1 INFO-VAX	Wed, 22 Sep 2004	Volume 2004 : Issue 526       Contents:, Re: a simple question: what the h*ll is MCR?) Re: AlphaServer DS10L - IDE disk question  Re: BASIC EDIT$ - 256  Re: BASIC EDIT$ - 256  China girl in Paris ( Re: Completed batch job still executing? Dapper Dan does a Kerry < Does anyone use Compaq C++ V6.5 on Alpha seriously with STL?@ Re: Does anyone use Compaq C++ V6.5 on Alpha seriously with STL? Mapping UNIX File Names  Re: Mapping UNIX File Names  Re: Mixed Cluster certification  Re: Off-the-wall CI QuestionP Re: OpenVMS Pearl -  Indian Railways Celebrates 10th anniversary using  HP TechnP OpenVMS Pearl -  Indian Railways Celebrates 10th anniversary using HP TechnologyP Re: OpenVMS Pearl -  Indian Railways Celebrates 10th anniversary using HP Techno& OT: Sun admits mistake of droping 8086* Re: OT: Sun admits mistake of droping 8086 Re: Program to create users  Re: Program to create users ( Readable error log while process running, Re: Readable error log while process running, Re: Readable error log while process running3 Saveset File that OpenVMS ZIP V2.3 will not archive 7 Re: Saveset File that OpenVMS ZIP V2.3 will not archive 7 Re: Saveset File that OpenVMS ZIP V2.3 will not archive 7 Re: Saveset File that OpenVMS ZIP V2.3 will not archive 7 Re: Saveset File that OpenVMS ZIP V2.3 will not archive & Re: Sun admits mistake of droping 8086  Re: TCP/IP cluster interconnect?  Re: TCP/IP cluster interconnect?  Re: TCP/IP cluster interconnect?  Re: TCP/IP cluster interconnect?  Re: TCP/IP cluster interconnect?  Re: TCP/IP cluster interconnect?  Re: TCP/IP cluster interconnect? user privilege Re: user privilege Re: user privilege Re: user privilege( Re: VAX 6000 replacement with CHARON-VAX( Re: VAX 6000 replacement with CHARON-VAX( Re: VAX 6000 replacement with CHARON-VAX5 Re: Windoze not rebooted monthly shuts down airports! 5 Re: Windoze not rebooted monthly shuts down airports! 5 Re: Windoze not rebooted monthly shuts down airports! 5 Re: Windoze not rebooted monthly shuts down airports! 5 Re: Windoze not rebooted monthly shuts down airports! 5 Re: Windoze not rebooted monthly shuts down airports! 5 Re: Windoze not rebooted monthly shuts down airports! 5 Re: Windoze not rebooted monthly shuts down airports! 5 Re: Windoze not rebooted monthly shuts down airports! 5 Re: Windoze not rebooted monthly shuts down airports! 5 Re: Windoze not rebooted monthly shuts down airports! 5 Re: Windoze not rebooted monthly shuts down airports! : Re: ZIP "-V" v. UNIX, et al.: Problem, possible solutions.: Re: ZIP "-V" v. UNIX, et al.: Problem, possible solutions.: Re: ZIP "-V" v. UNIX, et al.: Problem, possible solutions.: Re: ZIP "-V" v. UNIX, et al.: Problem, possible solutions.8 [DCL REQUEST] New ignore keyword for DIFFERENCES command Re: [OT]: McNeally understands  F ----------------------------------------------------------------------  # Date: Tue, 21 Sep 2004 23:48:09 GMT   From: John Santos <john@egh.com>5 Subject: Re: a simple question: what the h*ll is MCR? * Message-ID: <Z434d.4201$Bg5.2461@trnddc07>   Glenn Everhart wrote:  [...]   H > When in V3 it was decided that most all of the commands that had been  > uniqueJ > to compat mode code now had native mode VAX equivalents, compat mode wasM > dropped and moved to the VAX/RSX product. By that time people realized that M > "compatibility mode" was better named "rough similarity mode" in many cases   C I once overheard a Deccie at DECUS (at the next table at lunch) say J "Compatible means different, because if it was the same, we would say so."   [...]    --   John Santos  Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  % Date: Tue, 21 Sep 2004 18:04:41 -0400 2 From: "Paul A. Jacobi" <Paul.Jacobi@nospam.hp.com>2 Subject: Re: AlphaServer DS10L - IDE disk question, Message-ID: <4150a691$1@usenet01.boi.hp.com>  . "Gremlin" <not-here@all.mate> wrote in message) news:415017ae$1@duster.adelaide.on.net... < > Can a AlphaServer DS10L use a large IDE drive - say 200Gb?  H The SRM console and the VMS DQDRIVER has a 137Gb limit on IDE drives.  I/ don't know any plans to remove this limitation.   H Be aware that even Windows XP requires SP1 to support drives larger than: 137Gb.  See http://news.zdnet.com/2100-9584_22-946929.html       Paul A. Jacobi HP OpenVMS Systems Group
 Nashua, NH   ------------------------------  % Date: Tue, 21 Sep 2004 20:56:21 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net> Subject: Re: BASIC EDIT$ - 256+ Message-ID: <4150DBC5.54563590@comcast.net>     David B Sneddon - bigpond wrote: > ( > David J Dachtera mentioned in passing: > > Joe wrote: > >  > >>Alpha BASIC V1.5-000 > >>OpenVMS V7.3-2 > >>I > >>256 is suppose to "Do not alter characters inside quotes" but I guess D > >>I always figured they were talking about "balanced" ones. Has it > >>always been this way?? > >> > >>$ ty test.bas A > >>option type = explicit, size = integer long,                & # > >>        constant type = integer  > >>= > >>print edit$("COEUR D'ALENE, ID 83814",2+4+8+16+32+64+128) A > >>print edit$("COEUR D'ALENE, ID 83814",2+4+8+16+32+64+128+256)  > >>$  > >>$ run test > >>COEURD'ALENE,ID83814 > >>COEURD'ALENE, ID 83814 > >  > > A > > Not sure - never tested that. Looks like a "feature" to me...  > > 0 > > When you said "quotes", I assumed you meant: > > . > > A$ = "This is a ""quoted literal"" string" > > Print EDIT$( a$, 254 ) > > Print EDIT$( a$, 510 ) > > E > > (Embedded double-quotes in the value of a string variable, not an * > > embedded apostrophe ("single quote").) > > 
 > > D.J.D. > >  > H > BASIC allows strings to be quoted with either single or double quotes.  3 Hhmmm... YAUF (Yet Another Undocumented "Feature").   E > Since the string is parsed from left to right, the original example E > with the single quote will cause characters after that single quote C > to be unaltered expecting to find a matching quote BEFORE hitting  > the end of the string.  0 It'd be interesting to test the results of this:  . A$ = "This here's a ""quoted literal"" string" Print EDIT$( a$, 254 ) Print EDIT$( a$, 510 )  F Has double-quotes and apostrophe. Should prove amusing, but can't testH it right now (the little Alpha is set up for other stuff - don't want to break it down just yet).   D.J.D.   ------------------------------  % Date: Wed, 22 Sep 2004 02:22:55 +0000 7 From: David B Sneddon - bigpond <dbsneddon@bigpond.com>  Subject: Re: BASIC EDIT$ - 256* Message-ID: <4150E1FF.4050403@bigpond.com>  & David J Dachtera mentioned in passing:H >>BASIC allows strings to be quoted with either single or double quotes. >  > 5 > Hhmmm... YAUF (Yet Another Undocumented "Feature").  Not quite...   1.5.2 String Constants  I String constants are either string literals or named constants. A string  N literal is a series of characters enclosed in string delimiters. Valid string  delimiters are as follows:  &      * Double quotation marks ("text")&      * Single quotation marks ('text')  O You can embed double quotation marks within single quotation marks ('this is a  K "text" string') and vice versa ("this is a 'text' string"). Note, however,  L that BASIC does not accept incorrectly paired quotation marks and that only O the outer quotation marks must be paired. For example, the following character   strings are valid:   >  > E >>Since the string is parsed from left to right, the original example E >>with the single quote will cause characters after that single quote C >>to be unaltered expecting to find a matching quote BEFORE hitting  >>the end of the string. >  > 2 > It'd be interesting to test the results of this: > 0 > A$ = "This here's a ""quoted literal"" string" > Print EDIT$( a$, 254 ) > Print EDIT$( a$, 510 ) > H > Has double-quotes and apostrophe. Should prove amusing, but can't testJ > it right now (the little Alpha is set up for other stuff - don't want to > break it down just yet). >  > D.J.D. >    D.Sneddon> type b.bas  program test7          A$ = "This here's a ""quoted literal"" string"           Print EDIT$( a$, 254 )           Print EDIT$( a$, 510 )  end program  D.Sneddon> basic b+ Diagnostic on source line 2, listing line 2           source file: B.BAS;1   ?                  A$ = "This here's a ""quoted literal"" string" % ....................................1   : %E, 1:        found string constant when expecting one of:5                                           an operator :                                           end of statement  8 %E, B.BAS;1 completed with 1 diagnostic - object deleted   D.Sneddon> type a.bas  program test?          A$ = "This here's a " + '"quoted literal"' + " string"           Print EDIT$( a$, 254 )           Print EDIT$( a$, 510 )  end program  D.Sneddon> basic a D.Sneddon> link a  D.Sneddon> run a  THISHERE'SA"QUOTEDLITERAL"STRING$ THISHERE's a "quoted literal" string     Regards, Dave.  --  I David B Sneddon (dbs)    VMS Systems Programmer     dbsneddon@bigpond.com I Sneddo's quick guide ...          http://www.users.bigpond.com/dbsneddon/ I DBS freeware at ...   http://www.users.bigpond.com/dbsneddon/software.htm I "Life is what happens to you while you're busy making other plans" Lennon    ------------------------------   Date: 19 Sep 2004 06:53:37 GMT# From: " paris@" <gzgimsun@163.net>  Subject: China girl in Paris2 Message-ID: <cijadh$pe2504@imsp212.netvigator.com>  Z <HTML><HEAD><META http-equiv=3DContent-Type content=3D"text/html; charset=3Dwindows-1252">8 <STYLE></STYLE></HEAD><FONT face=3DArial><FONT size=3D2>a <BODY><DIV>Send email to gzgimsun@163.net or Contact me +8613802741877 Holland Kin.</DIV><BR><BR>    ------------------------------  % Date: Tue, 21 Sep 2004 15:19:32 -0500 ( From: brandon@dalsemi.com (John Brandon)1 Subject: Re: Completed batch job still executing? 1 Message-ID: <04092115193223@dscis6-0.dalsemi.com>    Chris Townley writes:  >  ...A > If I delete the entry it moves to aborting, and stays there - a  > further delet/entry returns * > %DELETE-E-NOTDELETED, error deleting 104E > -JBC-E-INCOMPLETE, previous incomplete operation prevents execution - > so I have to use stop/queue/reset to clear.  >  ...  ; I have encountered the same problem - with Version V7.2-1h1   < STOP/QUE/RESET worked to resolve it - which you already know   I have not looked into further.      J*o*h*n B*r*a*n*d*o*n  VMS Systems Administrator * firstname.lastname.spam.me.not@dalsemi.com   ------------------------------    Date: 22 Sep 2004 03:00:02 -00003 From: lcs Mixmaster Remailer <mix@anon.lcs.mit.edu>   Subject: Dapper Dan does a Kerry6 Message-ID: <20040922030002.23546.qmail@nym.alias.net>   flip-flop flip-flop    September 10  H "This report was not based solely on recovered documents, but rather on G a preponderance of evidence, including documents that were provided by  C unimpeachable sources, interviews with former Texas National Guard  E officials and individuals who worked closely back in the early 1970s  = with Colonel Kerry Killian and were well acquainted with his  - procedures, his character and his thinking,"    	 yesterday   D NEW YORK (AP) - At the behest of a CBS producer, an adviser to John A Kerry said he talked to a central figure in the controversy over  F President Bush's National Guard service shortly before the release of ; documents whose authenticity CBS said it could not verify.  F   Joe Lockhart said he made the call to retired Lt. Col. Bill Burkett C of the Texas National Guard at the suggestion of CBS producer Mary   Mapes.    ? Dan Rather said that he no longer had the "confidence in these  = documents that would allow CBS to continue vouching for them  B journalistically." He went on to apologize: "We made a mistake in H judgment, and for that I am sorry." Unfortunately, Mr. Rather's apology 5 rightly could be described as "too little too late."    ( ========================================B Just a teeny little mistake.  Moron wont admit they are forgeries.? Unimpeachable?  What balls.  Rather likes to tell big fat lies. B To quote Teresa Heinz-Kerry these scumbugs are liars.  Idiots too.: Another Teresa favorite.  But not unexpected after all sheH suffers from Moore's disease.  They can't even run a conspiracy well and- you want to vote for them to run the country?  That makes YOU the Idiot.    ------------------------------  % Date: Tue, 21 Sep 2004 23:26:54 +0200 2 From: Ondrej Brablc <news2004@brablc-removeme.com>E Subject: Does anyone use Compaq C++ V6.5 on Alpha seriously with STL? 2 Message-ID: <ciq6as$4i5$2@ulysses.news.tiscali.de>  	 Hi there,   D I started using Standard C++ Library havily on our OpenVMS 7.3 with  compiler C++ V6.5-004.  H I have started previous week on this project and spent 2 days debugging F compiler. I'm pretty confident that I have found related to templates F and occuring during link time and another one related to dynamic_cast H and appearing during runtime. I know compiler is always the first to be H blamed for programmer's wrong pointer arithmetics, but I can illustrate " the problems with simple examples.  H The nature of those errors brought the question written in the subject. G Is there someone working with dynamic_cast and templates and satisfied   with this compiler?   G I have send the bugs to our system operators, but this could take ages  G until I have some feedback. So I hope, someone in this group does some   C++ programming.   Ondrej   ------------------------------    Date: 21 Sep 2004 21:09:41 -07001 From: usenet_cpp@lehrerfamily.com (Joshua Lehrer) I Subject: Re: Does anyone use Compaq C++ V6.5 on Alpha seriously with STL? = Message-ID: <31c49f0d.0409212009.320dc707@posting.google.com>    Ondrej,   @ I use the C++ compiler (6.5-???) on 7.3-2 extensively.  I'm also5 somewhat of a C++ expert (ok, advanced-intermediate).   E Please post an explanation of your problem, including sample code, as F well as any and all compiler/linker flags/switches that you are using.E  Please try to trim your sample(s) down to as little code required to C demonstrate the problem.  This will make it easier for us to debug.   B I know there was a dynamic_cast<> bug in previous revisions of theF compiler (I'm on 6.5-041) that had to do with shared libraries.  ThereC are also some very subtle strict-ansi compliant template rules that > this compiler gets correct and many other compilers get wrong.  
 joshua lehrer  factset research systems NYSE:FDS   ------------------------------    Date: 21 Sep 2004 11:53:22 -07002 From: c00kiemonster69@hotmail.com (Cookie Monster)  Subject: Mapping UNIX File Names= Message-ID: <6da13cb0.0409211053.220383ff@posting.google.com>   F A file on a unix file system looks like "aCaseSENSITIVEFilename" would? map to "A$C$ASE$SENSITIVEF$ILENAME" on the Multinet OpenVMS NFS  server.   F I'm sure there is a program to convert the files names back and forth,D but the only one I can find is in fortran and I don't have a fortran@ compiler. I have compaq C (I guess that would be HP C now) V6.2.: Any one know of a C program or perl script that does this?   Thanks   ------------------------------    Date: 21 Sep 2004 16:05:28 -0700, From: JimStrehlow@data911.com (Jim Strehlow)$ Subject: Re: Mapping UNIX File Names= Message-ID: <4b6ec350.0409211505.71e2bc38@posting.google.com>   , c00kiemonster69@hotmail.com (Cookie Monster)H > A file on a unix file system looks like "aCaseSENSITIVEFilename" wouldA > map to "A$C$ASE$SENSITIVEF$ILENAME" on the Multinet OpenVMS NFS  > server...  >  > Thanks  ; Knowing how that occurs, we intentionally have designed our ; application to write either  ALL_UPPERCASE_FILENAMES   or    all_lowercase_filenames   B Oracle on another operating system platform (e.g. Windows or Unix)B writes a file to a NFS directory which is scanned and processed by jobs running on OpenVMS.  C Do you have any control over the programming or specifications to a < third-party application in order to change to a single case?  B Jim Strehlow, Data911 Law Enforcement Systems and Database Manager Alameda, CA, USA   ------------------------------    Date: 21 Sep 2004 13:28:49 -0500; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) ( Subject: Re: Mixed Cluster certification3 Message-ID: <tcrZJcKjG5v5@eisner.encompasserve.org>   c In article <cipp89$hb0$1@hercules.btinternet.com>, "Robert H" <robert.heyes@btinternet.com> writes: M > Can anyone direct me to a document which highlights what is possible for a  ) > mixed cluster, or tell me if they know.   F    You want the SPD.  If you can't find them on HP's VMS web site then$    grap the FAQ and it'll point you.   ------------------------------  % Date: Tue, 21 Sep 2004 21:00:33 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net>% Subject: Re: Off-the-wall CI Question + Message-ID: <4150DCC1.E8ECA795@comcast.net>    Keith Parris wrote:  >  > David J Dachtera wrote: ? > > I want to try connecting a CIPCA directly to an HSJ with no K > > intervening Star Coupler: that is, the TX of both channels go to the RX . > > ports of the other device, and vice-versa. > A > Bob is right: this overdrives the receivers and is unsupported.   ? O.k., so what would I need to pad that down to a usuable level?    > [snip]? > I'm sure this is freeing up CI hardware all over the country.   F Not likely. Existing clusters tend to stay that way (the SysAdmins gotE laid-off years ago, and they've been running without a reboot since - 1 clueless Borg drones don't like to mess with it).   	 > I see a J > star coupler on eBay right now for $300. I suspect you could call aroundH > and find any number of customers in the Chicago area who have old star< > couplers and cables they would probably lend you for free. > H > Alternatively, could you MSCP-serve the storage during the transition?  C No. The disk is all SAN/FC. Only the tape silo (TL826) is still CI.    D.J.D.   ------------------------------  % Date: Tue, 21 Sep 2004 22:18:18 -0400 # From: "John Smith" <a@nonymous.com> Y Subject: Re: OpenVMS Pearl -  Indian Railways Celebrates 10th anniversary using  HP Techn , Message-ID: <jrudneV8g5t2fc3cRVn-rw@igs.net>   JF Mezei wrote:  > Sue Skonetski wrote: >>C >> Congratulations to the Indian Railways who celebrated their 10th C >> anniversary of their online reservation system using OpenVMS and % >> Reliable Transaction Router (RTR).  > C > Sue, great story. Just a comment though. The title should contain  > "VMS" instead of "HP". > H > Since HP inherited VMS less than 10 years ago (and hasn't really takenE > ownership of it), my first interpretation of the titke was that the F > indian railways have been HP custoemrs for 10 years and as a result,  > would be running HP-UX of MPE.     How about the tag lines:- "Engineered by Digital Equipment Corporation" 2 "Lack of Marketing by Hewlett Packard Corporation"   ------------------------------    Date: 21 Sep 2004 17:28:13 -07001 From: susan_skonetski@hotmail.com (Sue Skonetski) Y Subject: OpenVMS Pearl -  Indian Railways Celebrates 10th anniversary using HP Technology = Message-ID: <857e9e41.0409211628.4037779f@posting.google.com>   @ Congratulations to the Indian Railways who celebrated their 10th@ anniversary of their online reservation system using OpenVMS and@ Reliable Transaction Router (RTR). Indian railways is the worlds> second largest railway and  annually transports over 4 billionD passengers and 492 million tons of freight.  For more details on the> Indian Railways please see their HP testimonial on the web at:4 http://h71000.www7.hp.com/openvms/brochures/indiarr/  2 The following is from the customer: (bold is mine)  = It is 10 years back we implemented the country wide passenger @ reservation systems. It was running on VAX and recently moved toE ALPHA. This is one of the largest distributed databases in the world.   ! It is an achievement of Open VMS.   ; 22 Sept 2004 marks the completion of 10 years of Successful 2 implementation of the online PRS system --CONCERT.  
 Warm Regards,  Sue    ------------------------------  % Date: Tue, 21 Sep 2004 21:35:32 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> Y Subject: Re: OpenVMS Pearl -  Indian Railways Celebrates 10th anniversary using HP Techno , Message-ID: <4150D6CC.5F8EE736@teksavvy.com>   Sue Skonetski wrote: > B > Congratulations to the Indian Railways who celebrated their 10thB > anniversary of their online reservation system using OpenVMS and$ > Reliable Transaction Router (RTR).  G Sue, great story. Just a comment though. The title should contain "VMS"  instead of "HP".  F Since HP inherited VMS less than 10 years ago (and hasn't really takenJ ownership of it), my first interpretation of the titke was that the indianN railways have been HP custoemrs for 10 years and as a result, would be running
 HP-UX of MPE.    ------------------------------  % Date: Tue, 21 Sep 2004 19:27:54 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> / Subject: OT: Sun admits mistake of droping 8086 , Message-ID: <4150B8E9.3FA66020@teksavvy.com>  M Excellent article on Sun doing a mea-culpa for not focusing on what customers J wanted and how its is now coming back strong with what customer want: 8086 based servers.  U > http://news.com.com/Sun+Weve+turned+over+a+new+leaf/2100-1010_3-5375931.html?tag=nl   L I think that David Dachtera will get a big kick out of this admission by SunJ of how important having your OS on the 8086 is. And maybe if Sun admits toG making mistakes, perhaps Carly could admit to the monumental mistake of M focusing on the failed IA64, the monumental mistake of choosing to ignore VMS  etc etc.   ------------------------------  % Date: Tue, 21 Sep 2004 19:50:26 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> 3 Subject: Re: OT: Sun admits mistake of droping 8086 , Message-ID: <4150BE30.FD7E3CF9@teksavvy.com>   Another interesting article at: I >http://news.com.com/Suns+fighting+chance/2010-1010_3-5374570.html?tag=nl    ##M Step No. 4: Play up the OS-plus-platform advantage. Sun is playing a very old M game here, but it will play it hard. The company is saying that you cannot be F a legitimate, long-term player without controlling and harmonizing theK operating system and the platform. You must have control over both to offer 5 easy and cost-effective solutions for your customer.    L Hewlett-Packard is letting HP-UX die in favor of Red Hat and Windows; IBM isM introducing Power systems that don't run AIX; and Dell never had an operating  system. I Watch for Microsoft and Sun to concoct some tough frontal attacks on IBM, H their avowed common enemy.  Sun makes the claim that it will be the onlyN vendor with a strong platform--Sparc at the high end, x86 at the low end--thatL also has a strong operating system to offer with Solaris at the high and low ends.  ##  N So, HP wants to deemphasize its own products in favour of Microsoft and Linux,D none of which generate revenus to HP. Sun wants to capitalise fo the2 copetitive edge that having your own OS gives you.   ------------------------------  % Date: Tue, 21 Sep 2004 14:53:42 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> $ Subject: Re: Program to create users, Message-ID: <415078B5.1D9EB3EA@teksavvy.com>   Javier Coego wrote: B >> - I need to create a program in C/C++ language which performs aE > massive user creation. - These users are supplied via a text file.    G With DCL,. just open the text file, read record by record, and for each H record, issue an MC AUTHORIZE CREATE username/UIC=[x,y]/PASSWORD/etc/etcM command, then issue the CREATE/DIR/OWNER=username/versions=2 $disk:[username]   M Your DCL will have to manually generate the UICs, probably starting with 1 in M group XX and making sure the user portion of UIC doesn't go beyond the limit, ? and if so, increase group by 1 and then restart at 1 for user).    ------------------------------  % Date: Tue, 21 Sep 2004 18:19:15 -0400 ( From: David Froble <davef@tsoft-inc.com>$ Subject: Re: Program to create users, Message-ID: <4150A8E3.8020303@tsoft-inc.com>   Javier Coego wrote:    > - Hi, everyone. -  > A > - I need to create a program in C/C++ language which performs a F > massive user creation. - These users are supplied via a text file. -     Define 'massive'.     G > - Has anyone tried to write a similar program? - I've taken a look to H > pwd.h functions, but all of them only allow to GET user information. - > : > - This program could be also written in DCL, perl, ... - >  > - Thanks in advance. - >  > - Kind regards. -  >  > Javi >   8 It's best if you let AUTHORIZE create the user accounts.  Q You can take the data from your text file and create one or more text files with  + commands to run AUTHORIZE.  Something like:    $ RUN SYS$SYSTEM:AUTHORIZE% ADD ????? /PASS=???? /UIC=???? ...... % ADD ????? /PASS=???? /UIC=???? ......  EXIT  L The file(s) then would be spawned as subprocesses, thus having the required ' priviledges that your user account has.    Dave   --  4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com 170 Grimplin Road  Vanderbilt, PA  15486    ------------------------------  # Date: Tue, 21 Sep 2004 17:53:07 GMT % From: "Paul Jones" <pjones@i-s-e.com> 1 Subject: Readable error log while process running < Message-ID: <7UZ3d.23950$Mx5.786@newssvr29.news.prodigy.com>  C Is there a trick to making a program's error log readable while the L executable is running when using the vms run statement in DCL? I'm using the format:    $       run program_xyz.exe-+               /error=program_xyz_error.log-                /output=_nla0:-                /input=_nla0:   B I know the VMS SYS$INPUT is equivalent to C's  stdin descriptor 0,L SYS$OUTPUT is equivalent to stdout descriptor 1, and SYS$ERROR is equivalentL stderr descriptor 2. At DCL level, I would like to be able to open the errorF log with a process write-lock and allow a world read privilege without& having to do this at the C code level.   Thanks in advance.
 Paul Jones   ------------------------------    Date: 21 Sep 2004 13:31:43 -0500; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) 5 Subject: Re: Readable error log while process running 3 Message-ID: <BEHfxX$4hw6F@eisner.encompasserve.org>   d In article <7UZ3d.23950$Mx5.786@newssvr29.news.prodigy.com>, "Paul Jones" <pjones@i-s-e.com> writes:E > Is there a trick to making a program's error log readable while the N > executable is running when using the vms run statement in DCL? I'm using the	 > format:  >  > $       run program_xyz.exe-- >               /error=program_xyz_error.log-  >               /output=_nla0:-  >               /input=_nla0:   C    VMS opens files for exclusive access unless specifically told to C    do otherwize.  Run doesn't tell it otherwize.  Sometimes you can E    get what you want via backup/ignore=interlock, but the tail of the '    file will almost certainly be stale.   F    Other solutions are to run the program in batch, allowing the errorF    output to go to the batch log file which is opened for shared read,G    or modify your program to write to a file which is explicitly opened     for shared read.    ------------------------------    Date: 21 Sep 2004 17:25:54 -0700- From: elementyl@hotmail.com (James Wilkinson) 5 Subject: Re: Readable error log while process running = Message-ID: <e2ac4c66.0409211625.73ac8006@posting.google.com>   ; koehler@eisner.nospam.encompasserve.org (Bob Koehler) wrote H >    Other solutions are to run the program in batch, allowing the errorH >    output to go to the batch log file which is opened for shared read,I >    or modify your program to write to a file which is explicitly opened  >    for shared read.   C Or stay with a detached process, run SYS$SYSTEM:LOGINOUT and use an A input command file to open a log file in the appropriate mode and > re-assign SYS$ERROR to use that file before running the target program.   James    ------------------------------  % Date: Tue, 21 Sep 2004 13:42:23 -0400 ) From: "Scott Greig" <jsgreig@geminaq.com> < Subject: Saveset File that OpenVMS ZIP V2.3 will not archive9 Message-ID: <KFZ3d.13389$pA.901273@news20.bellglobal.com>   
 Hello all:  6 I've tried sending this to Zip-Bugs@lists.wku.edu, but* receive only "User unknown" respone, so...  5 I have been using ZIP to compress many of my savesets 6 for transmission to remote systems.  Of five savesets,0 one (and only one) is not being archived by ZIP.
 This file is:    OpenVMS Version 7.3-2  ZIP Version 2.3 2 Source file: OpenVMS backup saveset named APPS.BCK-              Fixed length, 32256 byte records -              ODS-5 disk, size= 4503996 blocks   3 My guess (fwiw) is that the content of the APPS.BCK ' file must be confusing the ZIP program.    An except follows:  ( $ dir disk$other:[backups]apps.bck;/full   Directory DISK$OTHER:[BACKUPS]   APPS.BCK;2 File ID: (3074,1,0)" Size: 4503996/4504012 Owner: [1,1]  Created: 21-SEP-2004 02:04:12.65$ Revised: 21-SEP-2004 10:21:09.83 (3) Expires: <None specified>  Backup: <No backup recorded> Effective: <None specified>  Recording: <None specified>  Accessed: <None specified> Attributes: <None specified> Modified: <None specified> Linkcount: 1 File organization: Sequential  Shelved state: Online  Caching attribute: No_caching G File attributes: Allocation: 4504012, Extend: 0, Global buffer count: 0 !                  No version limit . Record format: Fixed length 32256 byte records Record attributes: None  RMS attributes: None Journaling enabled: None> File protection: System:RWED, Owner:RWED, Group:RWED, World:RE Access Cntrl List: None  Client attributes: None ( Total of 1 file, 4503996/4504012 blocks.   $ set def disk$stuff:[zips]  $ dir apps.zip! %DIRECT-W-NOFILES, no files found  $ show dev disk$stuff:B Device               Device   Error    Volume      Free  Trans MntB  Name                Status   Count     Label     Blocks Count CntB $1$DGA500:  (hostnm) Mounted      0  STUFF     138387125     1   2  3 $ zip -j "-V" apps.zip disk$other:[backups]apps.bck  adding: APPS.BCK (stored 0%) $ dir/fu apps.zip    Directory DISK$STUFF:[ZIPS]    APPS.ZIP;1 File ID: (16,15,0)  Size: 1/137 Owner: [1,1]  Created: 21-SEP-2004 10:27:07.93$ Revised: 21-SEP-2004 10:27:07.94 (3) Expires: <None specified>  Backup: <No backup recorded> Effective: <None specified>  Recording: <None specified>  Accessed: <None specified> Attributes: <None specified> Modified: <None specified> Linkcount: 1 File organization: Sequential  Shelved state: Online  Caching attribute: Writethrough C File attributes: Allocation: 137, Extend: 0, Global buffer count: 0 !                  No version limit , Record format: Fixed length 512 byte records Record attributes: None  RMS attributes: None Journaling enabled: None> File protection: System:RWED, Owner:RWED, Group:RWED, World:RE Access Cntrl List: None  Client attributes: None  Total of 1 file, 1/137 blocks.  3 All other savesets appear to be archived correctly.   3 When sent to the remote node, they UNZIP just fine.   # Any help would be much appreciated.    Scott Greig    ------------------------------  % Date: Tue, 21 Sep 2004 20:33:57 +0200 0 From: Keith Cayemberg <keith.cayemberg@arcor.de>@ Subject: Re: Saveset File that OpenVMS ZIP V2.3 will not archiveB Message-ID: <41507416$0$18554$9b4e6d93@newsread2.arcor-online.net>   Scott Greig wrote:   > Hello all: > 8 > I've tried sending this to Zip-Bugs@lists.wku.edu, but, > receive only "User unknown" respone, so... > 7 > I have been using ZIP to compress many of my savesets 8 > for transmission to remote systems.  Of five savesets,2 > one (and only one) is not being archived by ZIP. > This file is:  >  > OpenVMS Version 7.3-2  > ZIP Version 2.3 4 > Source file: OpenVMS backup saveset named APPS.BCK/ >              Fixed length, 32256 byte records / >              ODS-5 disk, size= 4503996 blocks  >  <SNIP> > % > Any help would be much appreciated.  > 
 > Scott Greig  >  >   	 Hi Scott,   H if I remember correctly, there is/was an issue with Zip (or Unzip) when < attempting to zip a file larger than 2GB (on 32bit Systems).   $ cb 4503996  !convert_blocks   < 4503996 Blocks equals 2199.21 Mebibytes or 2251.99 Megabytes     For more info please see... ' http://www.info-zip.org/FAQ.html#limits  and + http://www.info-zip.org/FAQ.html#corruption   2 for contact info for mailing lists and bugs see...& http://www.info-zip.org/FAQ.html#lists   Cheers anyway!   Keith Cayemberg    ------------------------------  + Date: Tue, 21 Sep 2004 14:33:12 -0500 (CDT)  From: sms@antinode.org@ Subject: Re: Saveset File that OpenVMS ZIP V2.3 will not archive) Message-ID: <04092114331256@antinode.org>   0 From: Keith Cayemberg <keith.cayemberg@arcor.de>   > Scott Greig wrote:  	 > > [...]  > > OpenVMS Version 7.3-2V > [...]pJ > if I remember correctly, there is/was an issue with Zip (or Unzip) when > > attempting to zip a file larger than 2GB (on 32bit Systems). > [...]   G    One might argue that VMS V7.3-2 does not run on a 32-bit system.  InrG any case, I've been told that large file support is among the hot items F in Info-ZIP develpoment these days.  (From which, one might reasonably infer that it's not there yet.)l       According to my DECC$TYPES.H:   [...]h   /*I ** On some Alpha OpenVMS systems support for files larger than 2 gigabyte L ** is available. This feature is enabled by specifying the MACRO _LARGEFILE.L ** When enabled all file offsets are accepted and returned as 64-bit values. **H ** This feature requires a version of the Compaq C compiler with support ** for #pragma __extern_prefix.i **   [...]M  P #       error " Support for large files not available before OpenVMS Alpha V7.2"   [...]a  H    Of course, if the Zip + UnZip code is not written to accomodate largeG files, merely defining _LARGEFILE at compile time is not likely to lead " to a significantly better outcome.  B   I assume that someone at HP is going through all these files and0 changing "Alpha" to "non-VAX" where appropriate.  H ------------------------------------------------------------------------  4    Steven M. Schweda               (+1) 651-699-98183    382 South Warwick Street        sms@antinode-orgt    Saint Paul  MN  55105-2547a   ------------------------------    Date: 21 Sep 2004 15:42:56 -07001 From: nothome@spammers.are.scum (Malcolm Dunnett) @ Subject: Re: Saveset File that OpenVMS ZIP V2.3 will not archive- Message-ID: <O0aL4D8IW1M0@malvm7.mala.bc.ca.>e  e In article <KFZ3d.13389$pA.901273@news20.bellglobal.com>, "Scott Greig" <jsgreig@geminaq.com> writes:s > Hello all: > 8 > I've tried sending this to Zip-Bugs@lists.wku.edu, but, > receive only "User unknown" respone, so... > 7 > I have been using ZIP to compress many of my savesets 8 > for transmission to remote systems.  Of five savesets,2 > one (and only one) is not being archived by ZIP. > This file is:n >  > OpenVMS Version 7.3-2e > ZIP Version 2.3t4 > Source file: OpenVMS backup saveset named APPS.BCK/ >              Fixed length, 32256 byte records-/ >              ODS-5 disk, size= 4503996 blocks  > 5 > My guess (fwiw) is that the content of the APPS.BCK ) > file must be confusing the ZIP program.i >    <SNIP> > 5 > $ zip -j "-V" apps.zip disk$other:[backups]apps.bck  > adding: APPS.BCK (stored 0%)  N    I've seen the same thing, I suspect ( as another poster suggested ) that itN has to do with the file being more than 2GB (ie a signed longword somewhere is overflowing ).  O    I've found that if one doesn't use the "-V" the file zips ok. This does meanaB that after unzipping the file at the remote end one needs to do a:  1 $SET FILE/ATTR=(rfm:fix,lrl:<blocksize>) file.BCKe  A   Where <blocksize> is the blocksize used in the original backup.    ------------------------------  % Date: Tue, 21 Sep 2004 22:10:39 -0500m2 From: David J Dachtera <djesys.nospam@comcast.net>@ Subject: Re: Saveset File that OpenVMS ZIP V2.3 will not archive+ Message-ID: <4150ED2F.71D6D4A4@comcast.net>    Scott Greig wrote: >  > Hello all: > 8 > I've tried sending this to Zip-Bugs@lists.wku.edu, but, > receive only "User unknown" respone, so... > 7 > I have been using ZIP to compress many of my savesets 8 > for transmission to remote systems.  Of five savesets,2 > one (and only one) is not being archived by ZIP. > This file is:5 >  > OpenVMS Version 7.3-2  > ZIP Version 2.3r4 > Source file: OpenVMS backup saveset named APPS.BCK/ >              Fixed length, 32256 byte records / >              ODS-5 disk, size= 4503996 blocks  > 5 > My guess (fwiw) is that the content of the APPS.BCKo) > file must be confusing the ZIP program.n >  > An except follows: > * > $ dir disk$other:[backups]apps.bck;/full >   > Directory DISK$OTHER:[BACKUPS] >   > APPS.BCK;2 File ID: (3074,1,0)$ > Size: 4503996/4504012 Owner: [1,1]" > Created: 21-SEP-2004 02:04:12.65& > Revised: 21-SEP-2004 10:21:09.83 (3) > Expires: <None specified>h > Backup: <No backup recorded> > Effective: <None specified>6 > Recording: <None specified>0 > Accessed: <None specified> > Attributes: <None specified> > Modified: <None specified> > Linkcount: 1 > File organization: Sequential0 > Shelved state: Online  > Caching attribute: No_caching I > File attributes: Allocation: 4504012, Extend: 0, Global buffer count: 0e# >                  No version limitt0 > Record format: Fixed length 32256 byte records > Record attributes: Nonel > RMS attributes: None > Journaling enabled: None@ > File protection: System:RWED, Owner:RWED, Group:RWED, World:RE > Access Cntrl List: None4 > Client attributes: NoneE* > Total of 1 file, 4503996/4504012 blocks. >  > $ set def disk$stuff:[zips]U > $ dir apps.zip# > %DIRECT-W-NOFILES, no files found  > $ show dev disk$stuff:D > Device               Device   Error    Volume      Free  Trans MntD >  Name                Status   Count     Label     Blocks Count CntD > $1$DGA500:  (hostnm) Mounted      0  STUFF     138387125     1   2 > 5 > $ zip -j "-V" apps.zip disk$other:[backups]apps.bckv > adding: APPS.BCK (stored 0%) > $ dir/fu apps.zipo >  > Directory DISK$STUFF:[ZIPS]  >  > APPS.ZIP;1 File ID: (16,15,0)i > Size: 1/137 Owner: [1,1]" > Created: 21-SEP-2004 10:27:07.93& > Revised: 21-SEP-2004 10:27:07.94 (3) > Expires: <None specified>t > Backup: <No backup recorded> > Effective: <None specified>u > Recording: <None specified>  > Accessed: <None specified> > Attributes: <None specified> > Modified: <None specified> > Linkcount: 1 > File organization: Sequentialu > Shelved state: Onlines! > Caching attribute: Writethrough"E > File attributes: Allocation: 137, Extend: 0, Global buffer count: 0n# >                  No version limitu. > Record format: Fixed length 512 byte records > Record attributes: Nonen > RMS attributes: None > Journaling enabled: None@ > File protection: System:RWED, Owner:RWED, Group:RWED, World:RE > Access Cntrl List: None  > Client attributes: Noned  > Total of 1 file, 1/137 blocks. > 5 > All other savesets appear to be archived correctly.  > 5 > When sent to the remote node, they UNZIP just fine.e > % > Any help would be much appreciated..  G Well, it would help to see what error messages you're getting from ZIP..  F However, as others have suggested, it's likely the 2GB file size limit8 again. Try breaking up the BACKUP into smaller savesets.  D Note, also, there is now some question about ZIP V2.3 being entirely	 "stable".    D.J.D.   ------------------------------  % Date: Tue, 21 Sep 2004 21:42:47 -0400o# From: "John Smith" <a@nonymous.com>t/ Subject: Re: Sun admits mistake of droping 8086s, Message-ID: <Mt6dnRsgSNvAQM3cRVn-gw@igs.net>   JF Mezei wrote:,E > Excellent article on Sun doing a mea-culpa for not focusing on whathB > customers wanted and how its is now coming back strong with what$ > customer want: 8086 based servers. >: >>L http://news.com.com/Sun+Weve+turned+over+a+new+leaf/2100-1010_3-5375931.html ?tag=nlt >sG > I think that David Dachtera will get a big kick out of this admissioneE > by Sun of how important having your OS on the 8086 is. And maybe ifeA > Sun admits to making mistakes, perhaps Carly could admit to the:C > monumental mistake of focusing on the failed IA64, the monumental(, > mistake of choosing to ignore VMS etc etc.    G She'll never do it, nor will HP's lapdog Bored of Directors....too muchg hubris.h   ------------------------------  % Date: Tue, 21 Sep 2004 21:21:02 +0100u. From: "Jim Brankin" <brankin at nildram.co.uk>) Subject: Re: TCP/IP cluster interconnect?e; Message-ID: <41508ce4$0$52214$65c69314@mercury.nildram.net>5   John,        Thanks for the reply.       L Why are you so against TCP? It would provide end-to-end error correction andF flow control. It would request the retransmission of lost or erroneous. packets and present a reliable network to SCS.      J My basic point is that we are moving to an IP-centric world. The level-twoC networks we use today to carry the cluster interconnect will not be F available. The recent discussion here regarding sending SCS over an IPA tunnel suggests that some people have already entered this world.t      H BTW I am not suggesting sending the cluster interconnect over the publicH internet. Obviously you would need some control over quality of service.       Jimc            <briggs@encompasserve.org> wroteI > I don't think a sane person would use an SCS <-> TCP/IP gateway.  You'deD > want SCS <-> IP or SCS <-> UDP.  I'll assume that's what you mean. > E > Latency, of course, is not an issue that is peculiar to IP.  If youeE > have a multi-hop delivery path you're going to have per-hop latencyoB > (barring cut-through switching anyway).  If you have congestion,F > you'll have queueing latency (or packet loss, pick your poison).  IfB > you have long distances, you'll have speed-of-light latency.  ItF > doesn't matter in principle whether the delivery path is IP, DECnet, > switched Ethernet or ATM.> >TB > Packet loss is also not peculiar to IP.  It is a function of theC > communications link and, in some cases, the result of congestion.D >DC > Objections to an SCS <=> IP gateway should, accordingly, be aimeddB > at the connectivity underlying the IP network rather than at IP. > F > An IP link from point A to point B with 10 ms latency and negligibleG > packet loss should be just as useful as a switched Ethernet link from2C > point A to point B with 10 ms latency and negligible packet loss.l > @ > While I wouldn't be happy clustering over the Internet, I'd be7 > perfectly happy to cluster over IP over ATM over DS3.g >4
 > John BriggsM   ------------------------------  % Date: Tue, 21 Sep 2004 17:34:23 -0400e- From: JF Mezei <jfmezei.spamnot@teksavvy.com>e) Subject: Re: TCP/IP cluster interconnect?V+ Message-ID: <41509E55.94D38C4@teksavvy.com>i   Jim Brankin wrote:N > Why are you so against TCP? It would provide end-to-end error correction andH > flow control. It would request the retransmission of lost or erroneous0 > packets and present a reliable network to SCS.  M If you have needs for retransmissions, then in all likelyhood, the link isn't # reliable enough to host clustering.0  L > My basic point is that we are moving to an IP-centric world. The level-twoE > networks we use today to carry the cluster interconnect will not ber > available.  N Not necessarily. Telcos are now providing ethernet services within cities, andK it is but a matter of time before this becomes between cities. And with thekL new trend with MPLS (or whatever the acronym is), the network protocol is noF longer so important since the telcos are building protocol-independant1 networks that can carry many different protocols.C  N Remember that switches work mostly on ethernet adresses and not interpretation of network stack addresses.   < Someone here (probably Keith Paris) said it very eloquently:  I If you need disaster tolerance across cities, then your system is missioni' critical enough that you don't want to:lJ 	1- share the cluster link other others who could potential cause problemsG 	2- have less experienced network managers play with the cluster link. nH 	   (the network equyivalent of having a MS weenie manage a VMS cluster)  I Remember that a network manager who insists on IP only probably lacks the I depth of knowledge and experience to manage a critical infrastructure andl@ knows only about a limited set of networks and network hardware.  M I think that the only semi-valid "anti VMS" decision would be one that forcesg@ DECNET to be tunneled in IP so it could use the existing IP onlyU infrastructure (consider that router vendors charge more for decnet routing options).e  C But for clustering, it is important enough to warrant its own link.s    N On the other side of the coin, *IF* VMS is ever going to be targetted at smallK businesses, then perhaps the ability to cluster acorss cheaper shared links K might be very good. Since there are already metropolitain ethernet servicesr5 available fronm telcos, perhaps that would be enough.o   ------------------------------  % Date: Tue, 21 Sep 2004 20:11:13 -0500i2 From: David J Dachtera <djesys.nospam@comcast.net>) Subject: Re: TCP/IP cluster interconnect?p+ Message-ID: <4150D130.B3CB79FF@comcast.net>    Jim Brankin wrote: >  > John,y >  > Thanks for the reply.r > N > Why are you so against TCP? It would provide end-to-end error correction andH > flow control. It would request the retransmission of lost or erroneous0 > packets and present a reliable network to SCS.  E The point is that you'd end up doing a lot of "front end" work (stuffrG that gets set up before VMS even boots) just to get the IP stack up andeD running. ...and it would still be TCP/IP: ubiquitous, but since when# does that mean anything to OpenVMS?i  L > My basic point is that we are moving to an IP-centric world. The level-twoE > networks we use today to carry the cluster interconnect will not beg > available.  B I guess VLANs are out-the-window, too, then, huh? Not likely, IMO.  = > The recent discussion here regarding sending SCS over an IP C > tunnel suggests that some people have already entered this world.h  G Some may have dabbled, perhaps alpha-ed a prototype (rather sounds likeoE Emerald, doesn't it?), but "entered" is likely a mischaracterization.e  J > BTW I am not suggesting sending the cluster interconnect over the publicJ > internet. Obviously you would need some control over quality of service.  C Absolutely. Which begs the question: why compromise quality when ite- would end up being a separate network anyway?e  D No offense intended, but folks who seem to want to pursue it may not fully understand the issue.e   That said, consider this model:i  E A separate cluster node that acts as an SCS "proxy" for remote TCP/IPdE linked cluster nodes. The proxy "acts locally, but thinks globally" -pA trying to keep the remote nodes in sync. with the local ones, andrB ordering CLUEXIT as appropriate when communications have gotten soF botched up that there's no way back. Consider the implications and the answer will soon be obvious.   D.J.D.   ------------------------------  % Date: Tue, 21 Sep 2004 20:16:27 -050032 From: David J Dachtera <djesys.nospam@comcast.net>) Subject: Re: TCP/IP cluster interconnect?m+ Message-ID: <4150D26B.B3989A96@comcast.net>    Bob Koehler wrote: > b > In article <414F7DF7.C54A6E94@comcast.net>, David J Dachtera <djesys.nospam@comcast.net> writes: > >IL > > Well, yes and no. SCS pre-dates the seven-layer model, and was a productE > > separate from VMS until it was more intimately integrated and notp > > installed spearately.  > E >    TCP/IP and DECnet Phase I through IV all predate the seven-layer7J >    model.  IIRC TCP/IP has four laters and DECnet on VMS has 6 (at leastE >    when using FAL), but other DECnet components and DECnet on othermG >    systems typically has 7.  And DECnet was a separate product beforel
 >    VMS 3.0.l  ? Technically, it still is. It's a .PCSI kit on the current AlphaE5 distro.'s (two kits, actually: Phase-V and Phase-IV).   B >    I believe SCS was latent in some versions of VMS prior to itsE >    official first ship, but I don't recall it ever being separatelye >    installable.   G Back around V4.x (late in the V4 series) and early V5. I may still haveaF a VAXcluster software 9-track or TK50 somewhere in my deepest, darkest	 archives..  E >    VMScluster is a technology that networks VMS kernels to create a E >    single tighly bound domain, a very different approach from othere9 >    "clusters" or networks, but a network none the less.   E Well, not as "tightly bound" as SMP, but better than anything else to3@ date. ...and yes, a network - as I said, the point is debatable.   D.J.D.   ------------------------------  % Date: Tue, 21 Sep 2004 22:01:05 -0400e- From: JF Mezei <jfmezei.spamnot@teksavvy.com>h) Subject: Re: TCP/IP cluster interconnect?., Message-ID: <4150DCC8.5958FF31@teksavvy.com>   David J Dachtera wrote:hG > The point is that you'd end up doing a lot of "front end" work (stuff I > that gets set up before VMS even boots) just to get the IP stack up and  > running. .  M Not of you have a network card that has a built-in TCPIP stack in its rom andeQ which woud appear as a simple ethernet or other card to the VMS operating system.:   ------------------------------  % Date: Tue, 21 Sep 2004 22:14:59 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net>) Subject: Re: TCP/IP cluster interconnect?d+ Message-ID: <4150EE32.93104AF8@comcast.net>t   JF Mezei wrote:, >  > David J Dachtera wrote:tI > > The point is that you'd end up doing a lot of "front end" work (stuff K > > that gets set up before VMS even boots) just to get the IP stack up ande > > running. . > O > Not of you have a network card that has a built-in TCPIP stack in its rom andvS > which woud appear as a simple ethernet or other card to the VMS operating system.s   J.F.:d   > David J Dachtera wrote:eI > > The point is that you'd end up doing a lot of "front end" work (stuffs+ > > that gets set up before VMS even boots)   $ Is that not what you just described?   D.J.D.   ------------------------------  % Date: Wed, 22 Sep 2004 00:25:04 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> ) Subject: Re: TCP/IP cluster interconnect?u, Message-ID: <4150FE9E.5415BC92@teksavvy.com>   David J Dachtera wrote:lK > > > The point is that you'd end up doing a lot of "front end" work (stuff-- > > > that gets set up before VMS even boots)8 > & > Is that not what you just described?  J Not sure. The way I had read the above was similar to SCS which gets setup% very early in thge VMS boot sequence.V  H What I described was a puerly hardware based solution with a TCPUP stackN implemented in a hardware card which becomes available as soon as the power is# applied (eg: even in console mode).    ------------------------------    Date: 21 Sep 2004 12:50:06 -0700 From: cljlk@hotmail.com (cljlk)g Subject: user privilegei< Message-ID: <de74637.0409211150.436d965b@posting.google.com>   I am not a VMS sytem person.8 Would anyone please show me how to check user privilege?4 I know "show user" for checking users on the system.  How to check the user privilege?$ And how to set privilege for a user?   Thanks   ------------------------------  # Date: Tue, 21 Sep 2004 19:59:12 GMT % From: Rob Brown <mylastname@gmcl.com>  Subject: Re: user privilege-E Message-ID: <Pine.LNX.4.61.0409211354310.22680@localhost.localdomain>t  ! On Tue, 21 Sep 2004, cljlk wrote:j   > I am not a VMS sytem person.: > Would anyone please show me how to check user privilege?6 > I know "show user" for checking users on the system." > How to check the user privilege?& > And how to set privilege for a user?  ' $ SHOW PROCESS/PRIVILEGE <process name>i   UseV   $ SHOW USER/FULL  C to map username to process name.  They are often the same, but not i always.   G Note that SHOW PROCESS shows active privileges at the instant you type gG the command.  Some users may be able to change their active privileges n from time to time.   What do you need to know?n     -- u  B Rob Brown                        b r o w n a t g m c l d o t c o mA G. Michaels Consulting Ltd.      (866)438-2101 (voice) toll free!e6 Edmonton                         (780)438-9343 (voice)5                                   (780)437-3367 (FAX)a2                                   http://gmcl.com/   ------------------------------  % Date: Tue, 21 Sep 2004 15:53:37 -0500l( From: brandon@dalsemi.com (John Brandon) Subject: Re: user privilege 1 Message-ID: <04092115533782@dscis6-0.dalsemi.com>e   cljlk@hotmail.com writes:  > I am not a VMS sytem person.: > Would anyone please show me how to check user privilege?6 > I know "show user" for checking users on the system." > How to check the user privilege?& > And how to set privilege for a user?  9 Keep in mind that there are multiple areas of priveleges.s  0 1) There is user priveleges & Rights Identifiers "Authorized Privileges:" "Default Privileges:", "Identifier"   $ MCR AUTHORIZE SHOW <user>e  * 2) There is file protection and ownership:( [<owner>]              (<s>,<o>,<g>,<w>)   $ DIRECTORY /SECURITY <file>  ! 3) And still more file protection  ACL "Access Cntrl List:"   $ DIRECTORY /FULL <file>  7 4) Then lest we forget installed images with priveleges  "Open Hdr SharAddr Prv"-   $ INSTALL LIST      9 One more thing.  Setting priveleges for a user?  Don't.  f     J*o*h*n B*r*a*n*d*o*n  VMS Systems Administratori* firstname.lastname.spam.me.not@dalsemi.com   ------------------------------    Date: 21 Sep 2004 17:05:46 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen)d Subject: Re: user privileget3 Message-ID: <9hByGIlxf$DT@eisner.encompasserve.org>   \ In article <04092115533782@dscis6-0.dalsemi.com>, brandon@dalsemi.com (John Brandon) writes:  ; > One more thing.  Setting priveleges for a user?  Don't.     > At least not until you have thoroughly read and understood the Guide to System Security.s   ------------------------------  % Date: Tue, 21 Sep 2004 21:44:37 +0200 2 From: "Robert Boers" <robert.boers@softresint.com>1 Subject: Re: VAX 6000 replacement with CHARON-VAX-7 Message-ID: <415084a8$0$21018$5402220f@news.sunrise.ch>F  ? "David J Dachtera" <djesys.nospam@comcast.net> wrote in message $ news:414CAE5E.2AA116A@comcast.net... snipH > Also, is there a similar adapter for any bus that supports CI or DSSI?C > (Acknowledged: CIPCA, but probably not supported by any PC BIOS).# snip  J For tests, we have connected a hardware DSSI controller via a Qbus adapterJ (made by The Logical Company) to CHARON-VAX emulating a Qbus VAX. Runs outI of the box. This way you can run a (CHARON) VAX DSSI cluster if you want,pI but the original DSSI hardware is very slow compared to current hardware.   B The recently announced VAX 6000 emulator has a direct disk clusterF capability that allows multiple CHARON-VAX nodes to form a similar VMSD cluster using an iSCSI target (using the standard Windows host iSCSIK initiators and a GB Ethernet) and this is an order of magnitude faster thankK with hardware DSSI. The design goal is to shrink a large VAX 6/7000 clusterh= into a single 19" rack with multiple Windows blade servers as8B CHARON-VAX/6000 nodes and an iSCSI target for shared data storage.   Robert   ------------------------------  % Date: Tue, 21 Sep 2004 17:52:33 -0400r( From: David Froble <davef@tsoft-inc.com>1 Subject: Re: VAX 6000 replacement with CHARON-VAXc, Message-ID: <4150A2A1.6050201@tsoft-inc.com>   JF Mezei wrote:o  P > If the charon product emulates a 6000 computer, why would there be any licenceN > transfer fees if you are moving from one 6000 to another 6000 machine within > the same organisation ?E >   J Operating systems are tied to the hardware with which they were sold.  To + transfer the OS license, a fee is required.c  N Now if you were to ask why transfer fees in general exist, you'd be attacking P the problem.  If you buy a copy (license to use) of windoz you can move it from Q one computer to another, as long as it's running on only one computer at a time. uL   DEC tied things up a bit more.  It may not be good, or fair, but it's the  agreement the buyer accepted.i   Dave   --  4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com 170 Grimplin Roade Vanderbilt, PA  15486    ------------------------------  % Date: Tue, 21 Sep 2004 19:59:31 -0500e2 From: David J Dachtera <djesys.nospam@comcast.net>1 Subject: Re: VAX 6000 replacement with CHARON-VAX + Message-ID: <4150CE73.416A2416@comcast.net>e   Robert Boers wrote:a > A > "David J Dachtera" <djesys.nospam@comcast.net> wrote in message & > news:414CAE5E.2AA116A@comcast.net... > snipJ > > Also, is there a similar adapter for any bus that supports CI or DSSI?E > > (Acknowledged: CIPCA, but probably not supported by any PC BIOS).u > snip > L > For tests, we have connected a hardware DSSI controller via a Qbus adapterL > (made by The Logical Company) to CHARON-VAX emulating a Qbus VAX. Runs outK > of the box. This way you can run a (CHARON) VAX DSSI cluster if you want,iK > but the original DSSI hardware is very slow compared to current hardware.A > D > The recently announced VAX 6000 emulator has a direct disk clusterH > capability that allows multiple CHARON-VAX nodes to form a similar VMSF > cluster using an iSCSI target (using the standard Windows host iSCSIM > initiators and a GB Ethernet) and this is an order of magnitude faster than M > with hardware DSSI. The design goal is to shrink a large VAX 6/7000 clusterM? > into a single 19" rack with multiple Windows blade servers as8D > CHARON-VAX/6000 nodes and an iSCSI target for shared data storage.  G What do you have up your sleeve for data migration? (DSSI/CI storage to  DAS/iSCSI/whatever?)   D.J.D.   ------------------------------  + Date: Tue, 21 Sep 2004 18:17:48 +0000 (UTC)g, From: lewis@PROBE.MITRE.ORG (Keith A. Lewis)> Subject: Re: Windoze not rebooted monthly shuts down airports!. Message-ID: <cipr8c$p5t$1@newslocal.mitre.org>   koehler@eisner.nospam.encompasserve.org (Bob Koehler) writes in article <5NXF4e+SI5L+@eisner.encompasserve.org> dated 21 Sep 2004 07:54:28 -0500:aK >   Techies who were working on IBM's (later canceled) 1980's FAA contract oI >   tell me they spent a lot of time trying to emulate VMS paradigms.  AndI >   understanding of the problem showed that what they really needed was  E >   the world's largest VMScluster, much larger geographically and in 9 >   number of nodes than VMScluster supports, even today.-  H I don't think VMScluster would have worked.  ATC facilities need to keepH running when any given part goes offline.  If the cluster communicationsJ link(s) went down, the VMS systems on the small partition would have hung.  0 --Keith Lewis              klewis {at} mitre.org> The above may not (yet) represent the opinions of my employer.   ------------------------------  % Date: Tue, 21 Sep 2004 14:49:13 -0400h- From: JF Mezei <jfmezei.spamnot@teksavvy.com>g> Subject: Re: Windoze not rebooted monthly shuts down airports!, Message-ID: <415077A8.1D030730@teksavvy.com>   Chris Doran wrote:F > This, and the wording in the article as "fix the _program_" suggestsF > that it's not the operating system but a programming bug which would2 > have occurred with _any_ system and/or compiler,  H The problem is that an organisation that chooses Microsoft solutions forN mission critical solutions is also more likely to hire inexperienced microsoftH weenies to do the programming. It is the mentality of organisations thatM choose "cheap trendy solutions" that is the problem. Microsoft is just takinge= advantage of this with purposely unreliable and dangerous OS.    ------------------------------  % Date: Tue, 21 Sep 2004 15:03:07 -0400s- From: JF Mezei <jfmezei.spamnot@teksavvy.com>c> Subject: Re: Windoze not rebooted monthly shuts down airports!, Message-ID: <41507AEA.D0AA7C03@teksavvy.com>   "Keith A. Lewis" wrote:tJ > I don't think VMScluster would have worked.  ATC facilities need to keepJ > running when any given part goes offline.  If the cluster communicationsL > link(s) went down, the VMS systems on the small partition would have hung.  I Each site would need to have its own cluster. And then you have decnet orO! whetever links between each site.t   ------------------------------  % Date: Tue, 21 Sep 2004 21:23:33 +0200t0 From: Keith Cayemberg <keith.cayemberg@arcor.de>> Subject: Re: Windoze not rebooted monthly shuts down airports!B Message-ID: <41507fb6$0$18557$9b4e6d93@newsread2.arcor-online.net>   Keith A. Lewis wrote:V   > koehler@eisner.nospam.encompasserve.org (Bob Koehler) writes in article <5NXF4e+SI5L+@eisner.encompasserve.org> dated 21 Sep 2004 07:54:28 -0500:. > K >>  Techies who were working on IBM's (later canceled) 1980's FAA contract cI >>  tell me they spent a lot of time trying to emulate VMS paradigms.  AnsI >>  understanding of the problem showed that what they really needed was  E >>  the world's largest VMScluster, much larger geographically and in 9 >>  number of nodes than VMScluster supports, even today.t >  > J > I don't think VMScluster would have worked.  ATC facilities need to keepJ > running when any given part goes offline.  If the cluster communicationsL > link(s) went down, the VMS systems on the small partition would have hung. > 2 > --Keith Lewis              klewis {at} mitre.org@ > The above may not (yet) represent the opinions of my employer.  < Any reason why one couldn't have multiple redundant cluster C interconnects (say 4 or 5 for good measure) between all sites with d different routes?d  H With the advent of the Multipath Failover enhancements in OpenVMS 7.3-1 H I believe a real no-single-point-of-failure multi-site cluster has been H worked-out in detail. But, I'm sure there are some cluster experts here 2 who understand the big picture better than myself.  C I think Bob Kohler's point is that VMS Clusters wouldn't (at least sC officially, probably also practically) not support enough multiple  C remote sites and systems to include the entire U.S. nationwide ATC  * distributed system as one OpenVMS cluster.  A However, any other competing distributed system would have their uA limitations in regard to implementing such a massive multi-site, hG concurrent-update single data image ATC system as appears to have been e	 the goal.t  H Although there are also some caveats concerning the reduced suitability B of clusters for hard real-time systems (predictability of service 7 latency would require a stand-alone system), I believe  D VMS-Clustering-based ATC System implemented as multiple (regional?) D clusters would still offer enough answers to the overall ATC system C problem space to provide an exceptional solution using some clever iD designing. I'm sure there would be some VMS clustering experts that B would greatly enjoy sinking their teeth into such a juicy problem.   Cheers!t   Keith Cayemberg>   ------------------------------  % Date: Tue, 21 Sep 2004 16:56:36 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com>e> Subject: Re: Windoze not rebooted monthly shuts down airports!, Message-ID: <4150957C.67CBEFFE@teksavvy.com>   re: clusters for ATC  M One big advantage of VMS clusters for 7/24 operations is the ability to bringaM down a node for servicing without affecting service much. I think that if the:L FAA had ordered VMS for its ATC service, they would have had enough power toJ force Hoff to stay at work a bit longer one night to provide a function toL "manually" force a node to get out of the cluster without having the cluster0 transition cause a few heartbeats to be skipped.  N One of the big problems FAA has had was the inability to gradually upgrade itsI systems over the years since downtime was unacceptable and any project toaE upgrade software/hardware turned out to be massive due to the need toeM transparently switch over. VMS clustering would have enabled smooth evolutionB of systems..  R And one big advantage that isn't talked about is the ability to roll back changes:  J The same way that you can upgrade a node without shutting others down, youK could go back to old software if you realise the new software has problems.s   ------------------------------  + Date: Tue, 21 Sep 2004 20:57:01 +0000 (UTC)u, From: lewis@PROBE.MITRE.ORG (Keith A. Lewis)> Subject: Re: Windoze not rebooted monthly shuts down airports!. Message-ID: <ciq4it$1gk$1@newslocal.mitre.org>   Keith Cayemberg <keith.cayemberg@arcor.de> writes in article <41507fb6$0$18557$9b4e6d93@newsread2.arcor-online.net> dated Tue, 21 Sep 2004 21:23:33 +0200: >Keith A. Lewis wrote: >n >> koehler@eisner.nospam.encompasserve.org (Bob Koehler) writes in article <5NXF4e+SI5L+@eisner.encompasserve.org> dated 21 Sep 2004 07:54:28 -0500: >> /L >>>  Techies who were working on IBM's (later canceled) 1980's FAA contract J >>>  tell me they spent a lot of time trying to emulate VMS paradigms.  AnJ >>>  understanding of the problem showed that what they really needed was F >>>  the world's largest VMScluster, much larger geographically and in: >>>  number of nodes than VMScluster supports, even today. >> P >>  K >> I don't think VMScluster would have worked.  ATC facilities need to keepaK >> running when any given part goes offline.  If the cluster communications M >> link(s) went down, the VMS systems on the small partition would have hung.s  = >Any reason why one couldn't have multiple redundant cluster hD >interconnects (say 4 or 5 for good measure) between all sites with  >different routes?  L It's funny, that was the kind of thinking that caused the program to crash. L Every system "needed" this or that capability, redundancy, etc.  It ended upC being more than even IBM could build.  By the time the contract wasC@ cancelled, it was massively bloated and years behind schedule.    I >With the advent of the Multipath Failover enhancements in OpenVMS 7.3-1 dI >I believe a real no-single-point-of-failure multi-site cluster has been aI >worked-out in detail. But, I'm sure there are some cluster experts here i3 >who understand the big picture better than myself.g >_D >I think Bob Kohler's point is that VMS Clusters wouldn't (at least D >officially, probably also practically) not support enough multiple D >remote sites and systems to include the entire U.S. nationwide ATC + >distributed system as one OpenVMS cluster.n  K The number of sites is far less than you'd think -- if you count only majorn  centers, it's less than 2 dozen.  E But I think you're both mis-applying the strengths of clustering to a-J problem for which the needs are different.  Multi-site clustering providesM geographic redundancy, where if site A goes down, B and C can take the load. s  I In ATC, you "need" the data at all the sites all of the time.  And if you-K lose your connections to the other sites, it's much better to continue withlI stale data than to hang.  Sure, distributed locking is a nice feature forcE the app, but if the OS depends on it, you're just asking for trouble.3  B >However, any other competing distributed system would have their B >limitations in regard to implementing such a massive multi-site, H >concurrent-update single data image ATC system as appears to have been 
 >the goal. > I >Although there are also some caveats concerning the reduced suitability sC >of clusters for hard real-time systems (predictability of service b8 >latency would require a stand-alone system), I believe E >VMS-Clustering-based ATC System implemented as multiple (regional?)  E >clusters would still offer enough answers to the overall ATC system dD >problem space to provide an exceptional solution using some clever E >designing. I'm sure there would be some VMS clustering experts that IC >would greatly enjoy sinking their teeth into such a juicy problem.i  L I have no doubt the DEC engineers would do a great job at such a design, but@ I'm sure what they came up with wouldn't be "one big cluster".    0 --Keith Lewis              klewis {at} mitre.org> The above may not (yet) represent the opinions of my employer.   ------------------------------  % Date: Tue, 21 Sep 2004 17:35:10 -0400n( From: David Froble <davef@tsoft-inc.com>> Subject: Re: Windoze not rebooted monthly shuts down airports!, Message-ID: <41509E8E.4020802@tsoft-inc.com>   Chris Doran wrote:  c > John Reagan <john.reagan@hp.com> wrote in message news:<VRD3d.11172$RN3.7572@news.cpqcorp.net>...: >  >>Bob Ceculski wrote:V >>9 >>>what fools, trying to run an airport on anything otherm >>>than VMS! >>> F >>>http://news.zdnet.co.uk/software/windows/0,39020396,39167074,00.htm >>>tJ >>A co-worker who read another article on the story which gave the actual J >>time before the system would die figured out that the duration is 2**32 H >>milliseconds (or something like that).  He guessed that some longword  >>counter overflows. >> > F > This, and the wording in the article as "fix the _program_" suggestsF > that it's not the operating system but a programming bug which wouldE > have occurred with _any_ system and/or compiler, including most VMS(H > ones, that doesn't trap integer overflow. So regrettably this incident& > isn't necessarily Microsoft's fault. >  > Chrisi >   N A counter to this is that if Microsoft hadn't fostered lesser expectations of P computers, and software, and prompted the use of less competant programmers and L designers, maybe the system would have been better designed and implemented.   Dave   -- p4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com 170 Grimplin Road: Vanderbilt, PA  15486:   ------------------------------  % Date: Tue, 21 Sep 2004 21:12:07 -0400 # From: "John Smith" <a@nonymous.com> > Subject: Re: Windoze not rebooted monthly shuts down airports!, Message-ID: <3p-dndC4f5f1TM3cRVn-sA@igs.net>   Keith A. Lewis wrote:t> > Keith Cayemberg <keith.cayemberg@arcor.de> writes in articleF > <41507fb6$0$18557$9b4e6d93@newsread2.arcor-online.net> dated Tue, 21 > Sep 2004 21:23:33 +0200: >> Keith A. Lewis wrote: >>C >>> koehler@eisner.nospam.encompasserve.org (Bob Koehler) writes iniE >>> article <5NXF4e+SI5L+@eisner.encompasserve.org> dated 21 Sep 2004d >>> 07:54:28 -0500:i >>> C >>>>  Techies who were working on IBM's (later canceled) 1980's FAA E >>>>  contract tell me they spent a lot of time trying to emulate VMS G >>>>  paradigms.  An understanding of the problem showed that what theyVC >>>>  really needed was the world's largest VMScluster, much larger0E >>>>  geographically and in number of nodes than VMScluster supports,S >>>> even today. >>>0 >>> G >>> I don't think VMScluster would have worked.  ATC facilities need to2 >>> keep= >>> running when any given part goes offline.  If the clustera >>> communicationsC >>> link(s) went down, the VMS systems on the small partition wouldd >>> have hung. > > >> Any reason why one couldn't have multiple redundant clusterE >> interconnects (say 4 or 5 for good measure) between all sites withs >> different routes? >iF > It's funny, that was the kind of thinking that caused the program to > crash.E > Every system "needed" this or that capability, redundancy, etc.  It1
 > ended upE > being more than even IBM could build.  By the time the contract wasa@ > cancelled, it was massively bloated and years behind schedule. >oD >> With the advent of the Multipath Failover enhancements in OpenVMSG >> 7.3-1 I believe a real no-single-point-of-failure multi-site clusterdF >> has been worked-out in detail. But, I'm sure there are some clusterB >> experts here who understand the big picture better than myself. >>E >> I think Bob Kohler's point is that VMS Clusters wouldn't (at leasthE >> officially, probably also practically) not support enough multiple E >> remote sites and systems to include the entire U.S. nationwide ATCu- >> distributed system as one OpenVMS cluster.  > G > The number of sites is far less than you'd think -- if you count only4 > major " > centers, it's less than 2 dozen. >uG > But I think you're both mis-applying the strengths of clustering to aQC > problem for which the needs are different.  Multi-site clusteringi
 > providesD > geographic redundancy, where if site A goes down, B and C can take > the load.o >wG > In ATC, you "need" the data at all the sites all of the time.  And ifo > youe? > lose your connections to the other sites, it's much better to  > continue withvG > stale data than to hang.  Sure, distributed locking is a nice featurem > for G > the app, but if the OS depends on it, you're just asking for trouble.a    L Without doing the math, it seems to me that there are perhaps 5,000 thousandJ flights flights under civilian ATC control at any point in time. How oftenD are the datum representing the transponder transmissions updated perJ aircraft and any 'skin paint' radar updates? Are we talking multiple times0 per second or once every few seconds for an ATC?  J With typical route and aircraft separation I'd hazard a guess that even ifK adjacent regional clusters backed each other up, the time to transition theyH cluster would be less than any prudent safety margin would allow. AnyoneI have any knowledge that would shed more light on the feasability of this?e   ------------------------------  % Date: Tue, 21 Sep 2004 21:57:36 -0400i- From: JF Mezei <jfmezei.spamnot@teksavvy.com>.> Subject: Re: Windoze not rebooted monthly shuts down airports!, Message-ID: <4150DBF6.201F8069@teksavvy.com>   John Smith wrote:nL > With typical route and aircraft separation I'd hazard a guess that even ifM > adjacent regional clusters backed each other up, the time to transition theoJ > cluster would be less than any prudent safety margin would allow. AnyoneK > have any knowledge that would shed more light on the feasability of this?'  M Consider the in-air collision over switzerland last year. A couple of seconds>H of delay in sounding the alarm helped cause it, as well as contradictoryN information given by ATC and he plane's one board collision avoidance systems.  M You don't want a cluster transition to cause a a few seconds delay that couldgM be fatal. Odds are that a cluster transition wouldn't matter, but if it does,o it will be a very bad day.   ------------------------------  % Date: Tue, 21 Sep 2004 22:11:36 -0400'# From: "John Smith" <a@nonymous.com>i> Subject: Re: Windoze not rebooted monthly shuts down airports!, Message-ID: <ONadnTBrB4bEQs3cRVn-ug@igs.net>   JF Mezei wrote:  > John Smith wrote:nE >> With typical route and aircraft separation I'd hazard a guess thatSG >> even if adjacent regional clusters backed each other up, the time toeF >> transition the cluster would be less than any prudent safety marginG >> would allow. Anyone have any knowledge that would shed more light ont >> the feasability of this?r > G > Consider the in-air collision over switzerland last year. A couple of D > seconds of delay in sounding the alarm helped cause it, as well asA > contradictory information given by ATC and he plane's one board4 > collision avoidance systems.  I As I recollect, it was the combination of bad ATC advice and the pilot(s)<J not invoking their training for head-on approach - pull up to starboard or something like that.      D > You don't want a cluster transition to cause a a few seconds delayB > that could be fatal. Odds are that a cluster transition wouldn't4 > matter, but if it does, it will be a very bad day.    I So NSK then. But, I'm told by some at HP and friends who are NSK gurus atrE banks that NSK can, under certain circumstances, not maintain logicaloF consistency between nodes - completely missed transactions/updates. So  you're back to the same problem.   ------------------------------  % Date: Tue, 21 Sep 2004 22:10:01 -0400l( From: David Froble <davef@tsoft-inc.com>> Subject: Re: Windoze not rebooted monthly shuts down airports!, Message-ID: <4150DEF9.3000103@tsoft-inc.com>   Keith Cayemberg wrote:   > Keith A. Lewis wrote:n > C >> koehler@eisner.nospam.encompasserve.org (Bob Koehler) writes in aE >> article <5NXF4e+SI5L+@eisner.encompasserve.org> dated 21 Sep 2004 t >> 07:54:28 -0500: >>C >>>  Techies who were working on IBM's (later canceled) 1980's FAA nE >>> contract  tell me they spent a lot of time trying to emulate VMS a >>> paradigms.  AnJ >>>  understanding of the problem showed that what they really needed was F >>>  the world's largest VMScluster, much larger geographically and in: >>>  number of nodes than VMScluster supports, even today. >> >> >>K >> I don't think VMScluster would have worked.  ATC facilities need to keepaK >> running when any given part goes offline.  If the cluster communicationsaH >> link(s) went down, the VMS systems on the small partition would have  >> hung.    Q I've also got to disagree with this conclusion.  It's not cluster communications dO that would be the issue.  If part of a cluster goes away, fine, as long as all bJ users have access to the part of the cluster that remains available.  The N communications problem would be to insure all users had multiple paths to the O network, and that the network would always have access to whatever part of the   cluster still available.  M With enough shadow sets, or however storage would be set up, all data should nC remain available to whatever part of the cluster remains available.e  P I'm not real knowledgable about running a cluster, so I'm sure there are others - who could provide solutions for the problems.-    3 >> --Keith Lewis              klewis {at} mitre.org1A >> The above may not (yet) represent the opinions of my employer.j >  > > > Any reason why one couldn't have multiple redundant cluster E > interconnects (say 4 or 5 for good measure) between all sites with r > different routes?  > J > With the advent of the Multipath Failover enhancements in OpenVMS 7.3-1 J > I believe a real no-single-point-of-failure multi-site cluster has been J > worked-out in detail. But, I'm sure there are some cluster experts here 4 > who understand the big picture better than myself. > E > I think Bob Kohler's point is that VMS Clusters wouldn't (at least wE > officially, probably also practically) not support enough multiple nE > remote sites and systems to include the entire U.S. nationwide ATC i, > distributed system as one OpenVMS cluster.    M I'm not sure why 96 nodes wouldn't be enough.  That's almost 2 per state, by  M average.  Most likely less than 96, possibly much less, could do the job.  I eO don't know the requirements so I'm just speculating, but I have to wonder just d* how many GS1280 systems would be required.    C > However, any other competing distributed system would have their iC > limitations in regard to implementing such a massive multi-site, tI > concurrent-update single data image ATC system as appears to have been b > the goal.o > J > Although there are also some caveats concerning the reduced suitability D > of clusters for hard real-time systems (predictability of service 9 > latency would require a stand-alone system), I believe  F > VMS-Clustering-based ATC System implemented as multiple (regional?) F > clusters would still offer enough answers to the overall ATC system E > problem space to provide an exceptional solution using some clever ,F > designing. I'm sure there would be some VMS clustering experts that D > would greatly enjoy sinking their teeth into such a juicy problem. > 	 > Cheers!  >  > Keith Cayembergr >      I'm inclined to agree with you.s     Dave   -- w4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com 170 Grimplin Roade Vanderbilt, PA  15486    ------------------------------  % Date: Wed, 22 Sep 2004 00:22:56 -04001- From: JF Mezei <jfmezei.spamnot@teksavvy.com><> Subject: Re: Windoze not rebooted monthly shuts down airports!, Message-ID: <4150FE1F.8E88A8DF@teksavvy.com>   David Froble wrote:fR > I've also got to disagree with this conclusion.  It's not cluster communicationsP > that would be the issue.  If part of a cluster goes away, fine, as long as allF > users have access to the part of the cluster that remains available.  # Perhaps Patrick Moreau can chip in.c  G I think one has to take a more global look at an ATC application, whichaM includes workstations as well as the connection between a workstation and theaA server infrastructure. Availability becomes an interesting issue.a  K There is an interesting transactional nature between regions, when a flightrL gets transfered from one regional centre to another. It isn't enough to justK tell the new centre is it now handling that flight, but there would also beyJ collision detection issues to ensure that two flights in adjoining regions2 wouldn't enter a 3rd region on a collision course.  H But inside the region, it isn't so much a transactional system as a dataH acquisition and processing to generate vectors and detect collision etc.  N > With enough shadow sets, or however storage would be set up, all data shouldE > remain available to whatever part of the cluster remains available.e  O But you are also more likely to see global sections that contain the live data.e  Q > I'm not real knowledgable about running a cluster, so I'm sure there are othersT/ > who could provide solutions for the problems.C  E Thing is that one should NOT view VMS clusters as the solution to ATC N problems. We should view the services provided by VMS clustering as tools thatL would allow such an application to be architected in a very elegant way. TheA application would be far more important than the OS in that case.   1 > I'm not sure why 96 nodes wouldn't be enough.  tP > don't know the requirements so I'm just speculating, but I have to wonder just, > how many GS1280 systems would be required.  K Whether you have standalone systems or galaxies/wildfires really depends ong the application architecture.t   ------------------------------  % Date: Tue, 21 Sep 2004 17:30:57 -0400s( From: David Froble <davef@tsoft-inc.com>C Subject: Re: ZIP "-V" v. UNIX, et al.: Problem, possible solutions. , Message-ID: <41509D91.2040802@tsoft-inc.com>   Soterro wrote:   > David J Dachtera wrote:l > F >> That said, in my DECUS presentation from Fall-99 (San Diego), I putJ >> forth the concept that the target platform should be borne in mind whenJ >> creating the archive. Preserving RMS attributes is really only of valueG >> when it's needed for OpenVMS. VMS .ZIP archives (using "-V" or /VMS)y >  > J > True, but this avoids the real issue. As an user and not a developer, I J > rather am of the original poster's opinion. I hate when developers come J > to me saying 'do it because we know what is good for you' and then stop B > listening to any feedback (this reminds me of the whole Mozilla  > multi-user profiles crap). > K > In this case, when I create an archive with "-V" I'd certainly expect to  H > have a 'normal' archive (normal like in multiplatform, that's the ZIP H > goal isn't it), and when and only when I unpack it on VMS to have the I > extra goodies packed by "-V" (like attributes). 'Platform independent' lJ > that is. I'm really looking forward for that, as at the moment I had to L > write Perl utilities to shape files between VMS Unix and Windows... awful. > J > I agree it's more work to develop, but that's what are developers for :D >  > S   3 Developers have another purpose.  To get paid.  :-).  O The problem with what you're asking is that if a version of ZIP were set up to  M have a place in the archive for RMS attributes so that they could be ignored tO when looking at/unpacking an archive on a non VMS system, every single variety -P of ZIP in use would have to know about this 'place' so that it could be ignored.  O You're basically asking for the ZIP format to change.  Not feasable.  How many 5A people out there will give a damn about a change VMS people want?o  G The non-standard option has been put into VMS versions of ZIP to allow oI attributes to be preserved between VMS systems.  It's not valid on other eQ systems.  If you want it to be, as mentioned, you'll first have to change ZIP on   all other systems.  P Ok, one of possibly many work arounds.  Place the custom stuff in another file, N and need both files to extract attributes.  The primary file would conform to & the format every other system expects.  L If you want this, feel free to implement it, or, did you want to pay for it?   Dave   --  4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com 170 Grimplin Road  Vanderbilt, PA  15486)   ------------------------------  % Date: Tue, 21 Sep 2004 20:33:35 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net>C Subject: Re: ZIP "-V" v. UNIX, et al.: Problem, possible solutions.0+ Message-ID: <4150D66F.4C4096E5@comcast.net>    Tim Shoppa wrote:v > I > sms@antinode.org wrote in message news:<04092017270038@antinode.org>... " > >  What you _do_ get is useless,K > > erroneous NUL padding.  (Or do you actually believe that the almost 16KhG > > of NULs is really a clever encoding of the VMS file attributes?  OrlI > > perhaps a file exactly 16K bytes long has exactly zero "extra stuff"?  > > Let's get serious.)g > K > Whoa!  Calm down there.  I'm not saying that the extra stuff isn't there."I > I agree fully that it is there.  And if you do any VMS file format that)1 > has RMS record lengths all those are there too.  > N > But, please, stop, take a step back, and look at the info-zip documentation: > C >        -V     [VMS] Save VMS file attributes.  zip archives  cre- C >               ated  with this option will generally not be usableh! >               on other systems.  > A > They (and I) make no guarantee that Info-zip's "-V" option will*M > produce what any particular person (including you) wants when unzipped on alI > non-VMS system.  In general there's a bunch of RMS junk between records5G > and potentially random junk at the end.  The case you're all hyped upv5 > about (stream LF) is indeed the best case but stillo > not perfect. > K > You're perfectly free to hack up the innards of Info-zip however you wanteF > to do whatever you think it should do.  But we still aren't into theI > realm of "problem" much less "solution" because it's working exactly ase
 > documented.  > I > >    Please, before telling me again that ZIP -V _should_ work this way L > > because it _does_ work this way, do us all a favor and look at the code. > > IT'S BROKEN. > ' > Look at the documentation.  IT WORKS..  H Um, well, no - it doesn't - at least, not in V2.3. Steve's correct thereD - I duplicated his results on VMS using ZIP V2.3 with /VMS and "-V",F then used UNZIP v5.51 to /LIST (-l). The sizes displayed as he posted: small files show as 16384.   $ dump/rec abc.seq  A Dump of file USER$ROOT:[WORK]ABC.SEQ;1 on 21-SEP-2004 20:25:11.69 : File ID (3875,2467,0)   End of file block 1 / Allocated 35  ? Record number 1 (00000001), 3 (0003) bytes, RFA(0001,0000,0000)(  <                               434241 ABC............. 000000   $ dump/rec abc.slf  A Dump of file USER$ROOT:[WORK]ABC.SLF;1 on 21-SEP-2004 20:25:17.69t9 File ID (3667,462,0)   End of file block 1 / Allocated 35l  ? Record number 1 (00000001), 3 (0003) bytes, RFA(0001,0000,0000)r  <                               434241 ABC............. 000000   $ dump/blo=cou=1 abc.seq  A Dump of file USER$ROOT:[WORK]ABC.SEQ;1 on 21-SEP-2004 20:25:31.52t: File ID (3875,2467,0)   End of file block 1 / Allocated 35  3 Virtual block number 1 (00000001), 512 (0200) bytes0  <  00000000 00000000 FFFF0043 42410003 ..ABC........... 000000<  00000000 00000000 00000000 00000000 ................ 000010<  00000000 00000000 00000000 00000000 ................ 000020<  00000000 00000000 00000000 00000000 ................ 000030<  00000000 00000000 00000000 00000000 ................ 000040<  00000000 00000000 00000000 00000000 ................ 000050<  00000000 00000000 00000000 00000000 ................ 000060<  00000000 00000000 00000000 00000000 ................ 000070<  00000000 00000000 00000000 00000000 ................ 000080<  00000000 00000000 00000000 00000000 ................ 000090<  00000000 00000000 00000000 00000000 ................ 0000A0<  00000000 00000000 00000000 00000000 ................ 0000B0<  00000000 00000000 00000000 00000000 ................ 0000C0<  00000000 00000000 00000000 00000000 ................ 0000D0<  00000000 00000000 00000000 00000000 ................ 0000E0<  00000000 00000000 00000000 00000000 ................ 0000F0<  00000000 00000000 00000000 00000000 ................ 000100<  00000000 00000000 00000000 00000000 ................ 000110<  00000000 00000000 00000000 00000000 ................ 000120<  00000000 00000000 00000000 00000000 ................ 000130<  00000000 00000000 00000000 00000000 ................ 000140<  00000000 00000000 00000000 00000000 ................ 000150<  00000000 00000000 00000000 00000000 ................ 000160<  00000000 00000000 00000000 00000000 ................ 000170<  00000000 00000000 00000000 00000000 ................ 000180<  00000000 00000000 00000000 00000000 ................ 000190<  00000000 00000000 00000000 00000000 ................ 0001A0<  00000000 00000000 00000000 00000000 ................ 0001B0<  00000000 00000000 00000000 00000000 ................ 0001C0<  00000000 00000000 00000000 00000000 ................ 0001D0<  00000000 00000000 00000000 00000000 ................ 0001E0<  00000000 00000000 00000000 00000000 ................ 0001F0   $ dump/blo=cou=1 abc.slf  A Dump of file USER$ROOT:[WORK]ABC.SLF;1 on 21-SEP-2004 20:25:35.13X9 File ID (3667,462,0)   End of file block 1 / Allocated 35   3 Virtual block number 1 (00000001), 512 (0200) bytes   <  00000000 00000000 00000000 0A434241 ABC............. 000000<  00000000 00000000 00000000 00000000 ................ 000010<  00000000 00000000 00000000 00000000 ................ 000020<  00000000 00000000 00000000 00000000 ................ 000030<  00000000 00000000 00000000 00000000 ................ 000040<  00000000 00000000 00000000 00000000 ................ 000050<  00000000 00000000 00000000 00000000 ................ 000060<  00000000 00000000 00000000 00000000 ................ 000070<  00000000 00000000 00000000 00000000 ................ 000080<  00000000 00000000 00000000 00000000 ................ 000090<  00000000 00000000 00000000 00000000 ................ 0000A0<  00000000 00000000 00000000 00000000 ................ 0000B0<  00000000 00000000 00000000 00000000 ................ 0000C0<  00000000 00000000 00000000 00000000 ................ 0000D0<  00000000 00000000 00000000 00000000 ................ 0000E0<  00000000 00000000 00000000 00000000 ................ 0000F0<  00000000 00000000 00000000 00000000 ................ 000100<  00000000 00000000 00000000 00000000 ................ 000110<  00000000 00000000 00000000 00000000 ................ 000120<  00000000 00000000 00000000 00000000 ................ 000130<  00000000 00000000 00000000 00000000 ................ 000140<  00000000 00000000 00000000 00000000 ................ 000150<  00000000 00000000 00000000 00000000 ................ 000160<  00000000 00000000 00000000 00000000 ................ 000170<  00000000 00000000 00000000 00000000 ................ 000180<  00000000 00000000 00000000 00000000 ................ 000190<  00000000 00000000 00000000 00000000 ................ 0001A0<  00000000 00000000 00000000 00000000 ................ 0001B0<  00000000 00000000 00000000 00000000 ................ 0001C0<  00000000 00000000 00000000 00000000 ................ 0001D0<  00000000 00000000 00000000 00000000 ................ 0001E0<  00000000 00000000 00000000 00000000 ................ 0001F0   $ zip/level=8/vms abcslf abc.*    adding: ABC.SEQ (deflated 86%)    adding: ABC.SLF (deflated 77%)   $ unzip/list abcslfc& Archive:  USER$ROOT:[WORK]ABCSLF.ZIP;1    Length     Date   Time    Name   --------    ----   ----    ----#     16384  09-20-04 13:14   abc.seqe#     16384  09-20-04 13:14   abc.slf #  --------                   -------d#     32768                   2 filesa   $ zip/level=8 abcslfu abc.*-   adding: ABC.SEQ (stored 0%)e   adding: ABC.SLF (stored 0%)0   $ unzip/list abcslfu' Archive:  USER$ROOT:[WORK]ABCSLFU.ZIP;1W    Length     Date   Time    Name   --------    ----   ----    ----#         4  09-20-04 13:14   abc.seqA#         4  09-20-04 13:14   abc.slfc#  --------                   -------t#         8                   2 filesg  . WinZip and PKZip for Windows agree with UNZIP.   $ zipi  Copyright (C) 1990-1999 Info-ZIP% Type 'zip "-L"' for software license. > Zip 2.3 (November 29th 1999). Usage: zip=="$disk:[dir]zip.exe"< zip zipfile[.zip] [list] [/EXCL=(xlist)] /options /modifiersD   The default action is to add or replace zipfile entries from list, exceptF   those in xlist. The include file list may contain the special name - toE   compress standard input.  If both zipfile and list are omitted, zip2   compresses stdin to stdout.L#   Type zip -h for Unix style flags.a   Major options include:H     /FRESHEN, /UPDATE, /DELETE, /[NO]MOVE, /COMMENTS[={ZIP_FILE|FILES}],@     /LATEST, /TEST, /ADJUST_OFFSETS, /FIX_ARCHIVE[=FULL], /UNSFX   Modifiers include:G     /EXCLUDE=(file list), /INCLUDE=(file list), /SINCE="creation time",a@     /QUIET, /VERBOSE[=MORE], /[NO]RECURSE, /[NO]DIRNAMES, /JUNK,H     /[NO]KEEP_VERSION, /[NO]VMS, /[NO]PKZIP, /TRANSLATE_EOL[={LF|CRLF}],F     /[NO]EXTRA_FIELDS /LEVEL=[0-9], /TEMP_PATH=directory, /BATCH[=list file]v   $ unzip H UnZip 5.51 of 22 May 2004, by Info-ZIP.  For more details see: unzip -v.  D Usage: unzip file[.zip] [list] [/EXCL=(xlist)] [/DIR=exdir] /options
 /modifiersG   Default action is to extract files in list, except those in xlist, toh exdir;G   file[.zip] may be a wildcard.  Type "unzip /ZIPINFO" for ZipInfo-mode  usage.  ; Major options include (type unzip -h for Unix style flags):dC    /[NO]TEST, /LIST, /[NO]SCREEN, /PIPE, /[NO]FRESHEN, /[NO]UPDATE,oF    /[NO]COMMENT, /DIRECTORY=directory-spec, /EXCLUDE=(file-spec1,etc.)   Modifiers include:,    /BRIEF, /FULL, /[NO]TEXT[=NONE|AUTO|ALL], /[NO]BINARY[=NONE|AUTO|ALL],@    /[NO]OVERWRITE, /[NO]JUNK, /QUIET, /QUIET[=SUPER], /[NO]PAGE,D    /[NO]CASE_INSENSITIVE, /[NO]LOWERCASE, /[NO]VERSION, /[NO]RESTORE  7 Examples (see unzip.txt or "HELP UNZIP" for more info): F    unzip edit1 /EXCL=joe.jou /CASE_INSENSITIVE    => extract all files exceptC       joe.jou (or JOE.JOU, or any combination of case) from zipfile 	 edit1.zip @    unzip zip201 "Makefile.VMS" vms/*.[ch]         => extract VMS Makefile andG       *.c and *.h files; must quote uppercase names if /CASE_INSENS not  usedF    unzip foo /DIR=tmp:[.test] /JUNK /TEXT /OVER   => extract all files to temp.=       directory without paths, auto-converting text files ande overwritingv   Does this help?    D.J.D.   ------------------------------  % Date: Tue, 21 Sep 2004 20:45:25 -0500c2 From: David J Dachtera <djesys.nospam@comcast.net>C Subject: Re: ZIP "-V" v. UNIX, et al.: Problem, possible solutions.g+ Message-ID: <4150D934.1DF5F906@comcast.net>n   Soterro wrote: >  > David J Dachtera wrote:oG > > That said, in my DECUS presentation from Fall-99 (San Diego), I put K > > forth the concept that the target platform should be borne in mind when K > > creating the archive. Preserving RMS attributes is really only of value-H > > when it's needed for OpenVMS. VMS .ZIP archives (using "-V" or /VMS) > I > True, but this avoids the real issue. As an user and not a developer, IoI > rather am of the original poster's opinion. I hate when developers come I > to me saying 'do it because we know what is good for you' and then stopnA > listening to any feedback (this reminds me of the whole Mozillat > multi-user profiles crap). > J > In this case, when I create an archive with "-V" I'd certainly expect toG > have a 'normal' archive (normal like in multiplatform, that's the ZIPn > goal isn't it),"   Um, well, no, it isn't.   7 > and when and only when I unpack it on VMS to have thefH > extra goodies packed by "-V" (like attributes). 'Platform independent'I > that is. I'm really looking forward for that, as at the moment I had to L > write Perl utilities to shape files between VMS Unix and Windows... awful.  G Learn to use ZIP and UNZIP to your advantage instead of working againstoF yourself. Then, you can scrap the perl stuff or keep it for reference.  J > I agree it's more work to develop, but that's what are developers for :D  = What I'll do at this point is refer you to some of my on-lineo documentation:  ) http://www.djesys.com/vms/mentor/rms.htmla  C I refer you to the RMS page because we need to remember that RMS is4E unique to the DEC world (although some third-party products exist foruA other platforms). In order for what you ask to be possible, everygD archive unpacker out there that understands Phil Katz's format wouldF have to also understand RMS so it can correcftly unpack the archive onE systems which lack RMS natively. Needless to say, such has never beeni& the case, nor is it likely to ever be.  4 http://www.djesys.com/vms/support/zipunzip/index.htm  7 In the ZIP/UNZIP presentation, see specifically slides:   5 http://www.djesys.com/vms/support/zipunzip/sld029.htme thru5 http://www.djesys.com/vms/support/zipunzip/sld038.htmo  5 http://www.djesys.com/vms/support/zipunzip/sld044.htmr  D If that leaves you scratching your head, write me privately (see theH Contact Us page on the website, or demung the reply-to on this post) and we can continue this off-line.   D.J.D.   ------------------------------  + Date: Tue, 21 Sep 2004 19:23:26 -0500 (CDT)i From: sms@antinode.orgC Subject: Re: ZIP "-V" v. UNIX, et al.: Problem, possible solutions. ) Message-ID: <04092119232598@antinode.org>t  ( From: David Froble <davef@tsoft-inc.com>   > Soterro wrote:  = > > [...] I rather am of the original poster's opinion. [...]5  6    Clearly a man of nice judgement and many resources.  Q > The problem with what you're asking is that if a version of ZIP were set up to  O > have a place in the archive for RMS attributes so that they could be ignored nQ > when looking at/unpacking an archive on a non VMS system, every single variety  R > of ZIP in use would have to know about this 'place' so that it could be ignored. > Q > You're basically asking for the ZIP format to change.  Not feasable.  How many mC > people out there will give a damn about a change VMS people want?e > [...]m  F    I always hate to suggest that anyone is a moron, but what makes youH think that a change would be required?  Rather than arguing in a vacuum,G you might take a quick glance at the Zip source code, paying particular 6 attention to occurrences of the phrase, "extra field".  I > The non-standard option has been put into VMS versions of ZIP to allow  K > attributes to be preserved between VMS systems.  It's not valid on other  S > systems.  If you want it to be, as mentioned, you'll first have to change ZIP on g > all other systems.  C    Is that bloated-rotting-dead horse still moving?  Let me beat itcB again.  I realize that for some folks no actual evidence will everF conquer a strongly held opinion reinforced by ignorance, but my latestB Wget kit is a ZIP -V archive, made with the Zip bug fixed.  PleaseG complain if you find any corrupted files when you unpack it on your VMSaD system.  (Please pay particular attention to the attribute-sensitiveF object files.)  I unpacked it on a Tru64 V5.1B system (and SunOS 5.9),# and it worked just fine there, too.e  F    I am reminded here of the words of the immortal Rocket J. Squirrel,  "Thank you, Mister Know-it-all."  H ------------------------------------------------------------------------  4    Steven M. Schweda               (+1) 651-699-98183    382 South Warwick Street        sms@antinode.orgy    Saint Paul  MN  55105-2547s   ------------------------------    Date: 21 Sep 2004 17:11:10 -0700. From: spamsink2001@yahoo.com (Alan E. Feldman)A Subject: [DCL REQUEST] New ignore keyword for DIFFERENCES commandd= Message-ID: <b096a4ee.0409211611.455028c7@posting.google.com>u  . I'd like to see /IGNORE=LEADING_DOLLAR_SIGNS.   C The purpose of this would be to compare DCL command procedures moreiC easily when using /IGNORE=(COMMENTS, TRAILING_SPACES, BLANK_LINES).dD The problem with using this is that it greatly increases the odds ofA DIFFERENCES losing synchronization between the two files. This isv@ because blocks of comments suddenly become lines containing only> leading dollar signs. The records within these blocks are thenC indistinguishable and are often mistakenly matched when they are inrD different parts of the file. Thus, a larger -- sometimes much larger9 -- /MATCH value is needed to reestablish synchronization.e  C By ignoring the leading dollar signs before ignoring comments, thene4 trailing spaces, and finally blank lines, lines like       $! exciting commentn  E completely "disappear" from the comparison and one can usually keep a/F small match value. Additionally, the clutter caused by rows of leading dollar signs is eliminated.a  C Currently, as a workaround, I remove the dollar signs myself before = running DIFFERENCES. Obviously, it would be much nicer if thee) DIFFERENCES command itself would do that.r   A project for Guy Peleg?   ------------------------------  % Date: Tue, 21 Sep 2004 13:57:12 -0400 # From: "John Smith" <a@nonymous.com>g' Subject: Re: [OT]: McNeally understandsn, Message-ID: <c-qdnbwQw8_l9s3cRVn-pA@igs.net>   Bob Ceculski wrote:t0 > "John Smith" <a@nonymous.com> wrote in message* > news:<X76dnUzreLH-u83cRVn-rg@igs.net>... >>C >> Joel Scotkin, CEO of Random Walk, which resells both Sun and IBMlF >> equipment to financial firms, said Wall Street customers are likelyC >> to welcome the features of Solaris 10. The operating system willa> >> offer a new file system and the ability to isolate parts ofD >> computers into "containers." These containers protect one part ofF >> the machine that supports a particular application from the rest of, >> the machine when another program crashes. >>F >> But Scotkin noted that IBM's Power5-based servers are well-received? >> on Wall Street and that IBM appears to be ahead in so-calledcB >> virtualization technology, which resembles Sun's containers and1 >> enables a customer to maximize a server's use.s >o> > VMS has had clusters for over 20 years now and Galaxy for 10> > I think ... welcome to something you could have had 10 years > ago ... :)    J VMS had clusters for 20 years....whoopee! Look at how much market share itF has. Now look how much market share Sun has of VMS's former customers.  J Sun has been able to sell, in the immortal words of Ken Olsen, 'snake oil'K because they decided to actively market it. Imagine what could be done withtL VMS if it were actively marketed to the types of companies that could really  use VMS features and robustness.  I I think it would be a slam-dunk to sell an extra 100,000 VMS servers overrL HP's current forecasts if they only thought "outside the box", which in HP's case is a coffin.    ------------------------------   End of INFO-VAX 2004.526 ************************