the comments in c.o.v, then where would they come nO from?  Are people suppost to be lied to, mislead, and have commitments broken, i< and just smile and say "thanks for repairing my hemmoroids"?  M Ok, it's pretty clear that when customers complain, that competitors benefit ,H from showing the customer dissatisfaction.  But who's job is it to keep N customers happy?  The customers, or the vendor?  I see just a bit too much of P "shoot the messenger" type of thinking here.  I'm not happy about any harm done P to VMS, and Alpha.  Even the harm of unhappy customer comments.  But, isn't the Q real problem that customers are unhappy, and not that they show that unhappiness?s   Dave   ------------------------------  # Date: Thu, 29 Aug 2002 20:55:59 GMT 1 From: "Terry C. Shannon" <terryshannon@attbi.com> + Subject: Re: HP-Compaq Merger Went Smoothlyd- Message-ID: <zTvb9.122566$aA.27353@sccrnsc02>   # "Andrew Harrison SUNUK Consultancy"e> <andrew_nospam.harrison_remove_this@sun#.com> wrote in message* news:aki7du$ljg$1@new-usenet.uk.sun.com... >  >S > Terry C. Shannon wrote:s >cK > > Well, seems to me that HPQ pretty much made its numbers, and it sure ast hellH > > doesn't look as bad as Sun (which might be a bargain at four bucks). > >  >d >e > How do you work that out.l   Duh, I look at my portfolio.   ------------------------------  # Date: Thu, 29 Aug 2002 21:00:17 GMT-1 From: "Terry C. Shannon" <terryshannon@attbi.com>r+ Subject: Re: HP-Compaq Merger Went Smoothlye. Message-ID: <BXvb9.225205$me6.30295@sccrnsc01>  : "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message& news:3D6DA1EB.138EF134@videotron.ca... > "Terry C. Shannon" wrote:o
 > > Sun $4.00y > >t > > IBM $78.00 > >I > > Enuff said >OB > Actually, Sun closed below $4.00 today, and Microsoft below $50. >nK > Sun's stock may be battered, but it still has an image of a leader in then unix > server market.  L Indeed it do! Carly and Curly could use a few lessons in marketing. Starting with Marketing 101.o   ------------------------------  # Date: Thu, 29 Aug 2002 19:21:44 GMTr# From: "John Smith" <a@nonymous.com>hC Subject: HP-UX and Tru64 - Side by Side Comparison for Cluster & DTXH Message-ID: <cvub9.21263$GK2.20416@news02.bloor.is.net.cable.rogers.com>    Announcing an HP Customer Event:8 HP-UX(tm) and Tru64(tm) UNIX  A Side by Side Comparison#   Clustering and Disaster Tolerancel  ? Presenter: Jan Mark Holzer, HP Senior Member of Technical Staff.  H Participants: This session is a limited registration event for customers only.t  - When: Thursday, Sept 5th at 10 -11:30am ET ORi%  Thursday, Sept 5th at 7  8:30 pm ET-   10:00 AM ET is'   8:00 AM in Colorado Springs, Coloradom   9:00 AM in Houston, Texasc   3:00 PM in London, England   4:00 PM in Munich, Germany  
 7:00 PM ET is0$   7:00 AM in Singapore and Hong Kong*   8:00 AM in Tokyo, Japan and Seoul, Korea!   9:00 AM in Melbourne, Australia #   11:00 AM in Auckland, New Zealandw"   4:00 PM in Cupertino, California    I Delivery: This session will be delivered via the web using the HP Virtualo= Classroom together with audio teleconferencing via the phone.eI Registration: To register for one of these sessions and learn more, visitEJ <http://www.hpbroadband.com> and enter your email address and the key word "tru64unix".G Recorded Playback: This webcast will be recorded and made available foraL playback subsequent to the live events.  Information on the availability andK access to the recording will be communicated after the live event has takena place.   Session Overview:(L No operating system is more critical to enterprise customers than UNIX, withK its ability to deliver robust, mission-critical solutions.  With the mergereK of HP and Compaq, HP is now the largest UNIX vendor in the world, with morenG than two million HP-UX and Tru64 UNIX installations worldwide. You, our C valued Tru64 UNIX customers, have expressed some concerns about thebK differences between these operating systems, and how those differences will2 affect your operational staffs.   J As HP moves to consolidate its Enterprise UNIX offerings around HP-UX withJ Tru64 UNIX features on Itanium(tm), it is important for you to be aware ofH the technical differences and similarities between Tru64 UNIX and HP-UX.F This webcast will focus on clustering and disaster tolerance.  It willI explore and compare the technical architecture and features of both Tru64gL UNIX TruCluster Server software and HP-UX MC/ServiceGuard for clustering andK disaster tolerance.  TruCluster Server technology will be incorporated intoaK a future version of HP-UX.  This talk is intended to increase understandingdB of the technical and architectural differences of the current highJ availability products available on HP-UX and Tru64 UNIX. As the TruClusterL Server technology is incorporated into HP-UX, this session will also provideF information on how existing installations can be upgrade to the future product offering.x    1 Session Agenda (subject to change) 1 hr in length  - Release historiesi1 - Design decisions and principles of each productV3     o Shared nothing and shared everything approache'     o Base operating system integrationI! - Product portfolio and solutionsa"     o High availability clustering     o Campus cluster solutions/     o Disaster tolerance features and solutionsi+ - Detailed comparison of cluster subsystemse-     o Hardware requirements and configurationg,     o I/O infrastructures and storage access)     o File system access and availabilitye*     o Cluster networking and client accessD     o Application integration in a MC/ServiceGuard/TruCluster Server environmenta'         - SG packages and CAA resourcesh     o Cluster internal servicesp*         - Membership and quorum management"         - Distributed Lock manager%         - Inter-cluster communications1 - TruCluster technology moving forward into HP-UX-.     o Features planned for TruCluster on HP-UXI - The transition from MC/ServiceGuard today to TruCluster/Serviceguard in  HP-UXg  4 Open questions and answers (15-30 minutes in length)   ------------------------------  # Date: Thu, 29 Aug 2002 18:17:06 GMT> From: danco@ns2.pebble.org7 Subject: Re: Is the HP/Compaq merger really going well?v1 Message-ID: <slrnamsqpa.e24.danco@ns2.pebble.org>a  F In article <oH+zOclRqOUn@eisner.encompasserve.org>, Bob Koehler wrote:p > In article <343f30ae.0208281118.774e09d7@posting.google.com>, SPAMSINK2001@YAHOO.COM (Alan E. Feldman) writes: >> pI >> And remember, CLAIMING NOON AND MIDNIGHT ARE EACH NEITHER AM NOR PM IST >> RIDICULOUS! THANK YOU.@ > F >    In that case I will assert the claim, especially since it's true. >    Often seen:u   "12:00 AM" and "12:00 Midnight"l  ) "12:00 Noon", "12:00 Meridian", "12:00 M"   8 (I saw that last one, "12:00 M" for 12 Noon, on RSTS/E.)   - Dani   ------------------------------  # Date: Thu, 29 Aug 2002 18:23:09 GMTh From: danco@ns2.pebble.org7 Subject: Re: Is the HP/Compaq merger really going well?>1 Message-ID: <slrnamsr4l.e24.danco@ns2.pebble.org>   K In article <BA52530E3149734A9BAABDBBFA808E49028BEA4F@rlghncst964.usps.gov>,a" Webb, William W Raleigh, NC wrote: > D > AM = Ante Meridian, which starts at 12:01 AM and ends at 11:59 AM C > PM = Post Meridian, which starts at 12:01 PM and ends at 11:59 PMt   I've also seen:c  @ M = Meridian, which starts at 12:00:00 M and ends at 12:00:59 M.  B (Saw the above on RSTS/E back in the early '80s.  At exactly noon,+ RSTS/E returned the time string "12:00 M".)a  M I wonder if the same people that use "M" for Meridian also use "N" for Nadir?e   - DanO   ------------------------------    Date: 29 Aug 2002 13:50:47 -0600- From: koehler@encompasserve.org (Bob Koehler)a7 Subject: Re: Is the HP/Compaq merger really going well?t3 Message-ID: <y16RgPV3i52p@eisner.encompasserve.org>   \ In article <3D6E3B18.986A4A94@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes:  G > Might as welll all live on GMT, or better yet, on VMS quadword time.  K > "Lets set the meeting at 01C73608 39A7B1F0"  There would be no confusion e > there.  G   When I log in I get the stardate, thanks to some little program I gotg@   off the web.  Isn't that now a univerally recognized standard?  "   OBTW, it's now stardate 37495.72   ------------------------------  % Date: Thu, 29 Aug 2002 17:23:06 -0400   From: John Santos <JOHN@egh.com>7 Subject: Re: Is the HP/Compaq merger really going well?-5 Message-ID: <1020829171648.1685A-100000@Ives.egh.com>B  / On Thu, 29 Aug 2002 danco@ns2.pebble.org wrote:n  M > In article <BA52530E3149734A9BAABDBBFA808E49028BEA4F@rlghncst964.usps.gov>,B$ > Webb, William W Raleigh, NC wrote: > > F > > AM = Ante Meridian, which starts at 12:01 AM and ends at 11:59 AM E > > PM = Post Meridian, which starts at 12:01 PM and ends at 11:59 PM  >  > I've also seen:a > B > M = Meridian, which starts at 12:00:00 M and ends at 12:00:59 M. > D > (Saw the above on RSTS/E back in the early '80s.  At exactly noon,- > RSTS/E returned the time string "12:00 M".)e > O > I wonder if the same people that use "M" for Meridian also use "N" for Nadir?o >  > - Dann  B RSTS/E used to use 12:00 PM for midnight and 12:00 M for noon, andA claimed (maybe in an article in a Software Dispatch), that it was B in accord with some international standard, but they changed it to: 12:00 AM for midnight and 12:00 PM for noon at about V9.0.   -- o John Santos  Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  # Date: Thu, 29 Aug 2002 23:06:21 GMTe1 From: "Terry C. Shannon" <terryshannon@attbi.com>eD Subject: Re: Is the HP/Compaq merger really going well?  YOU BET!!!!- Message-ID: <NNxb9.123520$aA.28031@sccrnsc02>p  @ Of course it's going well... czech out these Secret New Products  7 http://www.tru64.org/stories.php?story=02/08/15/4152136i  L The Marvel Mystery Oil has significantly reduced latency in my 88 BMW 325ix,G and the Hewlett Creme is good for darn near anything that ails you. ;-}   @ With stuff like this, who needs toner cartridges and peecees????   ------------------------------  % Date: Thu, 29 Aug 2002 11:44:14 -0700t. From: JMK <jeffrey.m.klopotic@lmco.nospam.com> Subject: Lint tool for HPQ CXX/ Message-ID: <3D6E6B7E.5803413F@lmco.nospam.com>r  D I'm tasked with porting a successful GCC program to DEC VMS CXX.  OnH CXX, it compiles and runs with no warnings or errors, although output is not as expected.  D Forensic analysis of the output indicates failure to convert stringsA into file names and other data identifiers necessary for internals workings of the program.    E Are there any tools which I could pass the source through which couldlC indicate possible problems?  As before, this passes compilation anda# linking with no warnings or errors.l   Any help appreciated.   
 Jeff Klopotico   ------------------------------    Date: 29 Aug 2002 14:47:22 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen):" Subject: Re: Lint tool for HPQ CXX3 Message-ID: <XJ8Wu6+IjIeQ@eisner.encompasserve.org>d  ` In article <3D6E6B7E.5803413F@lmco.nospam.com>, JMK <jeffrey.m.klopotic@lmco.nospam.com> writes:F > I'm tasked with porting a successful GCC program to DEC VMS CXX.  OnJ > CXX, it compiles and runs with no warnings or errors, although output is > not as expected. > F > Forensic analysis of the output indicates failure to convert stringsC > into file names and other data identifiers necessary for internalh > workings of the program.   > G > Are there any tools which I could pass the source through which couldTE > indicate possible problems?  As before, this passes compilation andh% > linking with no warnings or errors.D  H I have done such a project.  If you make it through the compiler withoutG any warning messages (and without disabling any warning messages :-), Ic0 think you are going to have to use the debugger.  G Beware if your program makes use of the "portable" RTL routine "dlsym".dH It isn't.   I presume there are others, and the debugger is how you find them.u   ------------------------------  % Date: Thu, 29 Aug 2002 13:47:35 -0700s. From: JMK <jeffrey.m.klopotic@lmco.nospam.com>" Subject: Re: Lint tool for HPQ CXX/ Message-ID: <3D6E8867.36772949@lmco.nospam.com>f   Larry, Thanks for the response.  H I'll check with warnings that may have been disabled by others elsewhere5 (we are a larger organization) -- thanks for the tip.t  A I'm using the debugger and finding that strtok has some different-1 responses cxx vs gnu, and they are less friendly.2  4 But aren't there any format/syntax checkers for CXX?   Jeff   ------------------------------    Date: 29 Aug 2002 16:26:30 -0600- From: koehler@encompasserve.org (Bob Koehler)c" Subject: Re: Lint tool for HPQ CXX3 Message-ID: <YBvuoeDwJUTi@eisner.encompasserve.org>A  ` In article <3D6E6B7E.5803413F@lmco.nospam.com>, JMK <jeffrey.m.klopotic@lmco.nospam.com> writes:F > I'm tasked with porting a successful GCC program to DEC VMS CXX.  OnJ > CXX, it compiles and runs with no warnings or errors, although output is > not as expected.      Successfull on what OS?   ------------------------------  # Date: Thu, 29 Aug 2002 21:40:01 GMT * From: "Paul A. Jacobi" <nospan@nospam.com>" Subject: Re: Lint tool for HPQ CXX1 Message-ID: <Rwwb9.44$pr.507988@news.cpqcorp.net>.  C Compilers on VMS include LINT-type functionality (unless disabled).e  C Unix is built around pipes and uses multiple images (compiler, lint C assembler) to accomplish the same functionality provided by the VMS 	 compiler.s  J In fact the VMS implementation usually does much better compile time error checking than Unix compilers.   K Unix code also tends to depend on uninitialized varibles being set to zero, ! which is always a bad assumption.g     Paul A. Jacobi Hewlett Packard Company ! OpenVMS Systems Group, ZKO3-4/U14n 110 Spitbrook Road Nashua, NH 03062-2698p$ Email: Paul dot Jacobi at hp dot com   ------------------------------    Date: 29 Aug 2002 17:03:51 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen)0" Subject: Re: Lint tool for HPQ CXX3 Message-ID: <cX63f1bIJQLx@eisner.encompasserve.org>t  ` In article <3D6E8867.36772949@lmco.nospam.com>, JMK <jeffrey.m.klopotic@lmco.nospam.com> writes:  J > I'll check with warnings that may have been disabled by others elsewhere  6 > But aren't there any format/syntax checkers for CXX?  @ There is a strong syntax checker, so some site disable warnings.   ------------------------------  % Date: Thu, 29 Aug 2002 22:47:26 -0000  From: sword7@speakeasy.org& Subject: Re: Old games for VAX system./ Message-ID: <umt93uae7ptsb5@corp.supernews.com>   + James Cameron <james.cameron@hp.com> wrote:oI > How does the TS10 compare, and where are the installation instructions?m  D Look into ka655.cfg for more information.  Read 'ReadMe' file in vaxB subdirectory have some instructions to how to install.  I still am= developing some instructions for other things.  On 1.9 Ghz P4o@ system, it is estimated VUPS 5.4.  Without -DDEBUG option, it isA estimated VUPS 7.4.  I only have vups.com that was posted on thisd
 newsgroup.   -- Tim Stark   -- -, Timothy Stark	<><	Inet: sword7@speakeasy.orgJ --------------------------------------------------------------------------F "For God so loved the world, that he gave his only begotten Son, that H whosoever believeth in him should not perish, but have everlasting life.. Amen." -- John 3:16 (King James Version Bible)   ------------------------------  % Date: Thu, 29 Aug 2002 22:54:54 -0000s From: sword7@speakeasy.org& Subject: Re: Old games for VAX system./ Message-ID: <umt9hutjfb7c98@corp.supernews.com>o  # Mark Hittinger <bugs@pu.net> wrote:>6 > "Zane H. Healy" <healyzh@shell1.aracnet.com> writes: >>  Empire 4.0    - Native   Me too.t  ) > I may have empire 5.0 around someplace.p  5 I remmeber that some time ago but do not have it yet.   K > So the question is begged - can Tim emulate the graphics subsystem of thetM > vaxstations :-) and the mouse :-)  Its probably not as important as gettingaK > the networking running etc. :-)  I'm only kidding Tim please don't chargee > off and write it tonight :-)  H Yes, I can emulate graphics (VCB02) for VAXstations but have not startedK development yet.  DELQA now works so well.  I am finishing DELQA emulation ,( soon.  No, TS10 emulator always is free.   -- Tim Stark   -- e, Timothy Stark	<><	Inet: sword7@speakeasy.orgJ --------------------------------------------------------------------------F "For God so loved the world, that he gave his only begotten Son, that H whosoever believeth in him should not perish, but have everlasting life.. Amen." -- John 3:16 (King James Version Bible)   ------------------------------  # Date: Fri, 30 Aug 2002 00:46:06 GMT # From: "John Smith" <a@nonymous.com>o) Subject: OT - re: From the Peanut GalleryaJ Message-ID: <ifzb9.133485$8aG1.69323@news01.bloor.is.net.cable.rogers.com>  < "Terry C. Shannon" <terryshannon@attbi.com> wrote in message( news:ifxb9.225968$me6.30410@sccrnsc01... >  >-L > Nope, I am a citizen of America (the greatest country in the world, by the > way) .....    ! Just pulling your chain a bit....:  H Which 'America' would that be...North America, Central America, or South America?  G A citizen of 'America' could be any resident from any of those regions.#  K More correctly, (in a fully qualified form) you are a citizen of the UnitedeJ States of America, which is merely one of many countries in 'America', andC one of several countries in North America. Geopolitically speaking,hJ Canadians and Mexicans have as much claim on the term 'American' as anyone  in the United States of America.  F Much as a 'European' is a resident of any one of a number of countriesG located in Europe (I just know I'm going to get some mail over this one- too).M  K As to the part of your statement "the greatest country in the world, by theeK way" -  don't let your hubris carry you away.....this is an ephemeral statedI of affairs as history has shown time and again...... Egypt, Greece, Rome,7H China, Macedonia (remember Alexander the Great?), Ottoman Empire, Spain,G England, the United States of America, each a 'superpower' in their own J right, with their own contributions to the advancement of world, each withE their own form of terror/enslavement/dominance, and their own form ofe 'Manifest Destiny'.-  I Who's next? Perhaps China yet again in the next 20-50 years. Look at pastmK history - most of the above noted countries were totalitarian or autocraticr@ regimes. Arguably Greece, England and the Untied States are/wereH quasi-democratic at the time, perhaps Greece more so than England or theH Untied States in pure political science terms. But think about what madeI these countries great...running an economic surplus and producing much of L what the world wanted. Today the United States is the world's largest debtorD and China and southeast Asia is making much of what the world wants.E England's decline wasn't so much the result of the rise of the Unitedg2 States, but more an accident of geography in 1914.  K  Remember the immortal words of Casey Stengle - "It's like deja vu all over  again."o     :-)n   ------------------------------  % Date: Thu, 29 Aug 2002 23:26:07 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca>w* Subject: PRODUCT (PCSI) and TCPIP services, Message-ID: <3D6EE5A4.859BCF0B@videotron.ca>  ? Ok, trying to install the latest patches for TCPIP services....r   PRODUCT SHOW  reveals I have:i  . DEC VAXVMS TCPIP V5.0-9   (Full LP, Installed)  # But when I try to PRODUCT INSTALL  .4 	DEC VAXVMS TCPIP_ECO V5.0-113  I get the following:  N %PCSI-E-APPLYTOERR, product DEC VAXVMS TCPIP, to which maintenance product DEC1 VAXVMS TCPIP_ECO V5.0-111 applies, was not found.a  D Terminating is strongly suggested, Do you want to terminate ? [YES].  M If I say "no", it seems to proceed but then does nothing. That kit seems verys- big, so it should be shuffling lots of files.a  M I am aware of the PCSI bug, so I downloaded the patch that seemed available: e 	DEC VAXVMS VMS72_PCSI V1.1P  ' But that made absolutely no difference.d  - The original TCPIP comes from the hobbyst CD.I  N The horrible HTML  README does mention that it expects TCPIP V5.0A. However, IM find no way to patch from 5.0-9 to 5.0A from the ftp.service.digital.com.  Ist such an upgrade possible ?  M I realise that this is very old, but since I do not have access to commercialfL VMS media anymore, I am stuck wiuth the hobbyist CD and what is available on
 the ftp site.w   ------------------------------  # Date: Thu, 29 Aug 2002 20:14:09 GMT + From: Jeff Campbell <jcampbell@ins-msi.com>  Subject: Re: re white boxeni+ Message-ID: <3D6E7549.17F0FD63@ins-msi.com>c   Thomas Harris wrote: > @ > In article <akjdkr$nq1$3@web1.cup.hp.com>, Hoff Hoffman wrote: > >so > > In article <slrnamp38r.44d.$USER@dinsdale.piranhabrothers.foo>, Thomas Harris <tharris@stewart.com> writes:U > >h; > >:I now am in the position of admining a slightly earlierc; > >:version alphaserver in VMS world,  which I'm reasonablyi> > >:familiar with,  but would like to use home alphaserver for> > >:playing in sandbox purposes,  rather than using production= > >:machinery.  Current alpha-montagnar hobbyist kit installs A > >:but pukes midway through boot.  Any help appreciated.  I haveu< > >:a 3100 vaxstation but builds would be a tad quicker on a > >:500mhz alpha.i > >eJ > >   Specific hardware details, please?  Specific details include but areK > >   not limited to SCSI and/or IDE details, graphics display, OpenVMS and I > >   DECwindows versions, and -- particularly since you don't want me toaH > >   hand you a "convenience bag" nor attempt to cannulate for an IV --L > >   details of exactly what you mean by "pukes", as that is not an OpenVMS% > >   error message that I recognize.s > > D > Yes I know all that.  I already went through this on it with Tru64G > which I was able to make run by editing the NVRAM to enable SRM Boot.i > % > DigitalServer 3000 Model 3305 6500Ao > SRM-Console V5.8-16n > ARC Console V5.70  > OpenVMS PALcode V1.21-3  > Digital UNIX PALcode V1.23-5 > Serial ROM V1.3  > 21164A-2 500mhz  > Motherboard 54-24803-02 Rev An > 256M in bank 0 > QLOGIC ISP 10x0  > 3 Quantum Viking drivesl
 > RRD46 DCROM  > S3TRio64/TRio32 videoS > Intel 82375 bridge > DE500 nicv > 
 > OpenVMS 7.2g > DECWindows V1.2-5d
 > DECnet V7.2  > TCP/IP v5.09 > > > I have a ps2 keyboard and environment value is set to PCXAL.; > One of the problems is even as it succesfully installs itC8 > sees "?" as "/".  I then succesfully complete install,  > reboot,  and it fails to load. > N >         %APB-F-BADSYSROOT System root (SYSA) does not exist, check bootflags   The SYSA is the clue. If you:g      >>> show b*  ( I'll bet boot_osflags is 'a' (or 'A' 8-)  7 APB is trying to access SYS$SYSDEVICE:[SYSA] instead ofHF SYS$SYSDEVICE:[SYS0] (where SYS$SYSDEVICE is your boot device). [SYS0]: is where the VMS installation kit will install by default.   To fix this you need to:      >>> set boot_osflags 0,00    >>> init0   >  >         Halted CPU 0 >         Halt code =5# >         Halt Instruction Executedi >         PC=200039bc-" >         Warning HWRPB is invalid > I > When I do the install I let it fully initialize the disk for VMS usage.r > M > So obviously I need to set some other bootflags or tinker with nvram a bit.c > Any ideas. > 	 > Thanks,A > Tom2    
 Jeff Campbellu n8wxs@arrl.net   ------------------------------  # Date: Thu, 29 Aug 2002 21:04:40 GMT ) From: Thomas Harris <tharris@stewart.com>  Subject: Re: re white boxen0? Message-ID: <slrnamt6ju.fl2.$USER@dinsdale.piranhabrothers.foo>h  @ In article <3D6E7549.17F0FD63@ins-msi.com>, Jeff Campbell wrote: > Thomas Harris wrote: >> 0 > The SYSA is the clue. If you:r >  >    >>> show b* > * > I'll bet boot_osflags is 'a' (or 'A' 8-) > 9 > APB is trying to access SYS$SYSDEVICE:[SYSA] instead of-H > SYS$SYSDEVICE:[SYS0] (where SYS$SYSDEVICE is your boot device). [SYS0]< > is where the VMS installation kit will install by default. >  > To fix this you need to: >  >    >>> set boot_osflags 0,0c
 >    >>> init  >  >> t >>         Halted CPU 0i >>         Halt code =5m$ >>         Halt Instruction Executed >>         PC=200039bc# >>         Warning HWRPB is invalido >> m> Sorted.  Thanks everyone.  Now to get on with breaking things.   Cheers,u Tome   ------------------------------  % Date: Thu, 29 Aug 2002 15:12:37 -0400g> From: "Michael Raspuzzi" <raspuzzi#spamblaster.emesen-dot-com> Subject: Re: Reverse LAT+ Message-ID: <#Vkun$4TCHA.1972@cpimsnntpa03>-  - "Hans Vlems" <hvlems@iae.nl> wrote in message00 news:akjck2$1irrkb$1@ID-143435.news.dfncis.de... > [...]  > snip > [...]r > > L > > The DECserver 100 can only act as a master.  Since it cannot function as a9K > > LAT slave, it does not have reverse LAT capability.  When SET HOST /LATD > wasEI > > introduced to OpenVMS, the project was called "LAT/Master" because we E > > essentially implemented the master portion of the LAT protocol inA addition! > > to the slave part on OpenVMS.2 > L > Thanks for that little bit of history. That must have happened well before+ > LAT development was contracted out, then?  >   L We did the LAT/Master development back in 1989 and shipped it as an extra inL VMS V5.4-1.  It was fully integrated in V5.5 and higher.  I believe that theJ only activity in the world of LAT and DECservers in the last few years hasH been sustaining engineering.  This work has been done by other companiesK outside of what was formerly DEC.  The last work I did with LAT was back inaJ 1995.  Support for multiple LAN adapters on OpenVMS was big and I had thatL working and integrated in some form of OpenVMS V7.  I do not know if supportL for that feature (failover to multiple adapters) ever made it to the outsideK world.  I did the leg work for the LAT V5.3 Architecture as well but again,II do not know if the licensees ever saw this.  Guess my age is showing ;-).1   Mike Former OpenVMS LAT developer   ------------------------------  % Date: Thu, 29 Aug 2002 18:35:16 -04004; From: "Brian Tillman" <tillman_brian@notnoone.notnohow.com>y Subject: Re: Reverse LAT$ Message-ID: <3d6ea1d7$1@news.si.com>  H >The DECserver 100 can NOT do reverse LAT.  It is a common misconception thatJ >being able to put a printer on a DECserver and accessing it from the host is% >the same as reverse LAT.  It is not.4  L Just noticed that.  No SET/DEFINE SERVICE commands on the DS100.  Yes, I was/ confising the two.  Thanks for the explanation.o --A Brian Tillman                   Internet: tillman_brian at si.comMA Smiths Aerospace                          tillman at swdev.si.coml= 3290 Patterson Ave. SE, MS      Addresses modified to prevento< Grand Rapids, MI 49512-1991     SPAM.  Replace "at" with "@"8        This opinion doesn't represent that of my company   ------------------------------  % Date: Thu, 29 Aug 2002 14:16:38 -0400o- From: JF Mezei <jfmezei.spamnot@videotron.ca>A+ Subject: Re: silliest thing heard this weekn, Message-ID: <3D6E64F9.454ED188@videotron.ca>   Rob Young wrote:M >         Typically, what I do is never purge when I am working on something.     H I concur to some extent. I will sometimes purge just after having testedM something with good results. Then i continue to accumulate versions until theBL next cycle. And at the end, once I know that the code is good, then I freezeM it, purge everything and that directory becomes an archives for that version.,    N I've had to mainain some software where customers were are different versions,J and I found this to be very convenient since you had 2 different directory structures to work with.   ------------------------------    Date: 29 Aug 2002 13:03:33 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen)l+ Subject: Re: silliest thing heard this weeki3 Message-ID: <gdN53M8mfir9@eisner.encompasserve.org>l  l In article <89cc5656.0208290848.607f6d3a@posting.google.com>, jlewocz@poczta.wprost.pl (John Lewocz) writes:  B > The other unix boy and myself on our team were trying to get our
 > development F > effort off on the right foot by trying to implement a source control > and builduC > procedure.   We of course campaigned for management to get us CMSrA > (which I'd used before), but they balked at the cost.  (We were-> > stunned too, since we're used to tools with roughly the sameE > functionality -- minus gui -- for free).  THe Vms boys in the groupf7 > not only did not help, but some rather looked down onmD > the entire concept of source control (apparently only people using > inferior OS'siG > need such a thing, and besides, it's just so cumbersome).   Also, thet  K Did I mention many shops have no qualification exam to be able to use VMS ?   F > Why can't DEC(or COMPAQ, or HP) add a lexical like f$dir_exists() toH > the language?   Furthermore, why is seem ok with VMSers?   It's almost  @ F$PARSE() _is_ "does directory exist" (well, combined with "does? device exist", but for your purposes I presume that presents no 	 problem).T  4 It is too bad your VMS folk did not understand that.  C If you are ever in that position again, just pop the question here.:E People waste enormous amounts of time trying to do simple things moren# elegantly than the previous poster.   F > VMSers all grieve when some VMS system gets replaced with pcs.   ButE > they never see their arrogance as a factor in its demise.   If some0G > third-shift operator whacks the wrong files because he's in the wrong:H > directory because his prompt is a meaningless 'MFGSYS> ' then everyone  C In a production shop, an operator should not have the privileges toe whack the files.   ------------------------------   Date: 29 AUG 2002 21:16:00 GMT+ From: Dave Greenwood <greenwoodde@ornl.gov>b+ Subject: Re: silliest thing heard this weekl2 Message-ID: <29AUG02.21160001@feda34.fed.ornl.gov>  D In a previous article, jlewocz@poczta.wprost.pl (John Lewocz) wrote: [snip]B > The other unix boy and myself on our team were trying to get our
 > development F > effort off on the right foot by trying to implement a source control > and buildoC > procedure.   We of course campaigned for management to get us CMS-A > (which I'd used before), but they balked at the cost.  (We werer> > stunned too, since we're used to tools with roughly the sameE > functionality -- minus gui -- for free).  THe Vms boys in the group 7 > not only did not help, but some rather looked down onfD > the entire concept of source control (apparently only people using > inferior OS'seG > need such a thing, and besides, it's just so cumbersome).   Also, theeG > claim was made that they knew ahead of time how much it cost and thatwG > management would shoot it down.   However, when we tried to write DCL7 > scripts to implement somekE > primitive source control us two unix weenies were on our own.   The7
 > VMS wizardsrG > weren't interested in helping.   So I guess it was never really aboute
 > the $$$. [snip]  D I'm a long-time VMS user (and full-fledged bigot) but I think sourceA code control is a must.  I've used CMS for years and  VMS versionsF numbers are *not* a substitute.  In *your* case, however, I'd use CVS.G Put the server on one of your unix boxes and get the VMS client.  ThereoA are some things I prefer about CMS over CVS but I've used CVS for G several projects.  You can download the VMS client for CVS from Richardr Byer's site:  3   http://www.ourservers.net/public/vms/programming/-   Dave --------------9 Dave Greenwood                Email: Greenwoodde@ORNL.GOViH Oak Ridge National Lab        %STD-W-DISCLAIMER, I only speak for myself   ------------------------------  % Date: Thu, 29 Aug 2002 15:02:10 -0700 . From: JMK <jeffrey.m.klopotic@lmco.nospam.com>+ Subject: Re: silliest thing heard this weekk/ Message-ID: <3D6E99E2.CAA6611C@lmco.nospam.com>    Carl,D   Let's be real.  D You know you are asking for problems when you use reserved terms forC other than intended uses.  This applies not only to DCL, but to alli- other kinds of things in all walks of life.  d  2 It can be done, but the result is usually trouble.   Jeff   ------------------------------    Date: 29 Aug 2002 18:28:23 -0700. From: spamsink2001@yahoo.com (Alan E. Feldman)+ Subject: Re: silliest thing heard this weekt= Message-ID: <b096a4ee.0208291728.663c5f8e@posting.google.com>l  q jlewocz@poczta.wprost.pl (John Lewocz) wrote in message news:<89cc5656.0208290848.607f6d3a@posting.google.com>...t [...]  (snippiness snipped :-) > ! > Here's what I was trying to do:L > B > The other unix boy and myself on our team were trying to get our
 > developmenthF > effort off on the right foot by trying to implement a source control > and build-C > procedure.   We of course campaigned for management to get us CMSvA > (which I'd used before), but they balked at the cost.  (We wereo> > stunned too, since we're used to tools with roughly the sameE > functionality -- minus gui -- for free).  THe Vms boys in the group 7 > not only did not help, but some rather looked down onmD > the entire concept of source control (apparently only people using > inferior OS'sw; > need such a thing, and besides, it's just so cumbersome).r    E If your management doesn't see much of a future for your VMS systems,.; then that's why they don't want to spend any money on them.o        Also, theG > claim was made that they knew ahead of time how much it cost and thatAG > management would shoot it down.   However, when we tried to write DCL? > scripts to implement someuE > primitive source control us two unix weenies were on our own.   The 
 > VMS wizards G > weren't interested in helping.   So I guess it was never really abouty
 > the $$$.    / Well, I'd like to hear their side of the story!s    H > Anyway, I took the stab at writing the scripts, and my intent was that > al< > user would set a logical like SC$LIBRARY to something likeA > [user.test.safe].   Then, hopefully, have created it as well.  e
 > However, if B > the user forgot either one, or both these steps (or fat-fingeredB > something), then the script would try to fix things as gently as > possible.   The last thingH > we needed was to have people complaining "see, told you source control > was cumbersome!".  > F > Regarding "official" source repositories: we could ensure that theseG > exist ahead of time.   We just wanted people on board and to see thatt# > this takes seconds of their time.i    # You wanted *who* on board for what?i    # > So we ended up with the following  > ! > $   DEFINE/NOLOG SYS$OUTPUT NL:e  > $   DEFINE/NOLOG SYS$ERROR NL: > $M( > $   CREATE/DIR 'f$trnlnm("SC$LIBRARY") > $ ? > $   IF f$locate("CREATE-I-EXISTS", f$message($STATUS)) .EQ. 1o
 > $   THEN, > $       DEFINE/NOLOG SYS$OUTPUT 'MY_OUTPUT* > $       DEFINE/NOLOG SYS$ERROR 'MY_ERRORH > $       WRITE SYS$OUTPUT "Library directory: ", f$trnlnm("SC$LIBRARY")
 > $   ELSE, > $       DEFINE/NOLOG SYS$OUTPUT 'MY_OUTPUT* > $       DEFINE/NOLOG SYS$ERROR 'MY_ERRORB > $       WRITE SYS$OUTPUT "Directory ", f$trnlnm("SC$LIBRARY"), " > couldn't be created" > $       EXIT > $   ENDIFi > $     E Why are you supressing error messages? What are the symbols MY_OUTPUT F and MY_ERROR for? And you didn't reset SYS$OUTPUT and SYS$ERROR at theC end of the code. Is this code for programmers to write code on VMS, 2 and they need help with creating directories? Huh?    = > It works.  But it took a ridiculously long amount of time. -    E I assume you meant a long time to write it. You are apparently tryingeE to write primitive AI code. What I mean is, something that is trivial @ to tell a person to do can be extremely hard to program. ImagineF writing code to tell a robot to go to the middle table across the roomF and bring you the book that is on the table. It is ridiculously eaiserF to ask someone to do it or just do it yourself. You are trying to holdB users by the hand. Why not just do it for them in the first place?B Also, if you are new to VMS, that will increase the time you spend writing DCL code.n      Is this! > the best way?   I don't know.        No.     &  Even after we did this I asked our 15 > yr VMS veteranF > if there was a better way and he screwed around for about 10 minutesE > with this problem before he came up with something equally ugly andl > almost as long.M    B Well, 10 minutes doesn't sound that long to write a program. Can't. judge if it was ugly since you didn't post it.    3    And we don't even know if that would have workedA > because we didn't test it.    2 Well, since you didn't test it, what's your point?  @ From all your posts, this is possibly what you want, or at least close:  E $    ON WARNING THEN EXIT          ! I prefer this to the DCL defaulti $    WSO := WRITE SYS$OUTPUT; $    SET NOON                      ! Disable error checking-4 $    CREATE/DIRECTORY SC$LIBRARY   ! Create the dir.B $    STATUS = $STATUS              ! Capture the completion status: $    SET ON                        ! Resume error checking0 $    IF (.NOT.STATUS)              ! Test status	 $    THEN ? $        WSO "Directory ",F$TRNLNM("SC$LIBRARY")," could not bel	 created."o
 $        EXITm
 $    ENDIFD $    DIR_EMPTY = F$SEARCH("SC$LIBRARY:*.*").EQS.""  ! Test if dir is emptya $    IF (.NOT.DIR_EMPTY) THEN -lF      WSO "WARNING: Directory ",F$TRNLNM("SC$LIBRARY")," is not empty."	 $    EXITt    ,    Even judging from the responses that I'veF > received every VMS guru has their own circuitous way of doing things > and they're all different.      B And in Unix 10 programmers would have come up with unvarying code?    +   Someone responded that I should have readcF > posts.  I was looking in this group before I began this endeavor andG > this is where I saw posts saying "find out if [foo.bar.baz] exists byIG > doing a f$search([foo.bar]baz.dir)".   Yeah, that's quick and elegantd > and general.    A This could be a difficult thing to look for in Usenet. But I justt) tried Google and searched comp.os.vms forl       check directory exists  F and the first result contained an example using F$PARSE. You'd have toA read more than just one article in the thread though. People makeiD mistakes when posting and you always have to watch out for that, and look for corrective posts.  C You could also just ask in comp.os.vms, or study the DCL manual and7E the User's manual. And many have posted the answer. F$PARSE will telllB you if a directory exists without having to transform it as above.    F > Why can't DEC(or COMPAQ, or HP) add a lexical like f$dir_exists() to > the language?i    @ Uh, F$PARSE is exactly that, and much more. Several posters have> already mentioned that. Perhaps your newsfeed is missing them.    + >  Furthermore, why is seem ok with VMSers?C    2 This is not an English sentence. Care to rephrase?        It's almostH > like a religion: "We run around with one shoe because Brian the Chosen
 > one did and E > saw that it was good.  It is heresy to put the other shoe on".   IffA > you look around, almost every other language undergoes periodictB > improvement.   The original unix shell, the Bourne shell, reallyF > sucked for interactive work.  Since then we have the Korn shell, and% > Bourne Again shells, which are bothVD > pretty spiffy once you learn the ropes.  And they include periodicH > improvements. The original VI was great.   But even vi lovers realizedF > there were improvements to be made.  Unlike the attitude I see here,E > there was no "Improve what God himself gave us?"   Now we have VIM,  > for VI Improved.@ > And it's not blasphemy to suggest improvements.   And many are > implemented.    F VMS and DCL each have a long history of improvements. Why do you think
 otherwise?    F > VMSers all grieve when some VMS system gets replaced with pcs.   But; > they never see their arrogance as a factor in its demise.e    E I think DEC and Compaq was largely responsible for this. Many claimed@D that DEC became quite arrogant at times. And their lack of marketingD of VMS didn't help. All I can tell you first hand is that whenever I@ called DEC/Compaq/HP for VMS support they were always excellent,> professional, and very helpful for both VMS and ALL-IN-1. I am> grateful to them and to those who post answers in comp.os.vms.    
    If someG > third-shift operator whacks the wrong files because he's in the wrong H > directory because his prompt is a meaningless 'MFGSYS> ' then everyone > but DCL gets the blame.     ? You should always be careful when deleting files. Instruct your C operators to run SHOW DEFAULT before deleting files. Alternatively,B@ they can explicitly include the disk and directory in the DELETE< argument. If they can't manage that, you need new operators.    '   How much coding do you think would be0F > involved having a prompt that tells you where you are without havingH > to resort to 'cd.com' or 32 character limits, or other such duct-tape?    E Perhaps DEC engineers were too busy with security and clustering, andn> ISAM files, and.... Actually, I've got my own gripes about SET< DEFAULT, but I enjoyed writing some DCL that addresses them.     Disclaimer: JMHO Alan E. Feldman  spamsink2001 at yahoo dot comE   ------------------------------  # Date: Fri, 30 Aug 2002 03:38:40 GMTL1 From: "David J. Dachtera" <djesys.nospam@fsi.net> + Subject: Re: silliest thing heard this weekd' Message-ID: <3D6EEDF7.859766F7@fsi.net>2   Carl Perkins wrote:  > _ > In article <3D6D7691.168EB72C@fsi.net>, "David J. Dachtera" <djesys.nospam@fsi.net> writes...n > }Carl Perkins wrote: > }>: > }> "David J. Dachtera" <djesys.nospam@fsi.net> writes... > }> }John Lewocz wrote: > }> }> [snip]U > }> }> Oh, and speaking of "college days", I swear DCL was written by some undergradhW > }> }> on spring break  -- in the 1940's.   Why is there no builtin way to distinguishaW > }> }> between an empty directory and one that doesn't exist?   Oh sure, I've seen the P > }> }> postings that say "if you want to check if [bozo.the.clown] exists, do aU > }> }> f$search([bozo.the]clown.dir).   But what if you don't know the path ahead ofl
 > }> }> time?e > }> }A > }> }$ PATH = F$PARSE( filespec,,, "DEVICE", "SYNTAX_ONLY" ) + -L7 > }> }F$PARSE( filespec,,, "DIRECTORY", "SYNTAX_ONLY" )t* > }> }$ IF F$PARSE( PATH ) .EQS. "" THEN -9 > }> }$ WRITE SYS$OUTPUT "% Path does not exist - ", PATH- > }> }David J. Dachtera- > }>G > }> This is good in theory, but it misses one case. If there is a filenH > }> called FOO.DIR that is not a directory file, the F$PARSE testing asF > }> above will fail by indicating that there is a [.FOO] subdirectory9 > }> even though there isn't (at least up to VMS V7.2-1).  > }rF > }One could argue that the example you cite constitutes a file systemG > }corruption, and that the lexicals and the underlying system services.1 > }were not intended to diagnose the file system.p > }i > }--- > }David J. Dachtera > $ > One could, but one would be wrong.   Not really.r  # > It is just a file called FOO.DIR. 0 > There isn't actually anything wrong with this.  A Well, yes there is. To borrow from your technique, here's another  example or two:h  
 Example 1:  E It's well known that executable images in Stream_LF format are valid.- So, let's try something:  ( DJAS01::DDACHTERA$ copy nla0: mytest.exe@ %COPY-S-COPIED, _NLA0: copied to DKA0:[DDACHTERA]MYTEST.EXE;1 (0 records)7 DJAS01::DDACHTERA$ set file/attr=(rfm=stmlf) mytest.exe . DJAS01::DDACHTERA$ append sys$input mytest.exe Davidh  Exit ! DJAS01::DDACHTERA$ run mytest.exer2 %DCL-W-ACTIMAGE, error activating image MYTEST.EXE7 -CLI-E-IMGNAME, image file DKA0:[DDACHTERA]MYTEST.EXE;1e8 -IMGACT-F-NOTNATIVE, image is not an OpenVMS Alpha image  C Hhmmm... Guess that extension doesn't mean what its says, does it? t  
 Example 2:  ! Let's take that a step further...t  ( DJAS01::DDACHTERA$ copy nla0: mytest.com@ %COPY-S-COPIED, _NLA0: copied to DKA0:[DDACHTERA]MYTEST.COM;3 (0 records)5 DJAS01::DDACHTERA$ append user$img:zip.exe mytest.comu> %APPEND-W-INCOMPAT, USER$ROOT:[EXE.ALPHA]ZIP.EXE;2 (input) andB DKA0:[DDACHTERA]MYTEST.COM;3 (output) have incompatible attributes DJAS01::DDACHTERA$ @mytest7 %RMS-W-RTB, 512 byte record too large for user's bufferr    Humpph! Another bogus extension!  D Y'know what? VMS sucks! So many inconsistencies! Let's all switch to WhineBloze!c   I'm kidding, of course.n  A See, the VMS documentation never really mentions the concept of a C "reserved" extension: .EXEs exclusively for executable images, datalG sections and the like, .COM exclusively for DCL command procedures, ...g  D DCL, the image activator, even compilers, the Librarian, the Linker,C etc. all apply certain default extensions to input filespec.'s, bute4 there is nothing to stop you from doing things like:   $ DIR [...]/OUT=[-]MYSTUFF.DIR  . $ APPEND File.1,File.2,File.3 File.COM!pendium  F ...and so on. Whether or not it's good practice would be a question ofF considerable debate, I should think. Back to what can be done Vs. what should be done, no?t   > In fact, the file system* > "knows" that it isn't a directory file.   B Well, I don't know as I'd go quite that far. Obviously, the systemB service(s) behind F$SEARCH() in your earlier example *DID* believe? FOO.DIR to be a directory, though some other piece of code took"
 exception.  G > Note that "Directory file" is a file attribute that the non-directorytI > file does not have. Why, exactly, it is trying to use it as a directoryaF > file when it "knows" that it isn't one is a question I can't answer.  H I've often thought that there is a "blurring" of the ODS and RMS layers,F that RMS is attributed with certain file system functions while ODS is2 attributed with certain record-oriented functions.  F Maybe VMS(, UN*X, Linux, etc.) needs a disk data storage equivalent toG the "seven layer model", huh? ...so there would be clearer distinctionsnD between the operations of the ODS (filesystem) and RMS (file system,B with a space between the words) layers, and which layer does what.   -- - David J. Dachtera9 dba DJE Systems0 http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/s   ------------------------------  # Date: Fri, 30 Aug 2002 03:49:13 GMTj1 From: "David J. Dachtera" <djesys.nospam@fsi.net>]+ Subject: Re: silliest thing heard this week ' Message-ID: <3D6EF070.77BDC52D@fsi.net>    "Alan E. Feldman" wrote: > ^ > carl@gerg.tamu.edu (Carl Perkins) wrote in message news:<29AUG200201172643@gerg.tamu.edu>...a > > In article <3D6D7691.168EB72C@fsi.net>, "David J. Dachtera" <djesys.nospam@fsi.net> writes...n > > }Carl Perkins wrote: > > }>< > > }> "David J. Dachtera" <djesys.nospam@fsi.net> writes... > > }> }John Lewocz wrote: > > }> }> [snip]W > > }> }> Oh, and speaking of "college days", I swear DCL was written by some undergradaY > > }> }> on spring break  -- in the 1940's.   Why is there no builtin way to distinguishcY > > }> }> between an empty directory and one that doesn't exist?   Oh sure, I've seen thedR > > }> }> postings that say "if you want to check if [bozo.the.clown] exists, do aW > > }> }> f$search([bozo.the]clown.dir).   But what if you don't know the path ahead ofn > > }> }> time?i > > }> }C > > }> }$ PATH = F$PARSE( filespec,,, "DEVICE", "SYNTAX_ONLY" ) + -e9 > > }> }F$PARSE( filespec,,, "DIRECTORY", "SYNTAX_ONLY" )f, > > }> }$ IF F$PARSE( PATH ) .EQS. "" THEN -; > > }> }$ WRITE SYS$OUTPUT "% Path does not exist - ", PATHr > > }> }David J. DachteraI > > }>I > > }> This is good in theory, but it misses one case. If there is a filegJ > > }> called FOO.DIR that is not a directory file, the F$PARSE testing asH > > }> above will fail by indicating that there is a [.FOO] subdirectory; > > }> even though there isn't (at least up to VMS V7.2-1).e > > }cH > > }One could argue that the example you cite constitutes a file systemI > > }corruption, and that the lexicals and the underlying system servicesE3 > > }were not intended to diagnose the file system.- > > }  > > }--h > > }David J. Dachtera > > H > > One could, but one would be wrong. It is just a file called FOO.DIR.K > > There isn't actually anything wrong with this. In fact, the file systemlK > > "knows" that it isn't a directory file. If you do a DIR/FULL on it, youe > > find something like this:n > >sH > > File attributes:    Allocation: 4, Extend: 0, Global buffer count: 0( > >                     No version limit > > G > > Whereas an actual directory file will give you something like this:a > >9H > > File attributes:    Allocation: 4, Extend: 0, Global buffer count: 0L > >                     No default version limit, Contiguous, Directory file > >5I > > Note that "Directory file" is a file attribute that the non-directory K > > file does not have. Why, exactly, it is trying to use it as a directorynH > > file when it "knows" that it isn't one is a question I can't answer. > >n > > --- Carl > C > The file system assumes that anything of the form name.DIR;1 is ahH > directory file. Everything else is considered a non-directory file. IfG > the name.DIR;1 file doesn't have the directory attribute set for someI; > reason, then it is considered a corrupted directory file.t  : See HELP Lex F$FILE ARG, expecially the DIRECTORY keyword.  
 > Apparently,m1 > F$PARSE doesn't check if it's corrupted or not.    Apparently:n    DJAS01::DDACHTERA$ cr mytest.dir David   Exit + DJAS01::DDACHTERA$ say f$parse("[.mytest]")  DKA0:[DDACHTERA.MYTEST].;'1 DJAS01::DDACHTERA$ say f$search( "[.mytest]*.*" )c  # DJAS01::DDACHTERA$ dir [.mytest]*.*eE %DIRECT-E-OPENIN, error opening DKA0:[DDACHTERA.MYTEST]*.*;* as inputd/ -RMS-E-FND, ACP file or directory lookup faileds0 -SYSTEM-W-BADIRECTORY, bad directory file format  H Note, however, that the facility reporting the error at the intermediate3 level is RMS, not ODS as I personally would expect.s  H To me, the "layers" should work like they in the more primitive o.s.-es:F ODS does the file lookups, using RMS as a record-access layer. Once itF retieves the extent pointers, it should hand off to RMS to perform theD $OPEN, retrieve the record, etc. and RMS should then hand off to theB program/application to interpret the data returned. In the message> above, the lowest level message has a facility name of SYSTEM.  ? The "borders" between the layers seem blurred or confused, IMO.o   > That's good becauseoH > the corruption could be caused by other reasons, and you would want toA > know that there is such a directory when you ask. IOW, by usingpD > F$PARSE you are asking "Is there such a directory?", not "Is there9 > such a directory that has not been corrupted somehow?".r  H Very true. Still, it does point up something that could be considered an inconsistency.   -- s David J. Dachterai dba DJE Systemst http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/T   ------------------------------  # Date: Fri, 30 Aug 2002 04:24:05 GMTl* From: "Bill Todd" <billtodd@metrocast.net>+ Subject: Re: silliest thing heard this week B Message-ID: <FrCb9.171517$Aw4.7336007@bin2.nnrp.aus1.giganews.com>  I OK, since no one with recent knowledge seems to be trying to explain thisr+ behavior, I'll try to dredge up what I can:   < "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message! news:3D6EF070.77BDC52D@fsi.net...T > "Alan E. Feldman" wrote:   ...(   > > Apparently, 3 > > F$PARSE doesn't check if it's corrupted or not.  > 
 > Apparently:   I IIRC $PARSE on the 11 performed no I/O:  it just looked at syntax (and, I J think, performed at least some logical name translation - but perhaps onlyE if no nodespec was discovered).  That, after all, is pretty much what K 'parse' means - and since there's no guarantee that it's being applied to aaG path specification on the local system, doing otherwise could be fairly  restrictive.  L Now, on VMS it may do some actual look-up if no nodespec is discovered (thatK may be controlled by the 'syntax only' qualifier mentioned), but it's stilltC not surprising that it might not check something that looked like aoK directory if it was the last path element specified (and it may not even be - possible:  see additional explanation below).    > " > DJAS01::DDACHTERA$ cr mytest.dir > Davidm >  Exiti- > DJAS01::DDACHTERA$ say f$parse("[.mytest]")i > DKA0:[DDACHTERA.MYTEST].;d3 > DJAS01::DDACHTERA$ say f$search( "[.mytest]*.*" )  >o% > DJAS01::DDACHTERA$ dir [.mytest]*.* G > %DIRECT-E-OPENIN, error opening DKA0:[DDACHTERA.MYTEST]*.*;* as input 1 > -RMS-E-FND, ACP file or directory lookup failedt2 > -SYSTEM-W-BADIRECTORY, bad directory file format >lJ > Note, however, that the facility reporting the error at the intermediate5 > level is RMS, not ODS as I personally would expect.d >'J > To me, the "layers" should work like they in the more primitive o.s.-es:H > ODS does the file lookups, using RMS as a record-access layer. Once itH > retieves the extent pointers, it should hand off to RMS to perform theF > $OPEN, retrieve the record, etc. and RMS should then hand off to theD > program/application to interpret the data returned. In the message@ > above, the lowest level message has a facility name of SYSTEM. >CA > The "borders" between the layers seem blurred or confused, IMO.m  A No, they just don't fit the particular (and arbitrary, though notc% unreasonable) model you came up with.   J ODS-1/2 *never* (IIRC) performed path lookups:  they just take a directoryK ID and a file name and return the FID (if any) associated with that name inAI that directory.  FCS and RMS on RSX, and RMS on VMS, invoke that facilityMF iteratively to follow a path.  Among other things, this keeps the fileI system (as distinct from RMS) from having to worry about things like node K specifications and what to do with them:  this may not be as big a deal now B with the XQP, but back in the ACP days it would have been awkward.   >u > > That's good because J > > the corruption could be caused by other reasons, and you would want toC > > know that there is such a directory when you ask. IOW, by usingtF > > F$PARSE you are asking "Is there such a directory?", not "Is there; > > such a directory that has not been corrupted somehow?".a > J > Very true. Still, it does point up something that could be considered an > inconsistency.  G Only if you invest certain file extensions with meaning, which the filecI system does not.  It *reserves* the combination .DIR;1 for its own use inrJ the sense that if you create a file with that extension you can't create aH directory under the same parent with the same name the file has (or viceJ versa), but it does not invest that combination with any meaning:  it uses* the file's attribute information for that.  G But that does mean that you can't unambiguously use F$PARSE to see if abI subdirectory FOO exists by looking under its parent for FOO.DIR;1.  So if(J you're depending upon that construct as the way to handle the problem thatL was originally posed, then the answer is simply that VMS doesn't provide the  facilities to solve it that way.   - bill   ------------------------------   Date: 30 Aug 2002 00:19 CDT ' From: carl@gerg.tamu.edu (Carl Perkins)e+ Subject: Re: silliest thing heard this weekh- Message-ID: <30AUG200200194745@gerg.tamu.edu>I  ] In article <3D6EEDF7.859766F7@fsi.net>, "David J. Dachtera" <djesys.nospam@fsi.net> writes...r }Carl Perkins wrote: }> n` }> In article <3D6D7691.168EB72C@fsi.net>, "David J. Dachtera" <djesys.nospam@fsi.net> writes... }> }Carl Perkins wrote:a }> }>o; }> }> "David J. Dachtera" <djesys.nospam@fsi.net> writes...l }> }> }John Lewocz wrote:n }> }> }> [snip] V }> }> }> Oh, and speaking of "college days", I swear DCL was written by some undergradX }> }> }> on spring break  -- in the 1940's.   Why is there no builtin way to distinguishX }> }> }> between an empty directory and one that doesn't exist?   Oh sure, I've seen theQ }> }> }> postings that say "if you want to check if [bozo.the.clown] exists, do ahV }> }> }> f$search([bozo.the]clown.dir).   But what if you don't know the path ahead of }> }> }> time? }> }> }wB }> }> }$ PATH = F$PARSE( filespec,,, "DEVICE", "SYNTAX_ONLY" ) + -8 }> }> }F$PARSE( filespec,,, "DIRECTORY", "SYNTAX_ONLY" )+ }> }> }$ IF F$PARSE( PATH ) .EQS. "" THEN -u: }> }> }$ WRITE SYS$OUTPUT "% Path does not exist - ", PATH }> }> }David J. Dachtera }> }>rH }> }> This is good in theory, but it misses one case. If there is a fileI }> }> called FOO.DIR that is not a directory file, the F$PARSE testing as1G }> }> above will fail by indicating that there is a [.FOO] subdirectory : }> }> even though there isn't (at least up to VMS V7.2-1). }> }G }> }One could argue that the example you cite constitutes a file systemGH }> }corruption, and that the lexicals and the underlying system services2 }> }were not intended to diagnose the file system. }> } }> }-- }> }David J. Dachteran }> e% }> One could, but one would be wrong.  }  }Not really. } $ }> It is just a file called FOO.DIR.1 }> There isn't actually anything wrong with this.  } B }Well, yes there is. To borrow from your technique, here's another }example or two: }  }Example 1:a } F }It's well known that executable images in Stream_LF format are valid. }So, let's try something:r } ) }DJAS01::DDACHTERA$ copy nla0: mytest.exenA }%COPY-S-COPIED, _NLA0: copied to DKA0:[DDACHTERA]MYTEST.EXE;1 (0 	 }records)t8 }DJAS01::DDACHTERA$ set file/attr=(rfm=stmlf) mytest.exe/ }DJAS01::DDACHTERA$ append sys$input mytest.exea }David } Exit -" }DJAS01::DDACHTERA$ run mytest.exe3 }%DCL-W-ACTIMAGE, error activating image MYTEST.EXEh8 }-CLI-E-IMGNAME, image file DKA0:[DDACHTERA]MYTEST.EXE;19 }-IMGACT-F-NOTNATIVE, image is not an OpenVMS Alpha image2 } D }Hhmmm... Guess that extension doesn't mean what its says, does it?    Of course it doesn't..   This is not Windows.  ? A files extension is just text. It does not "mean" anything. Ann? executable can have any extension you want to give it. "EXE" is]@ just the default that the RUN command uses if none is specified.= Likewsie, you can give any file the .EXE extension. That does'< not make in an executable just a file with a .EXE extension.   }Example 2:r } " }Let's take that a step further... } ) }DJAS01::DDACHTERA$ copy nla0: mytest.com A }%COPY-S-COPIED, _NLA0: copied to DKA0:[DDACHTERA]MYTEST.COM;3 (0e	 }records)t6 }DJAS01::DDACHTERA$ append user$img:zip.exe mytest.com? }%APPEND-W-INCOMPAT, USER$ROOT:[EXE.ALPHA]ZIP.EXE;2 (input) and,C }DKA0:[DDACHTERA]MYTEST.COM;3 (output) have incompatible attributesR }DJAS01::DDACHTERA$ @mytest 8 }%RMS-W-RTB, 512 byte record too large for user's buffer } ! }Humpph! Another bogus extension!   	 Nonsense.D   Once again, the is not Windows.n  = The extension is in no way bogus. An extension means nothing.   It is just part of the filespec.  B }See, the VMS documentation never really mentions the concept of aD }"reserved" extension: .EXEs exclusively for executable images, dataH }sections and the like, .COM exclusively for DCL command procedures, ...  E No kidding. That is because they are not "reserved". You can use themr as you see fit.i   }> In fact, the file systemi+ }> "knows" that it isn't a directory file. v } C }Well, I don't know as I'd go quite that far. Obviously, the system C }service(s) behind F$SEARCH() in your earlier example *DID* believen@ }FOO.DIR to be a directory, though some other piece of code took }exception.  } H }> Note that "Directory file" is a file attribute that the non-directoryJ }> file does not have. Why, exactly, it is trying to use it as a directoryG }> file when it "knows" that it isn't one is a question I can't answer.A } I }I've often thought that there is a "blurring" of the ODS and RMS layers,mG }that RMS is attributed with certain file system functions while ODS is 3 }attributed with certain record-oriented functions., } G }Maybe VMS(, UN*X, Linux, etc.) needs a disk data storage equivalent toAH }the "seven layer model", huh? ...so there would be clearer distinctionsE }between the operations of the ODS (filesystem) and RMS (file system, C }with a space between the words) layers, and which layer does what.R }--  }David J. Dachtera  E Or maybe it should just check the attribute in the appropriate places)E to make sure that it is supposed to be a directory file before tryingO( to do directory file type stuff with it.  C For the most part it is fine as it is, but it might be better if ittC checked. It would eliminate the situation that prevents the F$PARSE A test from being all you need to do to check for the exsistance ofa a directory.   --- Carl   ------------------------------  # Date: Fri, 30 Aug 2002 03:10:04 GMTn1 From: "David J. Dachtera" <djesys.nospam@fsi.net>s Subject: Re: Simple cluster"' Message-ID: <3D6EE746.42AE0F23@fsi.net>b   Robert Deininger wrote:a > = > In article <3D6D74C5.54B62C7B@fsi.net>, "David J. Dachtera"C  > <djesys.nospam@fsi.net> wrote: > L > >> > > As others have responded, this won't work.  With 2 systems having 1L > >> > > vote each, quorum is 2 ((n+1)/2).  If one system crashes, the other > >> > > hangs.- > >> > > >> > 2+1=30 > >> > 3/2=1 (integer math, remainder truncated) > >> >$ > >> > Therefore, quorum = 1, right? > >>$ > >> QUORUM = (EXPECTED_VOTES + 2)/2 > >>P > >> (2+2)/2 = 2 so the quorum is 2. If it were one the cluster could partition. > >y
 > >Try again:  > >i" > >QUORUM = (EXPECTED_VOTES + 1)/2 > >                           ^  > >2+1=3 > >3/2=1 >  > No!v > J > The manual has several paragraphs in a table, but the relevant line says@ > "QUORUM = (EXPECTED_VOTES + 2)/2        |        Rounded down"  E Actually, you're quite right. I shoulda looked it up. I just took the  poster at his word.h  D Not sure why, but at work I can't find the "OpenVMS Cluster Systems"@ manual in hardcopy, just the Cluster Configuration Guidelines...   -- t David J. Dachterar dba DJE Systemsa http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/t   ------------------------------  # Date: Fri, 30 Aug 2002 03:11:02 GMTe1 From: "David J. Dachtera" <djesys.nospam@fsi.net>y Subject: Re: Simple clustero& Message-ID: <3D6EE780.A960C0F@fsi.net>   Phillip Helbig wrote:f > 
 > > > > 2+1=3(1 > > > > 3/2=1 (integer math, remainder truncated)s > > > >c% > > > > Therefore, quorum = 1, right?c > > > % > > > QUORUM = (EXPECTED_VOTES + 2)/2/ > > >fQ > > > (2+2)/2 = 2 so the quorum is 2. If it were one the cluster could partition.- > >- > > Try again: > >e# > > QUORUM = (EXPECTED_VOTES + 1)/2m  > >                            ^	 > > 2+1=3]	 > > 3/2=1R > J > Sorry Dave, quorum is 2 for a 3-machine cluster.  See section 2.3 of theF > manual OPENVMS CLUSTER SYSTEMS, which explains the (correct) formula? > above.  Even without the math, you can see that it must be 2.c  = See my response to Robert D. at another point in this thread.m   -- t David J. DachteraC dba DJE Systems: http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  # Date: Fri, 30 Aug 2002 03:12:36 GMT 1 From: "David J. Dachtera" <djesys.nospam@fsi.net>  Subject: Re: Simple clustery' Message-ID: <3D6EE7DE.2C52B5BC@fsi.net>    JF Mezei wrote:n >  > Phillip Helbig wrote: L > > Sorry Dave, quorum is 2 for a 3-machine cluster.  See section 2.3 of the > J > > If it were 1, suppose you lose connection.  Each member could think IT* > > is the cluster.  Bad state of affairs. > N > Quorum could be 1 for a 96 machine cluster. Your boot node would have 1 voteN > and all others would have 0 votes. Lose the boot node and all other machinesK > freeze, but you could lose any number of satellites without affecting thePL > remaining cluster. If a satellite looses a connection to the boot node, itP > would freeze even if it has its own system disk since "its" cluster would have( > 0 votes of the 1 needed to stay awake.  H Hhmmm... Have to admit, I never thought of it that way, but you're quite right.   -- c David J. Dachterah dba DJE Systems( http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/l   ------------------------------  % Date: Thu, 29 Aug 2002 15:51:19 -0400s- From: JF Mezei <jfmezei.spamnot@videotron.ca>e Subject: Support for MPE (OT)h, Message-ID: <3D6E7B23.1A429944@videotron.ca>  U OK, I know it isn't directly related to VMS, but I found it of interest nevertheless.h- (PS. at this time HP is trading below $14.00)   9 Bradmark Reaffirms Its Support of the HP e3000 Community;U> Maintenance Savings Plan Secures Support and Savings On Future Upgrade Costs      8/29/02 9:00am 0  G HOUSTON--(BUSINESS WIRE)--Aug. 29, 2002--Bradmark Technologies Inc., an D established provider of data management solutions, today announced aN Maintenance Savings Plan aimed at easing the concerns of those loyal customers: who will be migrating from the HP 3000 to other platforms.  N Since Hewlett Packard's decision to discontinue support for the HP e3000, manyN users have developed their migration strategies, but in most cases these plansL will take some time to complete. In the meantime, HP e3000 users are lookingC for continuity and stability with their current software providers.a  G Bradmark's Maintenance Savings Plan allows current customers to receivecL preferential rates for software maintenance for up to 24 months, giving themK time to carefully consider their choices and fully implement the migration,iM knowing that their maintenance rates will not increase and that Bradmark will # continue to provide 24 x 7 support.c  M "We are committed to providing the highest level of support for our customers : by continuing to back the HP 3000 community  <etc etc etc>    N   This program also includes special pricing on Bradmark's DBGeneral and other; products for the HP 3000, such as SUPERDEX(R) <etc etc etc>   L   "HP 3000 users know what it means to have high availability systems," saidL Tashenberg. "As they migrate to other platforms and databases, Bradmark willE also be there with tools to help them maintain those high standards."   L   DBGeneral(R) and NORAD(TM) are available direct from Bradmark Technologies2 or through its  worldwide network of distributors.  "   About Bradmark Technologies Inc.  J   Bradmark Technologies Inc. provides management tools for Hewlett-Packard IMAGE databasese   <etc etc etc>    -  CONTACT: Bradmark Technologies Inc., Houstone   ------------------------------   Date: 29 Aug 2002 18:11:25 GMT* From: Casper H.S. Dik <Casper.Dik@Sun.COM>6 Subject: Re: The "Gold" Standard in Disaster Tolerance* Message-ID: <aklo4d$hnr$4@news1.xs4all.nl>  ( Joshua Cope <Joshua.Cope@hp.com> writes:   >Bill Gunshannon wrote:y >> uJ >> While I am not knocking the disaster tolerance of OpenVMS, there was/isJ >> a much cheaper solution to the disaster tolerance problem as related toE >> 9/11.  It's simple.  Put your critical datacenters in very obscureS0 >> locations and don't advertise where they are.  L >With the money you save on rent from moving out of the big city, maybe you H >could *afford* two sites! You're right; it fixes the problem as relatedJ >to 9/11, but if you're running a truly critical service, you're probably I >still awake nights worrying about earthquakes, flash floods, tornadoes, :M >hurricanes, chemical spills, fires, and the new college intern accidentally r >hitting the Big Red Button.  > Wyoming, Idaho?  (Only need a bullet/stampede proof building?)   Casper -- 'F Expressed in this posting are my opinions.  They are in no way related2 to opinions held by my employer, Sun Microsystems.? Statements on Sun products included here are not gospel and may  be fiction rather than truth.'   ------------------------------    Date: 29 Aug 2002 13:52:28 -0600- From: koehler@encompasserve.org (Bob Koehler) 6 Subject: Re: The "Gold" Standard in Disaster Tolerance3 Message-ID: <X7IB3dFbn2Tj@eisner.encompasserve.org>   e In article <akl6p3$1k2uct$1@ID-135708.news.dfncis.de>, bill@gw5.cs.uofs.edu (Bill Gunshannon) writes:p > I > While I am not knocking the disaster tolerance of OpenVMS, there was/isiI > a much cheaper solution to the disaster tolerance problem as related topD > 9/11.  It's simple.  Put your critical datacenters in very obscureJ > locations and don't advertise where they are.  With communications beingL > what it is today, what exactly was the advantage of putting the datacenter > in the WTC building??  o  <    Availability of large numbers of talented support people.   ------------------------------  % Date: Thu, 29 Aug 2002 21:22:34 +0200e0 From: "Philip Lewis" <FerrariTR512m@hotmail.com>6 Subject: Re: The "Gold" Standard in Disaster Tolerance- Message-ID: <aklsel$1mqb$1@news.cybercity.dk>-  
 Gentlemen,  = I giver the "Power Barge" http://www.bwsc.dk/power.asp?ppl=15C    > "Jerry Leslie" <LESLIE@JRLVAX.HOUSTON.RR.COM> wrote in message7 news:38rb9.371330$q53.12222637@twister.austin.rr.com...y< > Webb, William W Raleigh, NC (wwebb1@email.usps.gov) wrote: > :s= > : Any of those old Minuteman missile silos still available?w > :i( > : Talk about a Disaster-Tolerant site! > :o2 > : One of those would be just perfect for some of4 > : those 100,000 U racks I've got in the warehouse. > :d > : :^)M > :p1 > Taking a cue from Jack Welch, former CEO of GE:R >s( >    http://www.progress.org/corpw30.htm; >    The Export-Import Bank: Corporate Welfare At Its Worstv >tH >   "...In fact, while taking the Ex-Im Bank subsidies, GE was extremelyH >    public about it's "globalization" plans to lay off American workersI >    and move jobs to Third World countries. Jack Welch, the longtime CEOaJ >    of GE stated, "Ideally, you'd have every plant you own on a barge..." >aH > Perhaps it's time for barge- or ship-mounted datacenters that could be- > moved to the cheapest labor market du jour.  >,E > Or purchase inactive offshore drilling rigs and have them installedA outsiderI > the U.S. 12-mile limit, and just fly in workers from the cheapest laborw+ > market du jour: India, Vietnam, Kenya....a >wG > Ford Motor Company has already moved IT-enabled jobs from the U.S. tou% > India, saving $ 30-60 million/year:e >w1 >    http://www.computerworld.com/softwaretopics/r0 >    software/appdev/story/0,10801,58739,00.html2 >    Ford opens IT hub in India to save millions |) >    Computerworld News & Feature Stories  > C >   "..Ford expects to save $30 million to $60 million per year..."r >oA > The Chairman of the Programmers' Guild claims the U.S. paid for @ > the high-speed data links to India, so telecomm costs wouldn't > be an issue for Ford.e >c >a4 > --Jerry Leslie   (my opinions are strictly my own); >   Note: leslie@jrlvax.houston.rr.com is invalid for emailn   ------------------------------  % Date: Thu, 29 Aug 2002 17:02:18 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca>t6 Subject: Re: The "Gold" Standard in Disaster Tolerance, Message-ID: <3D6E8BC2.AEF8BEAA@videotron.ca>   Philip Lewis wrote:t >  > Gentlemen, > ? > I giver the "Power Barge" http://www.bwsc.dk/power.asp?ppl=15A  N How long and how fat an extention cord do they run between the power barge and the mainland ?   ------------------------------   Date: 29 Aug 2002 17:41 CDT ' From: carl@gerg.tamu.edu (Carl Perkins)e6 Subject: Re: The "Gold" Standard in Disaster Tolerance- Message-ID: <29AUG200217414067@gerg.tamu.edu>-  e In article <0z51$Bfaz$V9@eisner.encompasserve.org>, Kilgallen@SpamCop.net (Larry Kilgallen) writes...lf }In article <akl6p3$1k2uct$1@ID-135708.news.dfncis.de>, bill@gw5.cs.uofs.edu (Bill Gunshannon) writes:@ }> In article <Ws4b9.364974$q53.11990354@twister.austin.rr.com>,7 }> 	LESLIE@JRLVAX.HOUSTON.RR.COM (Jerry Leslie) writes:.  }>> Cypherpunk@nyc.rr.com wrote: }>> N }>>: OpenVMS, as the article indicated, was just one of many types of systems. }>>: hC }>>    http://www.openvms.compaq.com/openvms/brochures/commerzbank/ E }>>    hp Alphaserver technology helps Commerzbank tolerate disaster D }>>    on September 11 }>>  }> >J }> While I am not knocking the disaster tolerance of OpenVMS, there was/isJ }> a much cheaper solution to the disaster tolerance problem as related toE }> 9/11.  It's simple.  Put your critical datacenters in very obscuree0 }> locations and don't advertise where they are. } C }That works for terrorists, but most people who can afford disasterfF }tolerance also need something that works for hurricanes/tornados/etc. } A }So the disaster tolerance software features are still necessary,X9 }even though one potential disaster might be less likely.}  B In such remote locations it is also a lot less likely that you canC arrange for redundant lines of communication (unless "two differentaA wires in the same bundle" is what you call "redundant"). Thus youyJ may find that you excercise your "distaster" tolerance everytime Billy-BobH severs the line(s) with his backhoe. (We have lost our network and phoneI connection this way twice, I think it is, in the last year - each time ituE is out for at least a full day since it takes time to splice togetherlE all the wires in a bundle correctly. At some point we are supposed to.D get connected via the new fiber optic lines that are up on poles andG which run the other way down the street - thus a backhoe incident would.I only take out the phone lines, not the network, and anything that managesiG to take out the network by cutting a line up on poles, or knocking downEE a pole, would not affect the phone lines. Not exactly redundancy, butrA separation of the two different means of communication at least.)    --- Carl   ------------------------------  # Date: Fri, 30 Aug 2002 01:15:10 GMT1* From: "Bill Todd" <billtodd@metrocast.net>* Subject: tru64.org and openvms.org changesC Message-ID: <yGzb9.282498$m91.11451621@bin5.nnrp.aus1.giganews.com>H  G Tru64.org and openvms.org are a good deal quieter than they used to be, G apparently largely because the forums have been redefined to avoid manyAK kinds of discussions (both technical - e.g., the relative merits of currentnI and future hardware platforms - and non-technical - e.g., the behavior of L the parent corporation) that had until recently spiced them up.  If the restI of the material at the sites had changed similarly that would at least betJ consistent.  However, Ken still trots out articles of all types that mightL be of interest to his former readership, but now provides no forums in which to discuss them.  K For example, he just posted the HP press release about new Superdome 'world0F records' (thinly disguised as a Business Wire article, but with the HPI disclaimer at its end) in tru64.org, which really begs responses pointingeJ out that the TPC-C result is not any kind of world record at all, at leastB not at the tpc.org site, which still lists the Fujitsu SPARC-basedI PRIMEPOWER entry at the top of the TPC-C non-clustered list, with the newdG Superdome entry firmly in second place.  True, the Fujitsu box uses 128oI processors and the Superdome only 64, but then the IBM Regatta nipping atbJ Superdome's heels in third place uses only 32, so that's hardly a relevantI factor in Superdome's favor - though the article acts as if it were later H when comparing Superdome's per-processor performance to SunFire's in theB TPC-H area, even though the difference is far less dramatic there.I Similarly, the article points out the marginal cost/performance advantagefK Superdome enjoys over Regatta in TPC-C, but fails to point out the marginall1 advantage SunFire enjoys over Superdome in TPC-H.f  I And in openvms.org he posted another thinly-veiled HP press release aboutmJ its recent financial performance, which nearly completely glosses over theK massive loss (over $2 billion) in the just-completed quarter by emphasizing L 'pro forma' (and unaudited) results that simply keep sweeping expenses underI a rather arbitrary rug until the fudged numbers work out the way HP wouldoL have liked them to.  The one reference to *actual* performance both fails toK identify the dollar amount of the loss and presents the per-share loss withoI the notation 'GAAP':  one might easily believe that *that* was the result L that had only some kind of nominal, rather than substantive, significance ifD one did not know that GAAP stands for 'generally accepted accountingC practices' - i.e., the way results are *supposed* to be calculated.   L To Ken's credit, he also provides pointers to articles less favorable to HP.G But for sites that claim to be 'forums' for those interested in the twop? operating systems, they seem far more 'read-only' now than thatoL characterization used to suggest.  Since they don't even provide a means forD reader feedback (at least that I could find), I'm providing it here.   - bill   ------------------------------  % Date: Thu, 29 Aug 2002 22:14:05 -0500e7 From: "Craig A. Berry" <craigberry@nospam.telocity.com> . Subject: Re: tru64.org and openvms.org changesG Message-ID: <craigberry-42EBE5.22140529082002@news.directvinternet.com>p  C In article <yGzb9.282498$m91.11451621@bin5.nnrp.aus1.giganews.com>,F,  "Bill Todd" <billtodd@metrocast.net> wrote:  F >  for sites that claim to be 'forums' for those interested in the twoA > operating systems, they seem far more 'read-only' now than that $ > characterization used to suggest.   H Nonsense.  On openvms.org I see five fora listed where they always have G been, most or all linked from the home page under the "Forums" section aD on the left sidebar.  I think there are actually more of these than  there used to be.C  + > Since they don't even provide a means for F > reader feedback (at least that I could find), I'm providing it here.  E Ken's address is easy to find on the "about" page and you can always  H submit an article via the "contribute" page linked from a tab along the  top of the home page.   G The only substantive change I've seen lately is that you can now order RD goodies like t-shirts, coffee mugs and such that say OpenVMS.org on  them.t   ------------------------------  # Date: Fri, 30 Aug 2002 03:52:13 GMTt* From: "Bill Todd" <billtodd@metrocast.net>. Subject: Re: tru64.org and openvms.org changesB Message-ID: <NZBb9.170890$Aw4.7312173@bin2.nnrp.aus1.giganews.com>  B "Craig A. Berry" <craigberry@nospam.telocity.com> wrote in messageA news:craigberry-42EBE5.22140529082002@news.directvinternet.com... E > In article <yGzb9.282498$m91.11451621@bin5.nnrp.aus1.giganews.com>, . >  "Bill Todd" <billtodd@metrocast.net> wrote: >.H > >  for sites that claim to be 'forums' for those interested in the twoC > > operating systems, they seem far more 'read-only' now than that,% > > characterization used to suggest.R >=I > Nonsense.  On openvms.org I see five fora listed where they always haveiH > been, most or all linked from the home page under the "Forums" sectionE > on the left sidebar.  I think there are actually more of these than  > there used to be.   J There are 4 fora listed at the left on the home page, none of which is theI old general-purpose VMS forum that used to exist there (at the top of thebK list) and held, as I mentioned, often more interesting contents.  And whilesG Terry's forum is still listed on the right-hand side, it hasn't had anyr> activity for months (nor is it the old general forum, either).  8 So your definition of 'nonsense' may need a little work.  I However, I've just discovered that by going to one of those fora in whichmH I'm not interested, I can see a link to a 'forum list' - and, by George,I *that* list (not as far as I can determine accessible from, or in any waylI suggested by, the home page) *does* contain the old HP-Compaq and OpenVMSeJ fora.  And exactly the same is true over at tru64.org:  the old (and stillL only) technical/general forum is now invisible from the home page, but stillJ accessible if you go somewhere else you don't care about and thence to the magic 'forum list'.d  H One really can't help but wonder why the most active fora on these sitesL suddenly became so cloaked in obscurity:  it's just *so* reminiscent of HP'sJ inclination to 'manage' activity that it's uncomfortable with.  But ratherK than dwell on that notion, I'd be perfectly happy just to see them restoredf% to their old positions of prominence..   > - > > Since they don't even provide a means foroH > > reader feedback (at least that I could find), I'm providing it here. >d3 > Ken's address is easy to find on the "about" page:  L Had I been sufficiently motivated, I likely could have discovered that.  ButL since I was also interested in whether others might feel the same way I did,H I just noted the fact that there wasn't any prominently-displayed way to" contact him and came here instead.    and you can alwaysfI > submit an article via the "contribute" page linked from a tab along the1 > top of the home page.r  K I considered that, but it's not exactly an easy (or particularly timely, inpK terms of responsiveness) way to hold a discussion.  The old forum was both.(   >EH > The only substantive change I've seen lately is that you can now orderE > goodies like t-shirts, coffee mugs and such that say OpenVMS.org ong > them.m  L I guess you never visited the old forum, then.  Not that I think there's anyG reason you should have if you weren't interested in it, but that hardlycK makes you qualified to assess the effect of its complete disappearance froms" its former position of prominence.   - bill   ------------------------------  # Date: Thu, 29 Aug 2002 20:19:01 GMT 0 From: "warren sander" <warren.sander@compaq.com>F Subject: Re: White Paper on DT clustering - VMS as the "Gold Standard"1 Message-ID: <Vkvb9.42$Au.584828@news.cpqcorp.net>w  C or http://www.openvms.compaq.com/openvms/whitepapers/illuminata.pdfs   -warrenc  6 "Paul Sture" <p_sture@elias.decus.ch> wrote in message# news:cGCAVv6CmpRo@elias.decus.ch...a9 > White Paper from Illuminata on DT Clustering in variouso< > Unix flavours (emphasis on OpenVMS as the "Gold Standard") >rJ > White paper comparing clustering techniques in the light of DT (Disaster
 > Tolerance).s >o > >aI http://www.tru64unix.compaq.com/unix/illuminata_dt_unix_research_note.pdfr >e > I found this a good read.  > __ > Paul Sture
 > Switzerlandk   ------------------------------   End of INFO-VAX 2002.477 ************************