1 INFO-VAX	Thu, 25 Apr 2002	Volume 2002 : Issue 228       Contents:& Re: (GCC) GNU C/C++ compiler for Alpha8 (SOLVED) Apache (CSWS), TCPware and DCL cgi-bin problems Re: ACE propagation problems Re: ACE propagation problems# Re: ADA on Hobbyist CD? (VAXVMS072) 9 Re: AMD To Announce Microsoft Support For New Chip Family 9 Re: AMD To Announce Microsoft Support For New Chip Family 9 Re: AMD To Announce Microsoft Support For New Chip Family 9 Re: AMD To Announce Microsoft Support For New Chip Family 9 Re: AMD To Announce Microsoft Support For New Chip Family 9 Re: AMD To Announce Microsoft Support For New Chip Family 9 Re: AMD To Announce Microsoft Support For New Chip Family 9 Re: AMD To Announce Microsoft Support For New Chip Family 9 Re: AMD To Announce Microsoft Support For New Chip Family 9 Re: AMD To Announce Microsoft Support For New Chip Family 9 Re: AMD To Announce Microsoft Support For New Chip Family 9 Re: AMD To Announce Microsoft Support For New Chip Family 9 Re: AMD To Announce Microsoft Support For New Chip Family  Re: AOL  AST question Re: AST question Re: AST question Re: AST question Re: AST question Re: AST question Re: Best DIGITAL product ever?% Building Emacs 19.28 on Alpha VMS 7.2 ) Re: Building Emacs 19.28 on Alpha VMS 7.2 ) Re: Building Emacs 19.28 on Alpha VMS 7.2 ) Re: Building Emacs 19.28 on Alpha VMS 7.2 & CGI access problem using CSWS (Apache)* Re: CGI access problem using CSWS (Apache)6 CIA warns of China cyber attacks ... better be on VMS!: RE: CIA warns of China cyber attacks ... better be on VMS!: Re: CIA warns of China cyber attacks ... better be on VMS!: RE: CIA warns of China cyber attacks ... better be on VMS!: Re: CIA warns of China cyber attacks ... better be on VMS!: Re: CIA warns of China cyber attacks ... better be on VMS! closed user groups / x254 Re: creat function hanging -- Code snippet provided! Disk storage for DS10  Disk storage for DS10  Re: Disk storage for DS10  Re: Disk storage for DS10  Re: Disk storage for DS10  download VMS Re: download VMS RE: download VMS Re: download VMS Re: DSSI Interconnect  Re: DSSI Interconnect # EDIT/FDL changes colours in DECterm  Re: EDT or EVE Re: EDT or EVE Re: EDT or EVE Re: EDT or EVE Re: EDT or EVE Re: EDT or EVE Re: EDT or EVE Re: EDT or EVE Re: EDT or EVE Re: EDT or EVE Re: EDT or EVE Re: EDT or EVE (EDT) Re: EDT or EVE (TECO RULES!) Re: EDT or EVE (TECO!) Re: EDT or EVE (TECO!) Re: EDT or EVE (TECO!)@ How to configure DECvms to connect 3Com OfficeConnect ISDN ModemD Re: How to configure DECvms to connect 3Com OfficeConnect ISDN ModemD Re: How to configure DECvms to connect 3Com OfficeConnect ISDN ModemD Re: How to configure DECvms to connect 3Com OfficeConnect ISDN ModemD Re: How to configure DECvms to connect 3Com OfficeConnect ISDN ModemD Re: How to configure DECvms to connect 3Com OfficeConnect ISDN ModemD Re: How to configure DECvms to connect 3Com OfficeConnect ISDN ModemD Re: How to configure DECvms to connect 3Com OfficeConnect ISDN Modem- Re: I was right!  Alpha will live in Itanium! - Re: I was right!  Alpha will live in Itanium! 0 Re: Inquirer even says Alpha in Itanium in 2005!0 Re: Inquirer even says Alpha in Itanium in 2005!0 Re: Inquirer even says Alpha in Itanium in 2005!0 Re: Inquirer even says Alpha in Itanium in 2005!0 Re: Inquirer even says Alpha in Itanium in 2005!+ Itanium 2 hits ... Itanium 3 will be Alpha! ? Re: markup vs wysiwyg (was: Re: learning how to use a computer) ? Re: markup vs wysiwyg (was: Re: learning how to use a computer) ? Re: markup vs wysiwyg (was: Re: learning how to use a computer)  Re: Memory Corruption  Re: Memory Corruption  Re: Netscape and Mozilla5 Never lose the data in your Palm Desktop again. 38733  Re: Openvm 7.3 and DECevent # Re: Pathworks 6.1 exhausting KNBCBs P put up or shut up (was: RE: AMD To Announce Microsoft Support For New Chip Famil- Read this if you own a PalmOS handheld. 68655  SMTP sanity checks RE: SMTP sanity checks Re: SMTP sanity checks Re: SMTP sanity checks> Re: somewhat off-topic: is there a DSL router with a console?!> RE: somewhat off-topic: is there a DSL router with a console?!> Re: somewhat off-topic: is there a DSL router with a console?!  Re: Suggestion: SET DEVICE/RESET System slowdowns with RWSCS  Re: TCPWare Vs MultiNet O Re: The 'tone' of c.o.v. (was Re: Prediction:  VMS lives, merger or no merger!) O Re: The 'tone' of c.o.v. (was Re: Prediction:  VMS lives, merger or no merger!) O Re: The 'tone' of c.o.v. (was Re: Prediction:  VMS lives, merger or no merger!) O Re: The 'tone' of c.o.v. (was Re: Prediction:  VMS lives, merger or no merger!) O Re: The 'tone' of c.o.v. (was Re: Prediction:  VMS lives, merger or no merger!) O Re: The 'tone' of c.o.v. (was Re: Prediction:  VMS lives, merger or no merger!) & Re: VAX decommissioning/secure HD wipe& Re: VAX decommissioning/secure HD wipe& Re: VAX decommissioning/secure HD wipe& Re: VAX decommissioning/secure HD wipe& Re: VAX decommissioning/secure HD wipe& Re: VAX decommissioning/secure HD wipe VAXCLUSTER license alternative) Re: VT520 and VT420 terminal and PC mouse ) Re: VT520 and VT420 terminal and PC mouse # Web based/encrypted printing (O.T.) ' Re: Web based/encrypted printing (O.T.)  web hosting  Re: web hosting  Re: web hosting  Re: web hosting  Re: web hosting  Re: web hosting  Re: web hosting   Re: where can I find older DECC?- will memory leak program caused system crash? 1 Re: will memory leak program caused system crash? 1 Re: will memory leak program caused system crash?   F ----------------------------------------------------------------------  % Date: Thu, 25 Apr 2002 09:56:42 -0400 > From: Charlie McCutcheon <charlie.mccutcheon@NOSPAMcompaq.com>/ Subject: Re: (GCC) GNU C/C++ compiler for Alpha 0 Message-ID: <3CC80B1A.E45BEB23@NOSPAMcompaq.com>   Fabio Becherini wrote:   > Hi folks,  > . > We have OpenVMS V7.2 running in some Alphas,7 > and now we are looking for a (GCC) GNU C/C++ compiler  > for these machines.  > & > Does anyone have a solution for it ? >  > Best regards !  F Perahps try http://gcc.gnu.org/ .   I didn't look far enough to see if they have a download for VMS.    Charlie    ------------------------------    Date: 25 Apr 2002 01:35:59 -0700  From: theop@itex.je (Theo Platt)A Subject: (SOLVED) Apache (CSWS), TCPware and DCL cgi-bin problems < Message-ID: <31ff7ddd.0204250035.2af167f@posting.google.com>  @ Sometime ago I posted a series of messages about having problemsB running cgi-bin scripts on Apache (Compaq Secure Web Server) under TCPware.  F If I called the test-cgi-vms.com test script with the DEC TCP-IP stackC in place it worked fine but with the TCPware stack I'd get internal  server errors.  B Thanks for everyone who gave me help. It turns out that the reasonD TCPware failed was that the loopback interface had not been assignedC (LPB-0 127.0.0.1) - this is an option with the TCPware setup but is ? default with DEC TCP-IP (as our systems guys tell us). Once the < loopback address is correctly defined everything works fine.  C I hope this helps someone else who may have the same problem later.    Theo   ------------------------------    Date: 25 Apr 2002 07:24:24 -0000= From: Doc.Cypher <Use-Author-Supplied-Address-Header@[127.1]> % Subject: Re: ACE propagation problems 6 Message-ID: <20020425072424.32666.qmail@gacracker.org>  J On Thu, 25 Apr 2002, "Peter A. Stoll" <archae86_unspam@comcast.net> wrote:J >I've been struggling with the OpenVMS ACL file protection in a particular >need. > J >My desire is to allow four specific users to create and change files in a/ >subdirectory, and to allow all others to read.  > G >So long as just one of the four users touchs any one file, my existing  >scheme works fine.  > J >But when a second one of my "allowed four" changes a file or copies a newK >version of a file that a first one previously touched, the first one's ACE H >often disappears from the latest revision.  This denies that first user, >their ability to update or delete the file.   <snip>  J I'd really keep this more manageable by adding an identifier in the rightsH database and granting it to these four users. I would say that with fourH users you are at the limit of manageability. You're going to have a realG mess if you had to let another four users access this directory and the J files in it. (Modifying ACLs on files and directories, propogating changes with SET SECURITY/LIKE.)  ! Try working with the following...    $ mc authorize add/id TESTJ %UAF-I-RDBADDMSG, identifier TEST value %X80010006 added to rights databas $ mc authorize grant/id TEST DC . %UAF-I-GRANTMSG, identifier TEST granted to DC% [repeat as required for other users].     = I use SET SECURITY/EDIT to set up the ACLs, and afterwards...    $ dir/sec test.dir   Directory SYS$SYSROOT:[SYSMGR]  E TEST.DIR;1           [SYSTEM]                         (RWE,RWE,RE,RE)   E (IDENTIFIER=TEST,OPTIONS=DEFAULT+PROTECTED,ACCESS=READ+WRITE+EXECUTE+            DELETE+CONTROL) 5           (IDENTIFIER=TEST,ACCESS=READ+WRITE+EXECUTE) @           (IDENTIFIER=[*,*],OPTIONS=DEFAULT,ACCESS=READ+EXECUTE)0           (IDENTIFIER=[*,*],ACCESS=READ+EXECUTE)   Total of 1 file.    K I'm not sure if it is relevant, but on the second ACL I don't grant delete, E if I'm correct this is what you should to to prevent someone actually  deleting the directory itself.     Doc. --  6 The bigger the humbug, the better people will like it.K ~ Phineas Taylor Barnum.                             https://vmsbox.cjb.net    ------------------------------    Date: 25 Apr 2002 02:13:12 -0700) From: P.Young@unsw.EDU.AU (Patrick Young) % Subject: Re: ACE propagation problems = Message-ID: <55f85d77.0204250113.45a68960@posting.google.com>   ~ "Peter A. Stoll" <archae86_unspam@comcast.net> wrote in message news:<ukKx8.187173$K5.15803392@bin5.nnrp.aus1.giganews.com>...K > I've been struggling with the OpenVMS ACL file protection in a particular  > need.  > K > My desire is to allow four specific users to create and change files in a 0 > subdirectory, and to allow all others to read.  K You are on the right track, however the files should be owned by a resource N identifier and that identifier granted to the users (with attribute resource).  3 For example, the following should be what you want:   % Directory GRP$WWW:[ECONOMICS.COURSES]   C ECON1101.DIR;1       ECONOMICS_RES                    (RWED,RWED,,) G           (IDENTIFIER=[WWW,APACHE],OPTIONS=DEFAULT,ACCESS=READ+EXECUTE) 7           (IDENTIFIER=[WWW,APACHE],ACCESS=READ+EXECUTE) H           (IDENTIFIER=[WWW,ECONOMICS],OPTIONS=DEFAULT,ACCESS=READ+WRITE+           EXECUTE+DELETE) G           (IDENTIFIER=[WWW,ECONOMICS],ACCESS=READ+WRITE+EXECUTE+DELETE) G           (IDENTIFIER=[WWW,ECON1101],OPTIONS=DEFAULT,ACCESS=READ+WRITE+            EXECUTE+DELETE) F           (IDENTIFIER=[WWW,ECON1101],ACCESS=READ+WRITE+EXECUTE+DELETE)C           (DEFAULT_PROTECTION,SYSTEM:RWED,OWNER:RWED,GROUP:,WORLD:)            (CREATOR,ACCESS=NONE)   " H mc authorize sh/id economics_res=   Name                             Value           Attributes ;   ECONOMICS_RES                    %X81007000      RESOURCE    ------------------------------    Date: 25 Apr 2002 09:40:35 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen) , Subject: Re: ADA on Hobbyist CD? (VAXVMS072)3 Message-ID: <CwMOQsVCngpj@eisner.encompasserve.org>   j In article <3cc78649.7211248@news.nauticom.net>, tauberg@nauticom.no_spam_please.net (Jim Tauberg) writes: > Hi, ; > 	The Application Licenses include ADA, but i don't see it G > anywher on the Hobbyist CD (Vol:  VAXVMS072).  Does anyone know where  > to get it?  C Consolidated Distribution kits are often found for auction on eBay.    ------------------------------  # Date: Thu, 25 Apr 2002 06:05:54 GMT * From: "Bill Todd" <billtodd@metrocast.net>B Subject: Re: AMD To Announce Microsoft Support For New Chip FamilyC Message-ID: <61Nx8.185127$%l3.14899294@bin8.nnrp.aus1.giganews.com>   : "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message& news:3CC7903E.BD91D257@videotron.ca...   ...   H > Unless the Digital engineers were sold as slaves to Intel, there is no telling G > how many will stay with Intel. (or how long they will stay).   Should  Intel K > decide to revive Alpha, it would take a long time to rebuild the team and  get " > EV8 up and running as a project.  F They reportedly got the team pretty much intact (I heard that only oneH engineer to whom an offer was made didn't accept it - because the offersI were quite attractive).  But unless they had a secret plan right from the L start to continue work on EV8 (the likelihood of which I place out in OliverL Stone territory), they couldn't get it out before 2005 - about the same dateI they could get out an Itanic with on-chip glue that would likely get them H back into the higher-end 64-bit game if they can just hold on until then1 with the sub-par McKinley/Madison/Montecito line.   G And since that 2005 date would be for the existing EV8 design - running K Alpha instructions and using Alpha chip interfaces rather than being in any G way compatible with existing Itanic software and surrounding hardware - L going with the glue-improved Itanic instead is a *lot* more attractive, evenH if the result isn't the leadership product that EV8 would have been.  IfF they want something like an EV8 core executing Itanic instructions andI talking to Itanic support hardware, that pushes them out at least another I year - and the question then becomes whether the very real advantages EV8 H would have had if released in 2004 (as Compaq planned) would still be asL important 2 - 3 years later (when the Itanic core, inefficient as it may be,B may be small enough that several can be placed on a single chip) -E especially if executing an instruction set less closely-suited to the 	 hardware.   K So, again, unless Intel had right from the start a secret plan to adopt the K entire Alpha architecture - instruction set, support chips, etc. - into its K product set, thus having *three* mainstream general-purpose processor lines L (IA32, Itanic, and Alpha) in addition to embedded processors like StrongARM,L Alpha has lost at least a couple of years of lead in all this, and since itsK technical lead was really all it had to work with losing it likely means it G is well and truly toast.  Unless Hammer somehow disintegrates (which is L looking less and less likely, now that they have real working silicon out toK partners), then it, and Yamhill, will likely occupy most of the market that I Itanic was originally targeted for, leaving Itanic at best in a niche for B high-end users who can't afford POWER.  Whether that will generateD sufficient volume to make it worthwhile for Intel (and Microsoft) to6 continue to support it is the main remaining question.   - bill   ------------------------------  % Date: Thu, 25 Apr 2002 10:07:18 +0100 % From: Alan Greig <a.greig@virgin.net> B Subject: Re: AMD To Announce Microsoft Support For New Chip Family8 Message-ID: <pahfcuodf7bj9j8i0l03jmhp9o8e40n9s8@4ax.com>  F On Thu, 25 Apr 2002 06:05:54 GMT, "Bill Todd" <billtodd@metrocast.net> wrote:    L >So, again, unless Intel had right from the start a secret plan to adopt theL >entire Alpha architecture - instruction set, support chips, etc. - into itsL >product set, thus having *three* mainstream general-purpose processor linesM >(IA32, Itanic, and Alpha) in addition to embedded processors like StrongARM, M >Alpha has lost at least a couple of years of lead in all this, and since its L >technical lead was really all it had to work with losing it likely means itH >is well and truly toast.  Unless Hammer somehow disintegrates (which isM >looking less and less likely, now that they have real working silicon out to   ? Well it's official now. MS will support Hammer. Here's what the   infamous Dave Cutler had to say.O http://www.amd.com/us-en/Corporate/VirtualPressRoom/0,,51_104_543~19906,00.html   D  "At Microsoft, our vision for 64-bit computing is a highly scalableB and affordable platform that is easy to deploy, easy to manage andE easy to develop applications for, said Dave Cutler, Sr. Distinguished C Engineer, Microsofts Windows Team, and one of the worlds foremost F software architects. AMDs 8th-generation architecture gives customersD great 32-bit performance and 64-bit capabilities on a single system.D Together, AMDs 8th-generation processors and Windows should provideB customers a flexible platform and a compelling value proposition.  F Looks to me as if Dave Cutler (and the rest of MS) see the AMD chip as? the future of 64 bit computing and not the IA64 chip. Meanwhile F OpenVMS has no long term future on Alpha and depends on the IA64 beingB a success. So did Compaq senior management cock-up by EOLing AlphaC early and announcing an IA64 port for VMS  or did they know exactlypF where this will lead in the end? If Compaq magically conjure up HammerC based systems out of thin air in the next few months I think we canA2 assume they've been working on them for some time.  L >partners), then it, and Yamhill, will likely occupy most of the market thatJ >Itanic was originally targeted for, leaving Itanic at best in a niche forC >high-end users who can't afford POWER.  Whether that will generatepE >sufficient volume to make it worthwhile for Intel (and Microsoft) toe7 >continue to support it is the main remaining question.  >  >- billi >  >    -- Alan   ------------------------------  % Date: Thu, 25 Apr 2002 09:54:12 +0100 % From: Alan Greig <a.greig@virgin.net>8B Subject: Re: AMD To Announce Microsoft Support For New Chip Family8 Message-ID: <nugfcu45bv8e2050s3besn3qsm0mmkk0k2@4ax.com>  A On 24 Apr 2002 19:26:05 -0700, bob@instantwhip.com (Bob Ceculski)  wrote:     >cG >you just gave the evidence ... if EV8 can compete or beat hammer, then.H >Intel will have to use EV8 or go out of business ... if Itanic tanks as8 >you are suggesting above, then there is no alternative!  E You miss the obvious. Intel could use EV8 technology and engineers to C make their own "Hammer" compatible chip and not use them to improver	 the IA64.    -- Alan   ------------------------------    Date: 25 Apr 2002 05:28:42 -0700( From: bob@instantwhip.com (Bob Ceculski)B Subject: Re: AMD To Announce Microsoft Support For New Chip Family= Message-ID: <d7791aa1.0204250428.5cf61159@posting.google.com>e  H "Bill Todd" <billtodd@metrocast.net> wrote in message news:<aVKx8.170816  E why would Intel want to play second fiddle to AMD or Power or anybodyeB when they have the chip that would put them number one?  I predictD Intel will drop the Epic boat anchor and eventually, if not already,C use the Alpha engineers to continue to develop EV8, and then market)@ it ... why be number two, or just compete when you can smoke theD competition?  If they do not, then they are as dum as DEC and Q are!5 And no x86 processor will ever dominate the high end!o   ------------------------------    Date: 25 Apr 2002 05:34:46 -0700( From: bob@instantwhip.com (Bob Ceculski)B Subject: Re: AMD To Announce Microsoft Support For New Chip Family= Message-ID: <d7791aa1.0204250434.3bc83c49@posting.google.com>M  u "Bill Todd" <billtodd@metrocast.net> wrote in message news:<61Nx8.185127$%l3.14899294@bin8.nnrp.aus1.giganews.com>...fN >> So, again, unless Intel had right from the start a secret plan to adopt theM > entire Alpha architecture - instruction set, support chips, etc. - into its M > product set, thus having *three* mainstream general-purpose processor linesaN > (IA32, Itanic, and Alpha) in addition to embedded processors like StrongARM,N > Alpha has lost at least a couple of years of lead in all this, and since itsM > technical lead was really all it had to work with losing it likely means it I > is well and truly toast.  Unless Hammer somehow disintegrates (which iskN > looking less and less likely, now that they have real working silicon out toM > partners), then it, and Yamhill, will likely occupy most of the market that K > Itanic was originally targeted for, leaving Itanic at best in a niche forxD > high-end users who can't afford POWER.  Whether that will generateF > sufficient volume to make it worthwhile for Intel (and Microsoft) to8 > continue to support it is the main remaining question. >  > - bill  G that's what I have been saying all along ... there is a secret plan ...hI Intel knows Epic is a boat anchor and EV8 is the way to go ... again, why!J should Intel play second fiddle to Power or Hammer when they can be numberH one ... the secret plan is on!  I'll bet that's what the 2005 influction point is ...   ------------------------------  # Date: Thu, 25 Apr 2002 13:25:43 GMT 5 From: peter.watkinson1@ntlworld.com (Peter Watkinson)sB Subject: Re: AMD To Announce Microsoft Support For New Chip Family1 Message-ID: <3cc801dc.5799796@news.cable.ntl.com>n  C On Thu, 25 Apr 2002 09:54:12 +0100, Alan Greig <a.greig@virgin.net>e wrote:  B >On 24 Apr 2002 19:26:05 -0700, bob@instantwhip.com (Bob Ceculski) >wrote:  >n >o >>H >>you just gave the evidence ... if EV8 can compete or beat hammer, thenI >>Intel will have to use EV8 or go out of business ... if Itanic tanks as 9 >>you are suggesting above, then there is no alternative!u >HF >You miss the obvious. Intel could use EV8 technology and engineers toD >make their own "Hammer" compatible chip and not use them to improve
 >the IA64. >o >--  >Alan2     Check this out     Clawhammer prices projected    Gigabyte to make stable mobos & By Mike Magee, 25/04/2002 00:38:25 BST    A SOURCES CLOSE TO AMD's plans continue to tell us that the 28th of E October date we're currently projecting appears so far to be spot on. F And we also have projected prices of both processors and motherboards,D with CPUs costing $400 and mobos hovering around the high $200 mark,* as an initial thrust into the marketplace.  ? The reference boards and CPUs are already with the partners, it:B appears, with Clawhammers running at full, rather than half speed.  > When  and it now seems to be a when  Barton is released, the? Thoroughbred will replace the Duron until demand for the Athlont weakens, we understand.   E Then  following ineluctable laws of X86 processors  the Barton willr; be the "Duron" and the "Clawhammer" is the XP branded chip.o  B We believe that our source is also spot on with this observation A that AMD will produce two mobos with a Gigabyte branding, becausehF there still is a perception in the corporate marketplace that the firm: is all too ready to dump high quality and stable chipsets.  D As for SIS, Nvidia and Via, the same source reckons that these mightE be ready for action in the first quarter of next year, but perhaps aso early as January 2003.   @ * MEANWHILE WE HAVE reason to believe that the Kribibench stuff,F mentioned on Ace's Hardware recently, is very close to completion. You) can check out this page for more details.s  E ** YEAH, WE KNOW we should call it the Opteron but we prefer a reader C suggestion, Overthron, just in case Microsoft decides to turn roundt8 and screw both AMD and Intel and release its own CPU...   q -----------------------------------------------------------------------------------------------------------------T    F Maybe that comment on vanshardware.com is true about Intel doing doingB a turnaround in June and announcing Yamhill. As I commented on theF Axp-linux redhat list. I remember back in the 90's when I commented toF anyone or read someone saying, if you suggested buying an Alpha to runC Linux or NT they said why not buy a dual Pentium Pro or dual PII orlC whatever. Maybe purchasers will say the same thing when offered theu: choice to buy an IA64 or Opteron or whatever it's called.   F On a side note and probably the wrong place to ask does anyone know ifF 64bit Windows IA64 or x86-64 will come on the same CD or will there be( separate releases.? Could be confusing.            Peter Watkinson  peter.watkinson1@ntlworld.come   ------------------------------  % Date: Thu, 25 Apr 2002 14:50:52 +0100 % From: Alan Greig <a.greig@virgin.net>eB Subject: Re: AMD To Announce Microsoft Support For New Chip Family8 Message-ID: <ur1gcush4c4qsv42jm7vll5s59pc8bqlue@4ax.com>  A On 25 Apr 2002 05:28:42 -0700, bob@instantwhip.com (Bob Ceculski)e wrote:  I >"Bill Todd" <billtodd@metrocast.net> wrote in message news:<aVKx8.170816e >eF >why would Intel want to play second fiddle to AMD or Power or anybodyC >when they have the chip that would put them number one?  I predictoE >Intel will drop the Epic boat anchor and eventually, if not already,oD >use the Alpha engineers to continue to develop EV8, and then marketA >it ... why be number two, or just compete when you can smoke theR  F There would have been a time when Intel could have done this but if itF was to drop EPIC now where would that leave HP? Where would that leave@ al the OEMs who have thrown money at IA64? Do you think MS wouldC resurrect Alpha support in Windows? Absolutely nothing hinted at by F anbody at all indicates that Intel intends to resurrect Alpha. I don't< even think it has the right to do so even if it wanted to asF intellectual ownership of Alpha remains with Compaq. Capellas was veryB specific about that at the original joint announcement with Intel.D Intel could only resurrect EV8 after further agreements with Compaq.D Do you really think Compaq can go back to its customers now and tell? them "forget everything we said. Alpha is the future not IA64"?1  E >competition?  If they do not, then they are as dum as DEC and Q are! 6 >And no x86 processor will ever dominate the high end!  = Various Alpha designers now at AMD don't agree. Various Alpha . designers now at Intel *might* also not agree. -- Alan   ------------------------------    Date: 25 Apr 2002 07:52:27 -0700( From: bob@instantwhip.com (Bob Ceculski)B Subject: Re: AMD To Announce Microsoft Support For New Chip Family= Message-ID: <d7791aa1.0204250652.4c6bfa77@posting.google.com>   e Alan Greig <a.greig@virgin.net> wrote in message news:<nugfcu45bv8e2050s3besn3qsm0mmkk0k2@4ax.com>...sC > On 24 Apr 2002 19:26:05 -0700, bob@instantwhip.com (Bob Ceculski). > wrote: >  >  > >eI > >you just gave the evidence ... if EV8 can compete or beat hammer, theneJ > >Intel will have to use EV8 or go out of business ... if Itanic tanks as: > >you are suggesting above, then there is no alternative! > G > You miss the obvious. Intel could use EV8 technology and engineers tocE > make their own "Hammer" compatible chip and not use them to improve  > the IA64.   G why do that when they already have the Alpha engineers in tact, the EV8EG blueprints in hand ... that is a waste of more time and money!  And did K anyone ever wonder why there is a beta 3 release of windows 2000 for Alpha?E   ------------------------------  % Date: Thu, 25 Apr 2002 14:57:41 -0000 - From: wspencer@ap.nospam.org (Warren Spencer)LB Subject: Re: AMD To Announce Microsoft Support For New Chip Family7 Message-ID: <91FB60B9Dwarrenspencer1977@209.249.90.100>   ( a.greig@virgin.net (Alan Greig) wrote in. <pahfcuodf7bj9j8i0l03jmhp9o8e40n9s8@4ax.com>:   @ >Well it's official now. MS will support Hammer. Here's what the! >infamous Dave Cutler had to say.EJ >http://www.amd.com/us-en/Corporate/VirtualPressRoom/0,,51_104_543~19906,0 >0.html  >tE > "At Microsoft, our vision for 64-bit computing is a highly scalable C >and affordable platform that is easy to deploy, easy to manage and F >easy to develop applications for, said Dave Cutler, Sr. DistinguishedD >Engineer, Microsofts Windows Team, and one of the worlds foremostG >software architects. AMDs 8th-generation architecture gives customersaE >great 32-bit performance and 64-bit capabilities on a single system.gE >Together, AMDs 8th-generation processors and Windows should provideiC >customers a flexible platform and a compelling value proposition.b >sG >Looks to me as if Dave Cutler (and the rest of MS) see the AMD chip asd6 >the future of 64 bit computing and not the IA64 chip.  
 -- snip --  I I think that word "affordable" in Cutler's comment is telling. One could  L read this as Hammer being the budget-box, leaving wiggle-room for IPF being  the enterprise gorilla.    ws   -- 8   Warren Spencer' Senior Software Engineer (not a writer)e The Associated Press  < ** Time flies like an arrow.  Fruit flies like a bananna. **   ------------------------------  % Date: Thu, 25 Apr 2002 15:56:01 +0100e% From: Alan Greig <a.greig@virgin.net>lB Subject: Re: AMD To Announce Microsoft Support For New Chip Family8 Message-ID: <ov5gcu0aef1jae03belajo8mvs7duo99g4@4ax.com>  F On Thu, 25 Apr 2002 13:25:43 GMT, peter.watkinson1@ntlworld.com (Peter Watkinson) wrote:p    G >Maybe that comment on vanshardware.com is true about Intel doing doingdC >a turnaround in June and announcing Yamhill. As I commented on thel  C Just having a read of vanshardware.com now. Is this the comment youVE refer to or is there more? If MS does drop IA64 then all hope of IA64y& becoming mainstream ends right there.   + Microsoft Guru Throws Weight Behind Hammerst  D As we reported yesterday, Microsoft is committed to supporting AMDs@ x86-64, the open standard instruction set native to Hammer.  KeyF decision makers inside the software company have been so enamored withB x86-64 that the software giant persuaded Intel to adopt the Hammer	 language.e  @ One of most avid proponents of x86-64 is the legendary MicrosoftD programmer responsible for the NT kernel.  David Cutler, formerly of< DEC, has reportedly voiced inside Microsoft extremely strong> preferences for x86-64 over Intel IA-64. Allegedly, Cutler andF Microsoft do not want to expend the extra resources to produce ongoingA support of IA-64, an instruction set Cutler and others inside thee. software giant reportedly disdain as inferior.     -- Alan   ------------------------------  % Date: Thu, 25 Apr 2002 16:17:29 +0100u% From: Alan Greig <a.greig@virgin.net>aB Subject: Re: AMD To Announce Microsoft Support For New Chip Family8 Message-ID: <0b7gcuokrlkkce4kk3node2855eiefec9v@4ax.com>  A On 25 Apr 2002 07:52:27 -0700, bob@instantwhip.com (Bob Ceculski)a wrote:   >oH >why do that when they already have the Alpha engineers in tact, the EV8H >blueprints in hand ... that is a waste of more time and money!  And didL >anyone ever wonder why there is a beta 3 release of windows 2000 for Alpha?  D Because it hadn't been dropped by that point. Beta 3 wasn't even theF final version shipped to customers. Release Candidates 1 and 2 shipped? prior to Compaq announcing they would not support W2K on Alpha.t  E Compaq dropped support for W2K on Alpha *after* all the work had beene9 done and just weeks before it was scheduled for release.     -- Alan   ------------------------------  # Date: Thu, 25 Apr 2002 15:49:10 GMT 5 From: peter.watkinson1@ntlworld.com (Peter Watkinson)iB Subject: Re: AMD To Announce Microsoft Support For New Chip Family2 Message-ID: <3cc82538.14852562@news.cable.ntl.com>  C On Thu, 25 Apr 2002 15:56:01 +0100, Alan Greig <a.greig@virgin.net>t wrote:  G >On Thu, 25 Apr 2002 13:25:43 GMT, peter.watkinson1@ntlworld.com (Peter- >Watkinson) wrote: >  >eH >>Maybe that comment on vanshardware.com is true about Intel doing doingD >>a turnaround in June and announcing Yamhill. As I commented on the >sD >Just having a read of vanshardware.com now. Is this the comment youF >refer to or is there more? If MS does drop IA64 then all hope of IA64' >becoming mainstream ends right there.   >r, >Microsoft Guru Throws Weight Behind Hammers >tE >As we reported yesterday, Microsoft is committed to supporting AMDs A >x86-64, the open standard instruction set native to Hammer.  KeyAG >decision makers inside the software company have been so enamored with C >x86-64 that the software giant persuaded Intel to adopt the Hammer 
 >language. >cA >One of most avid proponents of x86-64 is the legendary MicrosoftnE >programmer responsible for the NT kernel.  David Cutler, formerly ofh= >DEC, has reportedly voiced inside Microsoft extremely strongw? >preferences for x86-64 over Intel IA-64. Allegedly, Cutler and G >Microsoft do not want to expend the extra resources to produce ongoingdB >support of IA-64, an instruction set Cutler and others inside the/ >software giant reportedly disdain as inferior.  >a >g >--c >Alan      http://www.vanshardware.coma   from 23rd 04 02t    & Microsoft is Committed to AMDs x86-64  E Sources attending Seattles WinHEC 2002 maintained that Microsoft hase? successfully pressured Goliath chipmaker Intel Corporation into8D adopting archrival AMDs new 64-bit chip language.  According to ourC confidential sources, the Redmond, Washington-based software vendora? expressed a clear preference for Advanced Micro Devices x86-64pB instruction set over Intels competing IA-64.  Furthermore, the OSA developer advised Intel to comply with x86-64 if the Santa Clara,4: California-based chipmaker desired ongoing 64-bit support.    l  F Used in AMDs upcoming Hammer line of 64-bit processors, x86-64 is a@ straightforward yet much more powerful evolution of the existingE 32-bit x86 instruction set.  Seen as the blood enemy of Intels IA64,eB a CPU lingo introduced with the flagging Merced Itanium, IntelsD adoption of AMDs x86-64 serves as a resounding victory for the much smaller MPU designer.       t  D Despite being severely handicapped in terms of relative size, AMD is? widely viewed as the superior innovator.  In fact, AMD has been F awarded more U.S. patents than Intel for each of the last three years.C With a total of 1,090 patents in 2001, AMD ranked 14th in the worldtD and 5th in America in number of awarded U.S. patents.  Intel trailed0 with only 811 U.S. patents during the same year.    -  F Our sources also allege that top Microsoft decision makers view x86-64C as the clearly superior solution over IA64, an underperforming VLIWtE architecture widely judged as Byzantine.  Intel is expected to make amC formal announcement of its x86-64 support, code-named Yamhill, in(= June or July during a face-saving presentation delivered with 
 Microsoft.    e   Peter Watkinsone peter.watkinson1@ntlworld.com    ------------------------------  # Date: Thu, 25 Apr 2002 17:08:28 GMTt* From: "Bill Todd" <billtodd@metrocast.net>B Subject: Re: AMD To Announce Microsoft Support For New Chip FamilyC Message-ID: <gKWx8.174554$3L2.15333010@bin7.nnrp.aus1.giganews.com>4  2 "Alan Greig" <a.greig@virgin.net> wrote in message2 news:ur1gcush4c4qsv42jm7vll5s59pc8bqlue@4ax.com...   ...B  H > There would have been a time when Intel could have done this but if itH > was to drop EPIC now where would that leave HP? Where would that leave, > al the OEMs who have thrown money at IA64?  H As I've said before, even Intel can't afford to make *too* many enemies.    Do you think MS would% > resurrect Alpha support in Windows?e  E Rumor has it that it may never have died in the first place.  But theaH problem now is Hammer (and maybe Yammer):  while Alpha emulation of IA32K binaries looked pretty good compared with Itanic's, it doesn't look good ate4 all compared with Hammer's native execution of them.  I Alpha was important to MS when there was no other good 64-bit platform in H sight (and I suspect they may have come to regret axing Win64 on Alpha).E With Hammer (and likely Yammer) ready in the wings, *no* other 64-bito* platform may look significant to them now.     Absolutely nothing hinted at by@ > anbody at all indicates that Intel intends to resurrect Alpha.  L True.  And while Compaq had every reason to leverage Alpha, Intel's case forH doing so (better basic platform) is severely compromised by its existing1 ties to two different incompatible architectures.o  J If Itanic sinks (soon, and incontrovertibly), *and* Yamhill can't fill theJ gap enough that Intel doesn't feel it needs another platform (i.e., that aF sufficiently large high-end niche remains that Yamhill can't address),G *then* Intel would have a business case for resurrecting Alpha.  I'd be K filled with joy to see that happen, but I don't expect it to even if it maynH make some sense:  I just don't expect Intel to concede defeat for ItanicG without much more of a fight, nor even if that happened for there to be I enough of a high-end niche left that Yammer can't fill to justify anotheri, product line from a volume-oriented company.    I don't> > even think it has the right to do so even if it wanted to asH > intellectual ownership of Alpha remains with Compaq. Capellas was veryD > specific about that at the original joint announcement with Intel.  K I don't think he was:  IIRC what he said was that Compaq retained the right.F to develop Alpha, but that all such rights were also licensed to IntelK (i.e., my impression is that Intel could indeed complete EV8 and sell it ifu it wished to).  F > Intel could only resurrect EV8 after further agreements with Compaq.F > Do you really think Compaq can go back to its customers now and tellA > them "forget everything we said. Alpha is the future not IA64"?c  I The only real reason Compaq couldn't would be lack of a future for Alpha,cI since the only real commitments Compaq customers have made because of thelL Intel deal so far are likely in the area of placing as much distance betweenI themselves and Compaq as they can.  If Intel actually did embrace EV8 andvI future Alpha generations, this would be the best of all worlds for CompaqeJ users:  a guaranteed hardware platform from a far more reliable owner thanJ Compaq has been, plus real pressure on Compaq to support the OSs that make that platform fly.   - bill   ------------------------------  % Date: Thu, 25 Apr 2002 07:42:50 -0600 0 From: "Mike Ober" <obermd.@.alum.mit.edu.nospam> Subject: Re: AOL0 Message-ID: <vJTx8.11$Sm3.18443@news.uswest.net>  D A lot of telecommunication companies in the US will have losses thisB quarter.  One of the Enron fall outs is that so called "good will"J investments and assets can only be valued for a limited period of time notJ and not indefinitely as they were before.  In reality, it makes absolutelyL no sense to place a monetary value on "good will" when it comes to balancingL your books, simply because if someone wants to buy your business to break it, up, they can't sell "good will" as an asset.  
 Mike Ober.  B "Peter Watkinson" <peter.watkinson1@ntlworld.com> wrote in message, news:3cc71e2f.18845453@news.cable.ntl.com... >e >h >t >eF > http://news.bbc.co.uk/hi/english/business/newsid_1949000/1949651.stm >r > he hea >n >l > Peter Watkinsona > peter.watkinson1@ntlworld.comn   ------------------------------   Date: 25 Apr 2002 10:54:29 GMT From: vanja@vmsbox.cjb.net Subject: AST questionaD Message-ID: <3cc7e065$0$10247$afc38c87@sisyphus.news.be.easynet.net>   Hi.r  D After reading a bit of OpenVMS documentation, it is my understandingD that ASTs are similar (remotely, but still 'similar' :) to 'signals' on UNIX.  C During an execution of some for(;;) loop (for example), AST routineaG can be invoked (for whatever reason), and when routine is done with itstE work, execution control is returned to the function that invoked AST.l  & That means that ASTs are blocking, no?  E The project I want to 'port' to OpenVMS could use the technique whichn# was suggested to me, and goes like:   . task1: does a bit, then queues task2 as an ast. task2: does a bit, then queues task3 as an astC task3: does a bit then deallocates the structures for that client. -  @ Now - working this way would make things look like they're beingC executed in sync, but in reality, they're not. Problem is that thisiF tool will be doing some virus scanning. Now, until the point of actualG scanning, everything is fine. But when a routine is invoked that has tonF scan 15MB ZIP archive, it's going to take a while, and all other tasks+ will be blocked while scanning is going on.m  E As an example - SYS$EXAMPLES:DB_SERVER.C shows usage of ASTs. For thewI sake of example, let's assume that process_request() has sleep(30) in it. > Would it mean that everything would be blocked for 30 seconds?  H So, does it mean I should skip ASTs, and look at another alternative? :)  D Also - I've played with sys$creprc a bit, and I also realized that IE can't actually use it either. Problem is that 'parent' process has to I initialize virus patterns, and keep them in memory all the time. There isnD a function (let's call it "scan_file") which then takes advantage ofJ patterns being already loaded in memory, and does the actual file scanningE very quickly (as opposed to having to initialize virus patterns everylG time a scan is required). Having each 'child' initialize virus patterns D beats the purpose of making this tool then, since one could just use- scanning tool supplied by Sophos (AV vendor).n  E It is my understanding that a process created with sys$creprc can notlC access to data that was 'initialized' by the parent. Is that right?   > Does it mean that using threads would be the only solution? :)   Thanks in advance.   Vanja    ------------------------------  # Date: Thu, 25 Apr 2002 12:05:34 GMTd# From: Jonas Lindholm <jlhm@usa.net>P Subject: Re: AST questionl' Message-ID: <3CC7F1DE.F627EBF0@usa.net>D   Hi!h  J I suggest you use threads instead of AST. With VMS 7.n you also can createN kernel threads so on a SMP machine two or more threads can execute at the same time.k  N If you want to create sub processes and share data between them and the parentL you can use global sections and map the section to each process thus sharing data between them.  G For any of these methods you need to synchronize the access between themL threads using mutex or using locks for processes. As long they only read theK data no synchronization is needed but every time someone need to update thenK data in the global section you need to protect the section so no one try tos# read the same data that is updated.   	 /Jonas L.h     vanja@vmsbox.cjb.net wrote:e   > Hi.  >hF > After reading a bit of OpenVMS documentation, it is my understandingF > that ASTs are similar (remotely, but still 'similar' :) to 'signals'
 > on UNIX. >BE > During an execution of some for(;;) loop (for example), AST routineiI > can be invoked (for whatever reason), and when routine is done with its G > work, execution control is returned to the function that invoked AST.b >3( > That means that ASTs are blocking, no? >:G > The project I want to 'port' to OpenVMS could use the technique which % > was suggested to me, and goes like:8 >n0 > task1: does a bit, then queues task2 as an ast0 > task2: does a bit, then queues task3 as an astD > task3: does a bit then deallocates the structures for that client. > B > Now - working this way would make things look like they're beingE > executed in sync, but in reality, they're not. Problem is that thisvH > tool will be doing some virus scanning. Now, until the point of actualI > scanning, everything is fine. But when a routine is invoked that has to H > scan 15MB ZIP archive, it's going to take a while, and all other tasks- > will be blocked while scanning is going on.- >-G > As an example - SYS$EXAMPLES:DB_SERVER.C shows usage of ASTs. For thegK > sake of example, let's assume that process_request() has sleep(30) in it.i@ > Would it mean that everything would be blocked for 30 seconds? > J > So, does it mean I should skip ASTs, and look at another alternative? :) >aF > Also - I've played with sys$creprc a bit, and I also realized that IG > can't actually use it either. Problem is that 'parent' process has torK > initialize virus patterns, and keep them in memory all the time. There isrF > a function (let's call it "scan_file") which then takes advantage ofL > patterns being already loaded in memory, and does the actual file scanningG > very quickly (as opposed to having to initialize virus patterns everysI > time a scan is required). Having each 'child' initialize virus patternseF > beats the purpose of making this tool then, since one could just use/ > scanning tool supplied by Sophos (AV vendor).A >sG > It is my understanding that a process created with sys$creprc can notiE > access to data that was 'initialized' by the parent. Is that right?  >D@ > Does it mean that using threads would be the only solution? :) >6 > Thanks in advance. >h > Vanjai   ------------------------------  % Date: Thu, 25 Apr 2002 14:43:21 -0000e- From: wspencer@ap.nospam.org (Warren Spencer)h Subject: Re: AST questione7 Message-ID: <91FB6A7B3warrenspencer1977@209.249.90.100>-  8 vanja@vmsbox.cjb.net wrote in <3cc7e065$0$10247$afc38c87 @sisyphus.news.be.easynet.net>:m   >Hi. >eE >After reading a bit of OpenVMS documentation, it is my understanding-E >that ASTs are similar (remotely, but still 'similar' :) to 'signals'i	 >on UNIX.t > D >During an execution of some for(;;) loop (for example), AST routine  
 -- snip --  J The application description you provide sounds like a very good match for L threads.  You can choose from pthreads and DECthreads - not sure if there's  others.n  K You can certainly achieve your goals using native OpenVMS services as well nL - AST's, locks, global sections, event flags, and mailboxes would likely be # useful allies if you go this route.    ws   -- e   Warren Spencer' Senior Software Engineer (not a writer)g The Associated Press  < ** Time flies like an arrow.  Fruit flies like a bananna. **   ------------------------------  + Date: Thu, 25 Apr 2002 15:01:43 +0000 (UTC)n1 From: sssslewis@spyder.mitre.org (Keith A. Lewis)e Subject: Re: AST questionf. Message-ID: <aa95on$993$1@newslocal.mitre.org>  K You obviously know a good deal about software engineering.  You should read K the "OpenVMS Programming Concepts" manual.  Specifically, the section aboutoI ASTs will answer your AST questions (they block each other if they're theoI same mode, etc.).  Also read about locks and event flags, which are greatAE for synchronizing processes.  And if you want to share memory betweene. processes, global sections will let you do it.  6 The manuals are at http://www.openvms.compaq.com:8000/   vanja@vmsbox.cjb.net writes in article <3cc7e065$0$10247$afc38c87@sisyphus.news.be.easynet.net> dated 25 Apr 2002 10:54:29 GMT:mE >After reading a bit of OpenVMS documentation, it is my understandingoE >that ASTs are similar (remotely, but still 'similar' :) to 'signals'e	 >on UNIX..  + --Keith Lewis              klewis$mitre.orga> The above may not (yet) represent the opinions of my employer.   ------------------------------    Date: 25 Apr 2002 09:54:03 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen)A Subject: Re: AST question 3 Message-ID: <py8PM4FgUlcZ@eisner.encompasserve.org>h  a In article <3cc7e065$0$10247$afc38c87@sisyphus.news.be.easynet.net>, vanja@vmsbox.cjb.net writes:c  G > The project I want to 'port' to OpenVMS could use the technique whiche% > was suggested to me, and goes like:o > 0 > task1: does a bit, then queues task2 as an ast0 > task2: does a bit, then queues task3 as an astE > task3: does a bit then deallocates the structures for that client.   > B > Now - working this way would make things look like they're being0 > executed in sync, but in reality, they're not.  F Ada Tasking handles this quite well, switching between tasks accordingF to rules that you specify.  Other languages can simulate this behavior by calling DECthreads.  E > executed in sync, but in reality, they're not. Problem is that thiseH > tool will be doing some virus scanning. Now, until the point of actualI > scanning, everything is fine. But when a routine is invoked that has tohH > scan 15MB ZIP archive, it's going to take a while, and all other tasks- > will be blocked while scanning is going on.r  H At the 15MB level, I would say it is appropriate to use another process.  F > Also - I've played with sys$creprc a bit, and I also realized that IG > can't actually use it either. Problem is that 'parent' process has to6K > initialize virus patterns, and keep them in memory all the time. There iseF > a function (let's call it "scan_file") which then takes advantage ofL > patterns being already loaded in memory, and does the actual file scanningG > very quickly (as opposed to having to initialize virus patterns everyeI > time a scan is required). Having each 'child' initialize virus patternsoF > beats the purpose of making this tool then, since one could just use/ > scanning tool supplied by Sophos (AV vendor).o > G > It is my understanding that a process created with sys$creprc can notoE > access to data that was 'initialized' by the parent. Is that right?n  E Rather than storing the data in process-private memory, have it storevD the data in a "Global Section" create with SYS$CRMPSC or a relative.7 Have the child access it with SYS$MGBLSC or a relative.n  @ > Does it mean that using threads would be the only solution? :)  B Generally on VMS one does not run into problems for which there isC only one solution.  Feel free to discuss it here, though, and folksoF can comment regarding why one solution might be preferable to another.   ------------------------------  % Date: Thu, 25 Apr 2002 12:07:41 -0400o( From: David Froble <davef@tsoft-inc.com> Subject: Re: AST questiona, Message-ID: <3CC829CD.8050302@tsoft-inc.com>   vanja@vmsbox.cjb.net wrote:    > Hi.M > F > After reading a bit of OpenVMS documentation, it is my understandingF > that ASTs are similar (remotely, but still 'similar' :) to 'signals'
 > on UNIX. > E > During an execution of some for(;;) loop (for example), AST routinegI > can be invoked (for whatever reason), and when routine is done with itsdG > work, execution control is returned to the function that invoked AST.     * Personal opinion on, others may not agree.  O An AST delivery should be to announce to your application that some async work oQ you set up to happen has completed.  The AST code then would do the minimal work  J necessary to cause the application to do whatever it needs to do with the  completed async work.s  M As an example, you queue up multiple (one or more) reads on multiple (one or oQ more) channels.  When a particular read is completed, and the data in the buffer pL waiting for you to do something with it, the AST is delivered, and your AST M routine executes.  It would then determine the channel on which the read has  F completed and set appropriate flags to alert the application.  If the Q application was idle, waiting for some work to do, then it would begin work upon nM completion of the AST routine.  If it was already doing work, then it would, 2F based upon application design, do the highest priority available work.  O The key is, you would NOT do the work in the AST routine.  This would prohibit 0D the application from working on the highest priority available work.    ( > That means that ASTs are blocking, no?    # When queued at the same level, yes.     G > The project I want to 'port' to OpenVMS could use the technique whicho% > was suggested to me, and goes like:h > 0 > task1: does a bit, then queues task2 as an ast0 > task2: does a bit, then queues task3 as an astE > task3: does a bit then deallocates the structures for that client. u    O AST routines are not jobs, they are to signal an event, such as the completion i3 of async work.  Don't do your work in AST routines.e    B > Now - working this way would make things look like they're beingE > executed in sync, but in reality, they're not. Problem is that thisrH > tool will be doing some virus scanning. Now, until the point of actualI > scanning, everything is fine. But when a routine is invoked that has tooH > scan 15MB ZIP archive, it's going to take a while, and all other tasks- > will be blocked while scanning is going on.     J That's why you don't do the work in the AST routine.  You set appropriate O information to allow the application to know that there is work to do, and let r= the application determine which work to do and in what order.b  Q Of course, while working on tasks, additional ASTs can be delivered, and the AST tI routine(s) can set additional flags for the application.  It's up to the  Q application to check the communication mechanisms and to determine what needs to m be done, and in what order.l  L If using threads, a new thread can be started to perform a task when an AST O delivery announces that there is work to be done, and the normal scheduling by aO the OS will perform the work in parallel.  If not using threads, or some other pN like mechanism, then the application must prioritize the work, and perform it 
 sequentially.   N Ok, you may find others using ASTs in a different manner.  I'm sure that many Q things are possible.  The above is my understanding of the 'proper' use of ASTs, i if there is such a thing.e  M I've seen applications where things are initialized, and then the process is lL idle unless executing an AST routine, and all work is done in AST routines. N Such seem to work.  I guess you design an application to work in this manner. 6 It's just not my understanding of the purpose of ASTs.    G > As an example - SYS$EXAMPLES:DB_SERVER.C shows usage of ASTs. For theeK > sake of example, let's assume that process_request() has sleep(30) in it.r@ > Would it mean that everything would be blocked for 30 seconds?     Yep!    J > So, does it mean I should skip ASTs, and look at another alternative? :)    . No, just keep the AST routine short and quick.    F > Also - I've played with sys$creprc a bit, and I also realized that IG > can't actually use it either. Problem is that 'parent' process has toiK > initialize virus patterns, and keep them in memory all the time. There is F > a function (let's call it "scan_file") which then takes advantage ofL > patterns being already loaded in memory, and does the actual file scanningG > very quickly (as opposed to having to initialize virus patterns everylI > time a scan is required). Having each 'child' initialize virus patternsiF > beats the purpose of making this tool then, since one could just use/ > scanning tool supplied by Sophos (AV vendor).n    N Not so.  Create a global section with the patterns, and have each process map L the global section.  It's shared memory.  Read only makes it much less of a M problem.  There's no loading involved once the global secion is set up.  The nQ 'parent' process doesn't even have to be kept around, if you don't need it.  The  Q global section will remain until the last connection to it is closed, if you set oK it up that way.  The code to create and load the global section could be a tK called routine that is invoked if the global section doesn't already exist.e    G > It is my understanding that a process created with sys$creprc can notnE > access to data that was 'initialized' by the parent. Is that right?      Not at all..    @ > Does it mean that using threads would be the only solution? :)    ( On VMS there are usually many solutions.     > Thanks in advance. >  > Vanja- >  >  >    Dave   ------------------------------   Date: 25 Apr 2002 16:51:23 GMT1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon)u' Subject: Re: Best DIGITAL product ever?e, Message-ID: <aa9c6b$11v5$2@info.cs.uofs.edu>  ; In article <HcEx8.43511$%s3.18066148@typhoon.ne.ipsvc.net>,i4  "Terry C. Shannon" <terryshannon@attbi.com> writes: |> m1 |> "John Smith" <a@nonymous.com> wrote in messageiE |> news:39Ex8.56905$VLV.50595@news01.bloor.is.net.cable.rogers.com...  |> > COLT 45 |> > . |> > .
 |> > Capellas  |> > Onk	 |> > Line  |> > TerminatorS |> >! |> > For the good of the company.t |> pG |> Ironically or otherwise, IBM made Colt .45s (actually Model 1911 .45g; |> automatic pistols) for the US Govt. during World War II.   F Long after that.  I was qualified on the Colt Automatic up through theE first half of the 80's.  I know they were looking at alternatives anduD did in fact have a 38 revolver for women MP's, but the basic sidearm' was still the 45 when I go out in 1984.o   bill   -- gJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   n   ------------------------------    Date: 25 Apr 2002 07:35:19 -0500B From: clubley@remove_me.eisner.decus.org-Earth.UFP (Simon Clubley). Subject: Building Emacs 19.28 on Alpha VMS 7.23 Message-ID: <P1Ai4$2rbH1s@eisner.encompasserve.org>s  	 Versions:T 	Alpha running VMS 7.2
 	DEC C 6.0 	MMK from the Freeware V4 CD  J Has anyone successfully built the Emacs 19.28 kit from the Freeware V4 kit% using the above or similar versions ?   H I am trying to build this kit and I am running into various errors whileH building, both with compiler warnings and the DESCRIP.MMS files as well.  H Note that, at the moment, instead of my listing the various errors here,I I am more interested in finding out if anyone has successfully built thissJ kit and what was involved, as I would like to know if I am at the start ofK a long path to get it working, where I fix one set of problems only to havep a new set appear.v   Thanks for any information,i   Simon.   --  B Simon Clubley, clubley@remove_me.eisner.decus.org-Earth.UFP       + Microsoft: The Lada of the computing world.    ------------------------------    Date: 25 Apr 2002 07:51:01 -0500B From: clubley@remove_me.eisner.decus.org-Earth.UFP (Simon Clubley)2 Subject: Re: Building Emacs 19.28 on Alpha VMS 7.23 Message-ID: <9y9GwXi8Os1Z@eisner.encompasserve.org>C  x In article <P1Ai4$2rbH1s@eisner.encompasserve.org>, clubley@remove_me.eisner.decus.org-Earth.UFP (Simon Clubley) writes: > L > Has anyone successfully built the Emacs 19.28 kit from the Freeware V4 kit' > using the above or similar versions ?  > J > I am trying to build this kit and I am running into various errors whileJ > building, both with compiler warnings and the DESCRIP.MMS files as well.'                                ^^^^^^^^wH Oops, I meant compiler errors. _For now_, in an attempt to get it build,D I've been able to convert the errors that I've encountered so far toH warnings by using /STANDARD=VAXC. I am now getting stopped with an errorI from MMK, which is making me wonder if it's possible to build Emacs usinghI MMK instead of MMS, so rather than start listing what may be the first ofOJ many errors, I am interested in finding out if someone has been successful* in building it with VMS 7.2/DEC C 6.0/MMK.   Thanks for any information,t   Simon.   -- ;B Simon Clubley, clubley@remove_me.eisner.decus.org-Earth.UFP       + Microsoft: The Lada of the computing world.    ------------------------------  % Date: Thu, 25 Apr 2002 09:37:11 -0500 C From: "Craig A. Berry" <craig.berry@nospam.SignalTreeSolutions.com> 2 Subject: Re: Building Emacs 19.28 on Alpha VMS 7.2H Message-ID: <craig.berry-A53D8C.09370525042002@news.directvinternet.com>  3 In article <P1Ai4$2rbH1s@eisner.encompasserve.org>,3D  clubley@remove_me.eisner.decus.org-Earth.UFP (Simon Clubley) wrote:   > 	MMK from the Freeware V4 CD  @ That's bound to be pretty ancient.  Try getting the latest from H www.madgoat.com.  There are some differences from MMS and it's possible @ you'll have to do a bit of porting work to the description file.   ------------------------------  # Date: Thu, 25 Apr 2002 14:38:49 GMTM From: system@SendSpamHere.ORGk2 Subject: Re: Building Emacs 19.28 on Alpha VMS 7.20 Message-ID: <00A0CFCD.D2A211F1@SendSpamHere.ORG>  x In article <P1Ai4$2rbH1s@eisner.encompasserve.org>, clubley@remove_me.eisner.decus.org-Earth.UFP (Simon Clubley) writes:
 >Versions: >	Alpha running VMS 7.2h >	DEC C 6.0b >	MMK from the Freeware V4 CDo >oK >Has anyone successfully built the Emacs 19.28 kit from the Freeware V4 kito& >using the above or similar versions ? > I >I am trying to build this kit and I am running into various errors whilenI >building, both with compiler warnings and the DESCRIP.MMS files as well.l >eI >Note that, at the moment, instead of my listing the various errors here, J >I am more interested in finding out if anyone has successfully built thisK >kit and what was involved, as I would like to know if I am at the start ofeL >a long path to get it working, where I fix one set of problems only to have >a new set appear. >u >Thanks for any information, >i >Simon.o >  >-- C >Simon Clubley, clubley@remove_me.eisner.decus.org-Earth.UFP       u, >Microsoft: The Lada of the computing world.      H You won't get far with DEC C 6.0.  Emacs relies on the C preprocessor toJ generate many of its build files.  Changes in the C preprocessor will makeJ this impossible to build.  Later versions of the C compiler V6.4 have someH corrections which will make this build compete.  Also, make sure you useH /STANDARD=VAXC or some other preprocessor oddities will impede building.   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COM             sJ   "And of course, I'm a genius, so people are naturally drawn to my fiery I   intellect.  Their admiration overwhelms their envy!" -- Calvin & Hobbesm   ------------------------------  % Date: Thu, 25 Apr 2002 18:15:31 +0200)% From: "Andrew Clark" <clarka@post.ch>s/ Subject: CGI access problem using CSWS (Apache) % Message-ID: <3cc82ba3$1@news.post.ch>e  ) Hi, I hope this is not a stupid question,e  L I am porting some intranet pages from Purveyor to CSWS (Apache), the HTML is= working fine, however we have a problem with the cgi scripts,c   in the Error_Log.n   ...o8 [Thu Apr 25 17:50:49 2002] [error] (1)not owner: exec of. /web_root/testweb/cgi-bin/debug_cgi.exe failed ...i  # the file ownership seems fine. e.g.e   ...*4 $  dir/owner WEB_ROOT:[TESTWEB.CGI-BIN]DEBUG_CGI.EXE  $ Directory WEB_ROOT:[TESTWEB.CGI-BIN]  * DEBUG_CGI.EXE;1      [AP_HTTPD,APACHE$WWW]   Total of 1 file. ...m   and the processes seem oku    $  show sys /proc=APACHE$* /fullE OpenVMS V7.2-1  on node WEBSERVER  25-APR-2002 18:12:46.94  Uptime  3a 10:12:42         AlphaServer 2100 4/200F   Pid    Process Name    State  Pri      I/O       CPU       Page flts PagesoF 26A02E51 APACHE$00       LEF      5     1515   0 00:00:11.38       612 625           [AP_HTTPD,APACHE$WWW] 5000KbF 26A02E74 APACHE$00006    LEF      5      944   0 00:00:10.74       511 590w          [AP_HTTPD,APACHE$WWW] 4720KbF 26A02E75 APACHE$00007    LEF      6      940   0 00:00:05.13       510 589.          [AP_HTTPD,APACHE$WWW] 4712KbF 26A02E76 APACHE$00001    LEF      6      938   0 00:00:08.51       510 589a          [AP_HTTPD,APACHE$WWW] 4712KbF 26A02E77 APACHE$00002    LEF      6      941   0 00:00:03.87       510 589e          [AP_HTTPD,APACHE$WWW] 4712KbF 26A0267B APACHE$00003    LEF      6      949   0 00:00:07.57       510 589           [AP_HTTPD,APACHE$WWW] 4712Kb $o    
 So any ideas?    thanks in advance-   Andrew  J BTW: If we crank up the privilages of the web user, everything works fine,K but then we would probably have the least secure VMS machine on the planet!s   ------------------------------  % Date: Thu, 25 Apr 2002 12:53:32 -0400t1 From: Michael Austin <maustin@firstdbasource.com>c3 Subject: Re: CGI access problem using CSWS (Apache) 2 Message-ID: <3CC8348C.11148340@firstdbasource.com>   Andrew Clark wrote:e > + > Hi, I hope this is not a stupid question,e > N > I am porting some intranet pages from Purveyor to CSWS (Apache), the HTML is? > working fine, however we have a problem with the cgi scripts,  >  > in the Error_Log.c >  > ...c: > [Thu Apr 25 17:50:49 2002] [error] (1)not owner: exec of0 > /web_root/testweb/cgi-bin/debug_cgi.exe failed > ...a > % > the file ownership seems fine. e.g.e >  > ...i6 > $  dir/owner WEB_ROOT:[TESTWEB.CGI-BIN]DEBUG_CGI.EXE > & > Directory WEB_ROOT:[TESTWEB.CGI-BIN] > , > DEBUG_CGI.EXE;1      [AP_HTTPD,APACHE$WWW] >  > Total of 1 file. > ...n >  > and the processes seem okn > " > $  show sys /proc=APACHE$* /fullG > OpenVMS V7.2-1  on node WEBSERVER  25-APR-2002 18:12:46.94  Uptime  3i
 > 10:12:42  >         AlphaServer 2100 4/200H >   Pid    Process Name    State  Pri      I/O       CPU       Page flts > Pages-H > 26A02E51 APACHE$00       LEF      5     1515   0 00:00:11.38       612 > 625u  >          [AP_HTTPD,APACHE$WWW] > 5000KbH > 26A02E74 APACHE$00006    LEF      5      944   0 00:00:10.74       511 > 590e  >          [AP_HTTPD,APACHE$WWW] > 4720KbH > 26A02E75 APACHE$00007    LEF      6      940   0 00:00:05.13       510 > 589d  >          [AP_HTTPD,APACHE$WWW] > 4712KbH > 26A02E76 APACHE$00001    LEF      6      938   0 00:00:08.51       510 > 589q  >          [AP_HTTPD,APACHE$WWW] > 4712KbH > 26A02E77 APACHE$00002    LEF      6      941   0 00:00:03.87       510 > 589e  >          [AP_HTTPD,APACHE$WWW] > 4712KbH > 26A0267B APACHE$00003    LEF      6      949   0 00:00:07.57       510 > 589o  >          [AP_HTTPD,APACHE$WWW] > 4712Kb > $e >  > So any ideas?t >  > thanks in advance  >  > Andrew > L > BTW: If we crank up the privilages of the web user, everything works fine,M > but then we would probably have the least secure VMS machine on the planet!e  D What does your httpd.conf file look like with regards to CGI. Do you! have this defined as exeutable?    something like:   - ScriptAlias /cgi-bin/ "/apache$root/cgi-bin/"d  " <Directory "/apache$root/cgi-bin">     AllowOverride None     Options None     Order allow,deny     Allow from all </Directory>  F Make sure Apache$WWW has at least execute ACL on all directories above cgi-bin.  7 What is the error that is returned on the failing page?0   -- f Regards,  7 Michael Austin            Registered Linux User #261163d7 First DBA Source, Inc.    http://www.firstdbasource.comg Sr. Consultant 704-947-1089 (Office)t 704-236-4377 (Mobile)'   ------------------------------    Date: 25 Apr 2002 06:07:06 -0700( From: bob@instantwhip.com (Bob Ceculski)? Subject: CIA warns of China cyber attacks ... better be on VMS!o= Message-ID: <d7791aa1.0204250507.23aefaff@posting.google.com>A  4 CIA Warns of Chinese Plans for Cyber-Attacks on U.S.B  Defense: Analysts fear government and private efforts to sabotage federal Internet sites.t    p        m% By ERIC LICHTBLAU, Times Staff Writer     F WASHINGTON -- U.S. intelligence officials believe the Chinese military= is working to launch wide-scale cyber-attacks on American ande? Taiwanese computer networks, including Internet-linked militaryhD systems considered vulnerable to sabotage, according to a classified CIA report.   E Moreover, U.S. authorities are bracing for a possible wave of hackingtF attacks by Chinese students against the United States in coming weeks,E according to the analysis. The confidential alert, which was reviewedh< by The Times, was sent to intelligence officials a week ago.  E Although U.S. officials have voiced concerns about individual hackersdC in China who have defaced federal and private Web sites, the United D States has resisted publicly linking the Chinese government to those* attacks or to broader cyber-style warfare.  ? The new CIA report, however, makes clear that U.S. intelligencew? analysts have become increasingly concerned that authorities ineA Beijing are actively planning to damage and disrupt U.S. computerpA systems through the use of Internet hacking and computer viruses.e  E Although the assessment concludes that China has not yet acquired thedA technical sophistication to do broad damage to U.S. and TaiwaneseaF systems, it maintains that this is the "intended goal" of the People's@ Liberation Army in China. "The mission of Chinese special forces= includes physical sabotage" of vulnerable systems, the reportnD says--which some analysts said is driven by China's hostility toward Taiwan.n  C The Chinese Embassy in Washington insisted Wednesday, however, that = Beijing is only conducting computer research that is strictlyr defensive in nature.  B "It is not the Chinese government's policy to disrupt the computer? system of any other country," said Larry Wu, an official in the ) embassy's science and technology section.d  E "We do research on the security of computers, of course--self-defenseoF to understand how a hacker can get into our computer systems so we canE defend it," he said. "But China has never assumed an offensive stance,% with regards to computer technology."s  E But several specialists in Chinese security and military affairs said D the CIA's conclusions jibe with their own observations about China's+ research into offensive-minded cyber-tools.o  D "We should be very worried about this issue," said James Mulvenon, aA China analyst at the Rand Corp. think tank who has done extensiveh+ studies into Chinese computer capabilities.l  E Taiwan, which China regards as a renegade province, appears to be the,A driving force behind the Chinese interest in hacking and viruses,iD Mulvenon said. Under one scenario, if China were to make good on itsF long-standing threat to invade Taiwan, the Chinese military could thenC seek to deploy widespread computer disruptions against American and ? Taiwanese military systems to slow any effort by U.S. forces to ' intervene in Taiwan's defense, he said.l  B The issue threatens to inflame what are invariably tense relationsF between the United States and the Communist regime in China, relationsD already frayed by a volley of charges and counter charges during the? last several years over alleged nuclear, military and political 
 espionage.  C Relations hit a low point last year after a U.S. spy plane collided)E with a Chinese jet fighter, triggering an international standoff overdB the return of the plane's 24 Navy crewmen. China detained the crewE members for 11 days and returned the disassembled plane months later.u  : Recent months have seen a warming in relations as the BushC administration secured China's cooperation in the war on terrorism.p? But China has become upset by what it sees as the White House'sm/ increasingly favorable overtures toward Taiwan.(  F The CIA's assessment discusses Taiwan and the United States, revealing@ that U.S. intelligence officials believe both are targets of the Chinese military.s   ------------------------------  % Date: Thu, 25 Apr 2002 09:12:02 -0400)* From: WILLIAM WEBB <WWEBB1@email.usps.gov>C Subject: RE: CIA warns of China cyber attacks ... better be on VMS!u- Message-ID: <0033000061511761000002L012*@MHS>v  ) =0AI could have said this in seven words:n  # "All your system are belong to us."s   WWWebb   -----Original Message-----/ From: Info-VAX-Request@Mvb.Saic.Com at INTERNETb& Sent: Thursday, April 25, 2002 9:07 AMB To: Webb, William W Raleigh, NC; Info-VAX@Mvb.Saic.Com at INTERNET? Subject: CIA warns of China cyber attacks ... better be on VMS!n    4 CIA Warns of Chinese Plans for Cyber-Attacks on U.S.B  Defense: Analysts fear government and private efforts to sabotage federal Internet sites.l    % By ERIC LICHTBLAU, Times Staff Writerc    F WASHINGTON -- U.S. intelligence officials believe the Chinese military= is working to launch wide-scale cyber-attacks on American andi? Taiwanese computer networks, including Internet-linked militaryoD systems considered vulnerable to sabotage, according to a classified CIA report.u  E Moreover, U.S. authorities are bracing for a possible wave of hacking F attacks by Chinese students against the United States in coming weeks,E according to the analysis. The confidential alert, which was reviewedB< by The Times, was sent to intelligence officials a week ago.  E Although U.S. officials have voiced concerns about individual hackersiC in China who have defaced federal and private Web sites, the UnitednD States has resisted publicly linking the Chinese government to those* attacks or to broader cyber-style warfare.  ? The new CIA report, however, makes clear that U.S. intelligence ? analysts have become increasingly concerned that authorities inpA Beijing are actively planning to damage and disrupt U.S. computer1A systems through the use of Internet hacking and computer viruses.   E Although the assessment concludes that China has not yet acquired theeA technical sophistication to do broad damage to U.S. and Taiwanese F systems, it maintains that this is the "intended goal" of the People's@ Liberation Army in China. "The mission of Chinese special forces= includes physical sabotage" of vulnerable systems, the report D says--which some analysts said is driven by China's hostility toward Taiwan.n  C The Chinese Embassy in Washington insisted Wednesday, however, thats= Beijing is only conducting computer research that is strictly  defensive in nature.  B "It is not the Chinese government's policy to disrupt the computer? system of any other country," said Larry Wu, an official in theP) embassy's science and technology section.e  E "We do research on the security of computers, of course--self-defenseDF to understand how a hacker can get into our computer systems so we canE defend it," he said. "But China has never assumed an offensive stancet% with regards to computer technology."r  E But several specialists in Chinese security and military affairs saideD the CIA's conclusions jibe with their own observations about China's+ research into offensive-minded cyber-tools.d  D "We should be very worried about this issue," said James Mulvenon, aA China analyst at the Rand Corp. think tank who has done extensiveg+ studies into Chinese computer capabilities.s  E Taiwan, which China regards as a renegade province, appears to be theoA driving force behind the Chinese interest in hacking and viruses,.D Mulvenon said. Under one scenario, if China were to make good on itsF long-standing threat to invade Taiwan, the Chinese military could thenC seek to deploy widespread computer disruptions against American andi? Taiwanese military systems to slow any effort by U.S. forces ton' intervene in Taiwan's defense, he said.i  B The issue threatens to inflame what are invariably tense relationsF between the United States and the Communist regime in China, relationsD already frayed by a volley of charges and counter charges during the? last several years over alleged nuclear, military and politicaln
 espionage.  C Relations hit a low point last year after a U.S. spy plane collidedCE with a Chinese jet fighter, triggering an international standoff over B the return of the plane's 24 Navy crewmen. China detained the crewE members for 11 days and returned the disassembled plane months later.a  : Recent months have seen a warming in relations as the BushC administration secured China's cooperation in the war on terrorism.l? But China has become upset by what it sees as the White House'sw/ increasingly favorable overtures toward Taiwan.e  F The CIA's assessment discusses Taiwan and the United States, revealing@ that U.S. intelligence officials believe both are targets of the Chinese military.=   ------------------------------  % Date: Thu, 25 Apr 2002 10:06:36 -0400f1 From: Michael Austin <maustin@firstdbasource.com>-C Subject: Re: CIA warns of China cyber attacks ... better be on VMS!"2 Message-ID: <3CC80D6C.FA7D0310@firstdbasource.com>   WILLIAM WEBB wrote:  > ( > I could have said this in seven words: > % > "All your system are belong to us."e >  > WWWebb >  > -----Original Message-----1 > From: Info-VAX-Request@Mvb.Saic.Com at INTERNET-( > Sent: Thursday, April 25, 2002 9:07 AMD > To: Webb, William W Raleigh, NC; Info-VAX@Mvb.Saic.Com at INTERNETA > Subject: CIA warns of China cyber attacks ... better be on VMS!s > 6 > CIA Warns of Chinese Plans for Cyber-Attacks on U.S.D >  Defense: Analysts fear government and private efforts to sabotage > federal Internet sites.: <snip>  > I have had to include a number of APAC IP address blocks in my; BAD-CLIENTS list.  I got tired of dealing with the garbage.H   -- > Regards,  7 Michael Austin            Registered Linux User #261163e7 First DBA Source, Inc.    http://www.firstdbasource.comV Sr. Consultant 704-947-1089 (Office)E 704-236-4377 (Mobile)e   ------------------------------  % Date: Thu, 25 Apr 2002 08:01:30 -0700o# From: "Tom Linden" <tom@kednos.com>oC Subject: RE: CIA warns of China cyber attacks ... better be on VMS! 9 Message-ID: <CIEJLCMNHNNDLLOOGNJIOEFJENAA.tom@kednos.com>m  = BTW, do you need to restart SMTP when you modify smtp.config?t   >   -----Original Message-----< >   From: Michael Austin [mailto:maustin@firstdbasource.com]* >   Sent: Thursday, April 25, 2002 7:07 AM >   To: Info-VAX@Mvb.Saic.ComsG >   Subject: Re: CIA warns of China cyber attacks ... better be on VMS!r >    >    >   WILLIAM WEBB wrote:  >   > , >   > I could have said this in seven words: >   > ) >   > "All your system are belong to us."y >   >  >   > WWWebb >   >   >   > -----Original Message-----5 >   > From: Info-VAX-Request@Mvb.Saic.Com at INTERNETe, >   > Sent: Thursday, April 25, 2002 9:07 AMH >   > To: Webb, William W Raleigh, NC; Info-VAX@Mvb.Saic.Com at INTERNETE >   > Subject: CIA warns of China cyber attacks ... better be on VMS!m >   > : >   > CIA Warns of Chinese Plans for Cyber-Attacks on U.S.H >   >  Defense: Analysts fear government and private efforts to sabotage >   > federal Internet sites.-
 >   <snip> >   B >   I have had to include a number of APAC IP address blocks in my? >   BAD-CLIENTS list.  I got tired of dealing with the garbage.l >    >   -- i >   Regards, >   ; >   Michael Austin            Registered Linux User #261163 ; >   First DBA Source, Inc.    http://www.firstdbasource.com: >   Sr. Consultant >   704-947-1089 (Office). >   704-236-4377 (Mobile)r >      ------------------------------  % Date: Thu, 25 Apr 2002 11:11:57 -0400t1 From: Michael Austin <maustin@firstdbasource.com> C Subject: Re: CIA warns of China cyber attacks ... better be on VMS!i2 Message-ID: <3CC81CBD.29AF3D02@firstdbasource.com>   <snip>   Tom Linden wrote:  > ? > BTW, do you need to restart SMTP when you modify smtp.config?e >    Yes.& I usually execute my restart_smtp.com:  n! $@sys$startup:tcpip$smtp_shutdown  $wait 00:00:02 $@sys$startup:tcpipsmtp_startupE   -- e Regards,  7 Michael Austin            Registered Linux User #261163A7 First DBA Source, Inc.    http://www.firstdbasource.come Sr. Consultant 704-947-1089 (Office)u 704-236-4377 (Mobile)n   ------------------------------  % Date: Thu, 25 Apr 2002 18:01:06 +0200S- From: Didier Morandi <Didier.Morandi@Free.fr>eC Subject: Re: CIA warns of China cyber attacks ... better be on VMS! ' Message-ID: <3CC82841.3BA9E4F4@Free.fr>    Next Clancy book?    D.   Bob Ceculski wrote:0 > 6 > CIA Warns of Chinese Plans for Cyber-Attacks on U.S.D >  Defense: Analysts fear government and private efforts to sabotage > federal Internet sites.    ------------------------------  % Date: Thu, 25 Apr 2002 16:50:28 +02007, From: "Andreas Bohn" <abohn@rzb-hamburg.com>! Subject: closed user groups / x25a6 Message-ID: <1019746227.809949@rzbgw1.rzb-hamburg.com>  $ OpenVMS AXP V7.0, DecNet/OSI Phase V  
 Hi everybody,   J trying to configure "closed user groups" I experience the problem that -asA soon as I use the 'closed_user_group' qualifier the PAD announcesAI "PAD-F-NOLINES, no lines are available for the call"- no matter whether I : specify a CUG and no matter what host I try to connect to.  : Every attempt like  set host <dte> /x29 /closed[=cug_name]J fails with the same error message while (the same) connections without the. closed qualifier are successfully established.  6 Has anybody got a clue how this problem can be solved?( Any help would be very much appreciated.   ------------------------------    Date: 25 Apr 2002 10:09:45 -0700, From: bubbapig@hotmail.com (Jeffrey Cameron)= Subject: Re: creat function hanging -- Code snippet provided!p= Message-ID: <b22333b7.0204250909.45348b64@posting.google.com>i  S Actually you can ... it works fine for the 5000+ other files I have run through it.]  2 I'll give it a try though, it sure can't hurt! :-)    Y John Forkosh <john@invalid.com> wrote in message news:<aa6paj$kdj$1@reader1.panix.com>... / > Jeffrey Cameron <bubbapig@hotmail.com> wrote:E9 > : I would like to thank you all for your advice so far.G >  ;5 > : The syntax I am passing to the creat function is   >   I > : filedesc = creat(temp, 0777, "alq=36", "mrs=length*4+4", "ctx=xplct",  > : "rfm=var");n > ; > I'm guessing you can't do arithmetic in "mrs=length*4+4".  > Try this instead:  >  > char mrs_string[99];* > sprintf(mrs_string,"mrs=%d",length*4+4);M > filedesc = creat(temp, 0777, "alq=36", mrs_string, "ctx=xplct", "rfm=var");0   ------------------------------  # Date: Thu, 25 Apr 2002 13:28:45 GMT6# From: "John N." <JNixon@cfl.rr.com>  Subject: Disk storage for DS10= Message-ID: <hwTx8.81127$nc.11701655@typhoon.tampabay.rr.com>A  H We have a cluster that contains (among other things, which will be goingI away soon),  a DS10 and an AS2100A.  These two are connected to an HSZ70.   I Due to maintenance costs of the 2100 and the HSZ, I would like to replace K the 2100 with something like a second DS10 or a DS20.  It could possibly be $ an ES40, but that might be overkill.  K These systems are used by our development group.  I need something that can.L support multiple (about 30) , fairly heavy users.  (They are used to workingK on a VAX 4705, so almost any new Alpha will be an improvement).  I need forkF both systems to have their own system disk so that we can keep them on different VMS versions.   K What would be a good, cost effective, shared  storage solution so that each L system could see all storage, even if one node is down?  They currently haveL about 75GBs, with about 20GB free, but I would like to multiply that several times.   ------------------------------  + Date: Thu, 25 Apr 2002 15:40:01 +0100 (MET)l9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com>e Subject: Disk storage for DS10; Message-ID: <01KGZHVZOBDW8Y7GCV@sysdev.deutsche-boerse.com>h  J > We have a cluster that contains (among other things, which will be goingK > away soon),  a DS10 and an AS2100A.  These two are connected to an HSZ70.r > K > Due to maintenance costs of the 2100 and the HSZ, I would like to replace-M > the 2100 with something like a second DS10 or a DS20.  It could possibly beb& > an ES40, but that might be overkill. > M > These systems are used by our development group.  I need something that cantN > support multiple (about 30) , fairly heavy users.  (They are used to workingM > on a VAX 4705, so almost any new Alpha will be an improvement).  I need forsH > both systems to have their own system disk so that we can keep them on > different VMS versions.f > M > What would be a good, cost effective, shared  storage solution so that each N > system could see all storage, even if one node is down?  They currently haveN > about 75GBs, with about 20GB free, but I would like to multiply that several > times.  D Why not shadow all disks with one member on each node?  No need for F fancy controllers, the disk will still be visible if one node is down H and you will be protected against a disk crash.  (Would your developers D be happy losing the work they do between backups if a disk crashes?)   ------------------------------  % Date: Thu, 25 Apr 2002 16:35:42 +0200o) From: Bart Zorn <B.Zorn@xs4all.nospam.nl>," Subject: Re: Disk storage for DS10/ Message-ID: <3CC8143E.2000104@xs4all.nospam.nl>t   Phillip Helbig wrote:-J >>We have a cluster that contains (among other things, which will be goingK >>away soon),  a DS10 and an AS2100A.  These two are connected to an HSZ70.j >>K >>Due to maintenance costs of the 2100 and the HSZ, I would like to replacehM >>the 2100 with something like a second DS10 or a DS20.  It could possibly bec& >>an ES40, but that might be overkill. >>M >>These systems are used by our development group.  I need something that can N >>support multiple (about 30) , fairly heavy users.  (They are used to workingM >>on a VAX 4705, so almost any new Alpha will be an improvement).  I need forlH >>both systems to have their own system disk so that we can keep them on >>different VMS versions.a >>M >>What would be a good, cost effective, shared  storage solution so that eachjN >>system could see all storage, even if one node is down?  They currently haveN >>about 75GBs, with about 20GB free, but I would like to multiply that several >>times. >  > F > Why not shadow all disks with one member on each node?  No need for H > fancy controllers, the disk will still be visible if one node is down J > and you will be protected against a disk crash.  (Would your developers F > be happy losing the work they do between backups if a disk crashes?) >   I Although this might be a good solution, it requires careful planning and eC system management. For instance, you always have to consider which tC system has the most recent version of the shadow sets, as this one dB should be the first to reboot in case of a entire cluster reboot. I Another issue is that shadow copy and merge operations can saturate your  E cluster interconnect. Remember that these kind of operations use SCS c5 interconnects, so a shared SCSI bus will not be used.h  	 Bart Zorno   ------------------------------  % Date: Thu, 25 Apr 2002 10:49:53 -0400d1 From: Michael Austin <maustin@firstdbasource.com> " Subject: Re: Disk storage for DS102 Message-ID: <3CC81791.4C42AFBA@firstdbasource.com>   "John N." wrote: > J > We have a cluster that contains (among other things, which will be goingK > away soon),  a DS10 and an AS2100A.  These two are connected to an HSZ70.e > K > Due to maintenance costs of the 2100 and the HSZ, I would like to replacenM > the 2100 with something like a second DS10 or a DS20.  It could possibly bee& > an ES40, but that might be overkill. > M > These systems are used by our development group.  I need something that can N > support multiple (about 30) , fairly heavy users.  (They are used to workingM > on a VAX 4705, so almost any new Alpha will be an improvement).  I need for4H > both systems to have their own system disk so that we can keep them on > different VMS versions.s > M > What would be a good, cost effective, shared  storage solution so that eachiN > system could see all storage, even if one node is down?  They currently haveN > about 75GBs, with about 20GB free, but I would like to multiply that several > times.  G A DS20 with an additional external disk array would do very nicely.. orgH you could even look at mulitple  DS10L (at $2K+ licenses 1u box with oneE internal disk)+PCI FiberChannel to a SAN or 2 DS10L+SCSI StorageWorks2 shelf using 80Gb disks.     E There are a million options it is all going to depend on how much yout want to spend. -- n Regards,  7 Michael Austin            Registered Linux User #261163h7 First DBA Source, Inc.    http://www.firstdbasource.coml Sr. Consultant 704-947-1089 (Office)  704-236-4377 (Mobile)o   ------------------------------  + Date: Thu, 25 Apr 2002 17:06:18 +0000 (UTC)o1 From: sssslewis@spyder.mitre.org (Keith A. Lewis)m" Subject: Re: Disk storage for DS10. Message-ID: <aa9d2a$9n8$1@newslocal.mitre.org>   Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> writes in article <01KGZHVZOBDW8Y7GCV@sysdev.deutsche-boerse.com> dated Thu, 25 Apr 2002 15:40:01 +0100 (MET):N >> What would be a good, cost effective, shared  storage solution so that eachO >> system could see all storage, even if one node is down?  They currently havenO >> about 75GBs, with about 20GB free, but I would like to multiply that severals	 >> times.e >lE >Why not shadow all disks with one member on each node?  No need for .G >fancy controllers, the disk will still be visible if one node is down eI >and you will be protected against a disk crash.  (Would your developers aE >be happy losing the work they do between backups if a disk crashes?)e  E Shadowing is a good idea if you have the disks to spare, but a 2-nodevH cluster will not provide automatic failover without a quorum disk (which@ must be visible by both nodes, possibly on a shared SCSI bus).    L You can set up an asymetric cluster with one "master" node having more votesL than the other, then if the slave goes down failover is automatic but if theH master goes down the slave will lose quorum and hang -- you can bring itI back, but you have to reset it physically.  One solution is to have equal I work nodes and a quorum cluster member (cheap, maybe a DEC 3000) to breako the tie.  + --Keith Lewis              klewis$mitre.orgo> The above may not (yet) represent the opinions of my employer.   ------------------------------  # Date: Thu, 25 Apr 2002 16:19:17 GMTn5 From: peter.watkinson1@ntlworld.com (Peter Watkinson)  Subject: download VMSa2 Message-ID: <3cc82bbb.16519531@news.cable.ntl.com>   Hi,a  F I was wondering if it would be a good idea for Compaq (HP?) to let youE download VMS through the internet. Particularly the Hobbyist version.lE With Linux and BSD it so easy to download and install via FTP so long B as you have a fast connection. I don't think it would be neccesaryF with windows as a 5min drive to the local PC world can pick up a copy.8 Could be a good idea to bring VMS into the 21st century.    e   Peter Watkinsone peter.watkinson1@ntlworld.como   ------------------------------  + Date: Thu, 25 Apr 2002 16:50:55 +0000 (UTC) * From: bdwheele@indiana.edu (Brian Wheeler) Subject: Re: download VMSs2 Message-ID: <aa9c5f$odd$1@wilson.uits.indiana.edu>  2 In article <3cc82bbb.16519531@news.cable.ntl.com>,8 	peter.watkinson1@ntlworld.com (Peter Watkinson) writes: >  > Hi,e > H > I was wondering if it would be a good idea for Compaq (HP?) to let youG > download VMS through the internet. Particularly the Hobbyist version. G > With Linux and BSD it so easy to download and install via FTP so longsD > as you have a fast connection. I don't think it would be neccesaryH > with windows as a 5min drive to the local PC world can pick up a copy.: > Could be a good idea to bring VMS into the 21st century. >  >    >  > Peter Watkinson  > peter.watkinson1@ntlworld.comr  J I think it would be great, especially if they included some of the layeredG products.  Since you have to have a license anyway to use the binaries,n; making the media freely available can't hurt license sales.    Briane   ------------------------------  % Date: Thu, 25 Apr 2002 09:53:09 -0700m# From: "Tom Linden" <tom@kednos.com>e Subject: RE: download VMSh9 Message-ID: <CIEJLCMNHNNDLLOOGNJICEGCENAA.tom@kednos.com>e  ; sun does it.  We make the PL/I kits available for downloadsi   >   -----Original Message-----5 >   From: Brian Wheeler [mailto:bdwheele@indiana.edu]r* >   Sent: Thursday, April 25, 2002 9:51 AM >   To: Info-VAX@Mvb.Saic.ComN >   Subject: Re: download VMS  >w >e6 >   In article <3cc82bbb.16519531@news.cable.ntl.com>,< >   	peter.watkinson1@ntlworld.com (Peter Watkinson) writes: >   > 	 >   > Hi,i >   >nL >   > I was wondering if it would be a good idea for Compaq (HP?) to let youK >   > download VMS through the internet. Particularly the Hobbyist version.tK >   > With Linux and BSD it so easy to download and install via FTP so long,H >   > as you have a fast connection. I don't think it would be neccesaryL >   > with windows as a 5min drive to the local PC world can pick up a copy.> >   > Could be a good idea to bring VMS into the 21st century. >   >- >   >c >   >C >   > Peter Watkinson # >   > peter.watkinson1@ntlworld.comg >DB >   I think it would be great, especially if they included some of >   the layeredsK >   products.  Since you have to have a license anyway to use the binaries,W? >   making the media freely available can't hurt license sales.M >S	 >   BrianI >R   ------------------------------  % Date: Thu, 25 Apr 2002 10:19:44 -0700g% From: Dean Woodward <deanw@rdrop.com>T Subject: Re: download VMSh) Message-ID: <3CC83AB0.6C9C0672@rdrop.com>    Brian Wheeler wrote: > 4 > In article <3cc82bbb.16519531@news.cable.ntl.com>,A >         peter.watkinson1@ntlworld.com (Peter Watkinson) writes:t > >  > > Hi,  > >eJ > > I was wondering if it would be a good idea for Compaq (HP?) to let youI > > download VMS through the internet. Particularly the Hobbyist version.tI > > With Linux and BSD it so easy to download and install via FTP so longpF > > as you have a fast connection. I don't think it would be neccesaryJ > > with windows as a 5min drive to the local PC world can pick up a copy.< > > Could be a good idea to bring VMS into the 21st century. > >R > >A > >s > > Peter Watkinsonk! > > peter.watkinson1@ntlworld.comD > L > I think it would be great, especially if they included some of the layeredI > products.  Since you have to have a license anyway to use the binaries,m= > making the media freely available can't hurt license sales.   G Being able to run VMSINSTAL and specify a URL as the kit location wouldcG be sweet!  And, as stated in another thread, CONDIST kits are available G next to free on ebay now, and it takes a slew of licenses to make it go C in any case, so it's not entirely like there's a huge concern about  unlicensed VMS running around.  @ What about making the enough of the hobbyist kit to get a systemF bootstrapped (Say, the VMS base CD? I installed off of work's CONDIST,G so I haven't even a poor, soggy clue how the hobbyist kit is organized)CG and connected to the net available for download as an ISO images?  That C would be enough to keep a lot of beginners happy for some time, I'dlC think.  Is there an official opinion on the copyright implications?n   ------------------------------    Date: 25 Apr 2002 02:22:49 -0700$ From: issinoho@slayme.com (issinoho) Subject: Re: DSSI Interconnect= Message-ID: <d0141774.0204250122.64a924c7@posting.google.com>   A Well spotted, my goggle-eyed friend, not many people remember it!s  Y Nic Clews <sendspamhere@127.0.0.1> wrote in message news:<3CC69869.3E2501E4@127.0.0.1>...f > issinoho wrote:- > > C > > I have a 2-node cluster where each Alpha is connected to an HSD I > > storage unit with a single DSSI interconnect. The 2 storage units areuH > > also connected directly together (presumably using DSSI) to create a > > daisy-chain effect.f > > G > > My question is: does this scenario provide for direct access to alleG > > disks in both storage units by both Alpha systems? i.e. if one nodeyF > > failed will the remaining node still see all disks in both storage
 > > units? > F > Answer is, it depends. It depends on the DSSI controller. I'm fairlyG > sure that all DSSI controllers can see disks when VMS has been booteds > (note the use of past tense).4 > H > However some controllers cannot show you the devices at console level,I > and this also means you cannot boot from an Alpha system disk connectediD > to the DSSI. However that does not rule out network booting for anI > Alpha, then you could 'see' the disks. However I don't think that would  > help you in your case. > J > This is off topic as far as your question is concerned, but if you had aH > VAX with a DSSI connection to the Alpha system disk, you could networkJ > boot the Alpha from the VAX, and when VMS is far enough up on the Alpha,H > it could form its own direct 'physical' connection to the disks rather > than being MSCP served.  >  7F > > My need for clarification arises because the Cluster Config manualI > > only describes scenarios where each node is connected to both storagee' > > units using multiple interconnects.h > G > The SPD (software product description) for clustering should identifyaH > which are supported and which are not. The manual tends to be a little
 > generic. > 
 > > Issinoho.t >  > BTW, isn't this from   >  >   thIS SIde up
 > use NO HOksa >  > Kids UK TV from 1970's?    ------------------------------    Date: 25 Apr 2002 03:12:28 -0700$ From: issinoho@slayme.com (issinoho) Subject: Re: DSSI Interconnect< Message-ID: <d0141774.0204250212.ac94ba9@posting.google.com>  D Doing a SHO PORT in SDA, I have a PA and PE driver running, althoughB further analysis shows that all cluster interconnects are runnning through the PE driver.  7 Why should this be? Surely DSSI is the preferred route?eE Should I be worried about this? Will it affect anything you mentionedI1 previously regarding the visibility over the bus?,  s hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) wrote in message news:<_NCx8.21$jx5.568374@cacnews.cac.cpqcorp.net>... f > In article <d0141774.0204240239.57e4be43@posting.google.com>, issinoho@slayme.com (issinoho) writes: > F > :My question is: does this scenario provide for direct access to all5 > :disks in both storage units by both Alpha systems?4 > H >   Yes, assuming a functional DSSI bus and functional DSSI controllers. > D >   DSSI functions both as a cluster communications bus -- except inD >   the specific case of KFQSA controllers -- and a cluster storage E >   bus.  The former means host-to-host communications are possible, -F >   and the latter indicates multi-host host-to-storage is possible.   > F >   The KFQSA Q-bus to DSSI is limited to storage access.  Most (all?)E >   other DSSI controllers provide communications and storage access.n > G >   Contrast DSSI with SCSI: SCSI is not a cluster communications bus, 2 >   but is a storage bus.t >  > P >  ---------------------------- #include <rtfaq.h> -----------------------------P >       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    P >  --------------------------- pure personal opinion ---------------------------N >    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  + Date: Thu, 25 Apr 2002 14:36:05 +0100 (MET)e9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com>o, Subject: EDIT/FDL changes colours in DECterm; Message-ID: <01KGZFOM4R4Y8Y7GCV@sysdev.deutsche-boerse.com>e  H A standard-colour DECterm changes to black background, white foreground G during EDIT/FDL and is not reset after exiting.  Is this a known bug?  f( (VMS 7.3, most stuff rather up to date.)   ------------------------------    Date: 25 Apr 2002 01:58:16 -0700) From: P.Young@unsw.EDU.AU (Patrick Young)b Subject: Re: EDT or EVEh= Message-ID: <55f85d77.0204250058.1e93f258@posting.google.com>K  l "Sandeep Yelwatkar" <Sandeep_Yelwatkar@bmc.com> wrote in message news:<uce00t53bua477@corp.supernews.com>... > HiH >     I am new to the OpenVMS system. Just wondering which is the better9 > editor EVE or EDT. Would like to know your suggesstion.e >   C You need to know both really, or at least EVE and TECO, if not EDT.0  B This is because you will still need to edit if you don't have a VTF compatible terminal - ie: booted minimum system and need to edit files, from the console. EDT and TECO will do this.  F EVE is more powerful and is my choice (with EDT keypad enabled) simplyE because you can define learn sequences and bind them to a key, and its+ supports longer than 255 character records.   I You will want to add TPU SET(MOUSE,OFF) to your F$TRNLNM("EVE$INIT") filefE before you even look at using it, otherwise you will be driven %$^@!#sH *INSANE* by not being able to cut and paste using the mouse - who's idea was this anyway?  B I never did work out why EDT does not have LEARN where as KED did.   ------------------------------  % Date: Thu, 25 Apr 2002 09:44:41 -0400t1 From: Michael Austin <maustin@firstdbasource.com>  Subject: Re: EDT or EVE 2 Message-ID: <3CC80849.36E8F9BD@firstdbasource.com>   Jerry Leslie wrote:e > 6 > Sandeep Yelwatkar (Sandeep_Yelwatkar@bmc.com) wrote: > : HiJ > :     I am new to the OpenVMS system. Just wondering which is the better; > : editor EVE or EDT. Would like to know your suggesstion.- > :3
 > : Thanks > : Sandeeps > :lC > What operating systems/editors are you already comfortable with ?gC > If *nix, then there are versions of 'vi' available for VMS. Using0D > a familiar editor in a strange operating system will let you focus > on that operating system.  > G > I use Charles Sandmann's EDT-like ED editor on VMS, *nix, and WindowsSE > (DOS box) because I've been primarily a VMS person since the 1980s.  > For more info on ED, see:u >  >   http://clio.rice.edu/m > F > ED is free, and comes with the source. It includes an ftp client forA > editing files across a TCP/IP network, as well as a newsreader.  > J > --Jerry Leslie   leslie@clio.rice.edu  (my opinions are strictly my own); >   Note: leslie@jrlvax.houston.rr.com is invalid for email>  G I like NuTPU for Unix/NT.  this way I get to use a real editor in Unix.j  1 http://www.advsyscon.com/products/nuTPU/nuTPU.asp  --   Regards,  7 Michael Austin            Registered Linux User #261163d7 First DBA Source, Inc.    http://www.firstdbasource.como Sr. Consultant 704-947-1089 (Office)a 704-236-4377 (Mobile)    ------------------------------  % Date: Thu, 25 Apr 2002 10:42:55 -0400 1 From: Michael Austin <maustin@firstdbasource.com>o Subject: Re: EDT or EVEe2 Message-ID: <3CC815EF.DB0DD10E@firstdbasource.com>   Patrick Young wrote: > n > "Sandeep Yelwatkar" <Sandeep_Yelwatkar@bmc.com> wrote in message news:<uce00t53bua477@corp.supernews.com>... > > HiJ > >     I am new to the OpenVMS system. Just wondering which is the better; > > editor EVE or EDT. Would like to know your suggesstion.r > >s > E > You need to know both really, or at least EVE and TECO, if not EDT.t > D > This is because you will still need to edit if you don't have a VTH > compatible terminal - ie: booted minimum system and need to edit files. > from the console. EDT and TECO will do this. > H > EVE is more powerful and is my choice (with EDT keypad enabled) simplyG > because you can define learn sequences and bind them to a key, and itF- > supports longer than 255 character records.  > K > You will want to add TPU SET(MOUSE,OFF) to your F$TRNLNM("EVE$INIT") filey  E I have been looking for something like this when using Motif DECTerm.   H Yes this is turning into a holy war of biblical proportions... :)  I useB both and EDT keypad when in TPU.  They accomplish the same task in different ways when  necessary.a  ! one last comment on EVE features:t  : command: what line         !tells you what line you are onF command: set cursor bound  !make the cursor behave the same as in EDT. -- w Regards,  7 Michael Austin            Registered Linux User #261163n7 First DBA Source, Inc.    http://www.firstdbasource.com  Sr. Consultant 704-947-1089 (Office)f 704-236-4377 (Mobile)    ------------------------------    Date: 25 Apr 2002 08:36:48 -0700. From: SPAMSINK2001@YAHOO.COM (Alan E. Feldman) Subject: Re: EDT or EVEe< Message-ID: <343f30ae.0204250736.252fa54@posting.google.com>  n P.Young@unsw.EDU.AU (Patrick Young) wrote in message news:<55f85d77.0204250058.1e93f258@posting.google.com>...n > "Sandeep Yelwatkar" <Sandeep_Yelwatkar@bmc.com> wrote in message news:<uce00t53bua477@corp.supernews.com>... [snip]H > EVE is more powerful and is my choice (with EDT keypad enabled) simplyG > because you can define learn sequences and bind them to a key, and itn- > supports longer than 255 character records.  > K > You will want to add TPU SET(MOUSE,OFF) to your F$TRNLNM("EVE$INIT") filecG > before you even look at using it, otherwise you will be driven %$^@!# J > *INSANE* by not being able to cut and paste using the mouse - who's idea > was this anyway? > D > I never did work out why EDT does not have LEARN where as KED did.  E Why doesn't EDT's Ctrl/K count as a LEARN equivalent? Press ^K, pressyF the key you want to define, press the key sequence you want memorized,D press the period key, then press Enter. Viola! You've defined a key.  3 And please pardon my ignorance, what is or was KED?l   Disclaimer: JMHO Alan E. Feldmane" afeldman atski gfigroup dotski com   ------------------------------  + Date: Thu, 25 Apr 2002 17:41:17 +0100 (MET)e9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com>  Subject: Re: EDT or EVE ; Message-ID: <01KGZM5PCK9U8Y7GCV@sysdev.deutsche-boerse.com>l  D > I never did work out why EDT does not have LEARN where as KED did. > G > Why doesn't EDT's Ctrl/K count as a LEARN equivalent? Press ^K, pressdH > the key you want to define, press the key sequence you want memorized,F > press the period key, then press Enter. Viola! You've defined a key.  H Although an EDT fan, I have not spent much time on ^K.  I have tried it F a few times (as following your example), but get "invalid subcommand".  F A few folks have mentioned here "I use the EDT keypad in EVE and thus H have both".  This gives you JUST the EDT keypad; there is a lot more to  EDT.   ------------------------------  % Date: Thu, 25 Apr 2002 17:52:01 +0200a) From: Bart Zorn <B.Zorn@xs4all.nospam.nl>n Subject: Re: EDT or EVE / Message-ID: <3CC82621.1030704@xs4all.nospam.nl>i   Alan E. Feldman wrote:p > P.Young@unsw.EDU.AU (Patrick Young) wrote in message news:<55f85d77.0204250058.1e93f258@posting.google.com>... > n >>"Sandeep Yelwatkar" <Sandeep_Yelwatkar@bmc.com> wrote in message news:<uce00t53bua477@corp.supernews.com>... >  > [snip] > H >>EVE is more powerful and is my choice (with EDT keypad enabled) simplyG >>because you can define learn sequences and bind them to a key, and itu- >>supports longer than 255 character records.i >>K >>You will want to add TPU SET(MOUSE,OFF) to your F$TRNLNM("EVE$INIT") fileyG >>before you even look at using it, otherwise you will be driven %$^@!#oJ >>*INSANE* by not being able to cut and paste using the mouse - who's idea >>was this anyway? >>D >>I never did work out why EDT does not have LEARN where as KED did. >  > G > Why doesn't EDT's Ctrl/K count as a LEARN equivalent? Press ^K, press H > the key you want to define, press the key sequence you want memorized,F > press the period key, then press Enter. Viola! You've defined a key. > 5 > And please pardon my ignorance, what is or was KED?t  C The BIG difference is that with TPU you can see what you are doing  C during the learn sequence. With EDT you have to do it in the blind.S  F KED was the predecessor of EDT on RT-11 and RSX. RSTS/E never had KED   but did have EDT V1.0 and later.  	 Bart ZornT   ------------------------------  % Date: Thu, 25 Apr 2002 17:58:14 +0200e- From: Didier Morandi <Didier.Morandi@Free.fr>a Subject: Re: EDT or EVEc' Message-ID: <3CC82795.8D06B2DC@Free.fr>    Patrick Young wrote: >  ../.. A > EVE is more powerful and is my choice (with EDT keypad enabled)t   Mwouaaaaaahhh !!!!!!!!!!!f MDR, LOL, I die laughing...y   :-))   D.   ------------------------------  % Date: Thu, 25 Apr 2002 11:58:53 -0400e1 From: Michael Austin <maustin@firstdbasource.com>  Subject: Re: EDT or EVE-2 Message-ID: <3CC827BD.6181EC1E@firstdbasource.com>   "Alan E. Feldman" wrote: > p > P.Young@unsw.EDU.AU (Patrick Young) wrote in message news:<55f85d77.0204250058.1e93f258@posting.google.com>...p > > "Sandeep Yelwatkar" <Sandeep_Yelwatkar@bmc.com> wrote in message news:<uce00t53bua477@corp.supernews.com>... > [snip]J > > EVE is more powerful and is my choice (with EDT keypad enabled) simplyI > > because you can define learn sequences and bind them to a key, and itc/ > > supports longer than 255 character records.w > >sM > > You will want to add TPU SET(MOUSE,OFF) to your F$TRNLNM("EVE$INIT") file I > > before you even look at using it, otherwise you will be driven %$^@!# L > > *INSANE* by not being able to cut and paste using the mouse - who's idea > > was this anyway? > >sF > > I never did work out why EDT does not have LEARN where as KED did. > G > Why doesn't EDT's Ctrl/K count as a LEARN equivalent? Press ^K, pressiH > the key you want to define, press the key sequence you want memorized,F > press the period key, then press Enter. Viola! You've defined a key. > 5 > And please pardon my ignorance, what is or was KED?  >  > Disclaimer: JMHO > Alan E. Feldman $ > afeldman atski gfigroup dotski com   It doesn't behave the same.h   try:  4 ^K <somekey> <gold>-<find> database <find> <delchar># <delchar><delchar><delchar> <ENTER>s  E in a file that has the word database on line 10 out of 30 lines using,
 both editors.:  + There are some simialarities but not exact.c --   Regards,  7 Michael Austin            Registered Linux User #261163 7 First DBA Source, Inc.    http://www.firstdbasource.com9 Sr. Consultant 704-947-1089 (Office)e 704-236-4377 (Mobile)o   ------------------------------  % Date: Thu, 25 Apr 2002 17:59:57 +0200i- From: Didier Morandi <Didier.Morandi@Free.fr>y Subject: Re: EDT or EVEu& Message-ID: <3CC827FD.88688DD@Free.fr>  8 because you forgot to terminate the sequence with a dot.   D.   Phillip Helbig wrote:   I > Although an EDT fan, I have not spent much time on ^K.  I have tried ithH > a few times (as following your example), but get "invalid subcommand".   ------------------------------    Date: 25 Apr 2002 11:09:05 -0500+ From: young_r@encompasserve.org (Rob Young)1 Subject: Re: EDT or EVEh3 Message-ID: <EK+rzz3QpM15@eisner.encompasserve.org>m  w In article <01KGZM5PCK9U8Y7GCV@sysdev.deutsche-boerse.com>, Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> writes:iE >> I never did work out why EDT does not have LEARN where as KED did.c >>  H >> Why doesn't EDT's Ctrl/K count as a LEARN equivalent? Press ^K, pressI >> the key you want to define, press the key sequence you want memorized,eG >> press the period key, then press Enter. Viola! You've defined a key.e > J > Although an EDT fan, I have not spent much time on ^K.  I have tried it H > a few times (as following your example), but get "invalid subcommand". > H > A few folks have mentioned here "I use the EDT keypad in EVE and thus J > have both".  This gives you JUST the EDT keypad; there is a lot more to  > EDT.    = 	But whatever is in there is in there.  How do you extend it?n 	s 	Here is my tpu subdirectory:n  < ADD_CHAR.TPU;1      BALANCE.TPU;1       DISPLAY.TPU;1       F EVE_BACKGROUND.TPU;1                    EVE_DIGITAL_REPLACEMENTS.TPU;1< EVE_EDT_CORE.TPU;1  EVE_UTILITY.TPU;1   FIND_FUNCTIONS.TPU;1O INCREMENTAL_FIND.TPU;1                  JUSTIFY.TPU;1       MATCHDELIM.TPU;1   o< MY_DATE_TIME.TPU;1  MY_DATE_TIME_PROC.TPU;1                 O ORIGINAL_EVE_EDT_CORE.TPU;1             PAREN_CHECK.TPU;1   PRINTSCREEN.TPU;1  .O SWITCH_EVE_DCL.TPU;1                    TABS.TPU;1          TEST.TPU;1         iA TEST2.TPU;1         TOGGLE_WIDTHS.TPU;1 TPU_CODING_TEMPLATE.TPU;1n VI_MATCH.TPU;1     	Many of those written by:     ! EVE_DIGITAL_REPLACEMENTS.TPU !D+ !               Written by:     Chris Yoder 3 !                       for:    Harvey Mudd Colleger6 !                               Mathematics Department/ !                       on:     Febuary 8, 1989l  A 	Additionally, in a prior existence I wrote a c module that firednB 	up TPU and it stayed fired up in the background so all your learnE 	sequences that day were always available, very powerful as "editors"aB 	gained just that... all large learn sequences were carried across 	editing sessions:  E         sprintf(cmd_string,"map_save_buffer_in (save_edit_buffer);");b&         do_a_tpu_command(cmd_string);   C         /* Hand off to tpu$control, finally get to edit the file */   L                 return_status = tpu$control(); /* 1 --> suppress messages */     int  do_a_tpu_command(cmd_str)f   char *cmd_str;  	         {          int     stat;u  2         $DESCRIPTOR(tpu$cmd_string_desc, cmd_str);  ;         tpu$cmd_string_desc.dsc$w_length = strlen(cmd_str);w9         stat = tpu$execute_command(&tpu$cmd_string_desc); $         if (!(stat)) LIB$STOP(stat);  	         }r   ---n  @ 	In many ways, EDT is quite uninteresting in comparison as there0 	isn't an underlying language to interface with.   				Robw   ------------------------------  % Date: Thu, 25 Apr 2002 12:16:42 -0400r1 From: Michael Austin <maustin@firstdbasource.com>t Subject: Re: EDT or EVEd2 Message-ID: <3CC82BEA.6FC782D9@firstdbasource.com>   Phillip Helbig wrote:  > F > > I never did work out why EDT does not have LEARN where as KED did. > >SI > > Why doesn't EDT's Ctrl/K count as a LEARN equivalent? Press ^K, press-J > > the key you want to define, press the key sequence you want memorized,H > > press the period key, then press Enter. Viola! You've defined a key. > I > Although an EDT fan, I have not spent much time on ^K.  I have tried it H > a few times (as following your example), but get "invalid subcommand". > G > A few folks have mentioned here "I use the EDT keypad in EVE and thus I > have both".  This gives you JUST the EDT keypad; there is a lot more toy > EDT.  B ... and there is a lot more to EVE. ... another useful feature BOX@ SELECT CUT/PASTE. I can cut/paste columns from the middle of the	 document.5  D result of BOX SELECT then CUT then PASTE from dir/size/owner/out=x.x      145  [TCPIP$AUX,TCPIP$SMT      1  [TCPIP$AUX,TCPIP$SMT     72  [TCPIP$AUX,TCPIP$SMT      1  [TCPIP$AUX,TCPIP$SMT      1  [TCPIP$AUX,TCPIP$SMT      1  [TCPIP$AUX,TCPIP$SMT      1  [TCPIP$AUX,TCPIP$SMT   just as an example.l --   Regards,  7 Michael Austin            Registered Linux User #261163a7 First DBA Source, Inc.    http://www.firstdbasource.comc Sr. Consultant 704-947-1089 (Office)s 704-236-4377 (Mobile)    ------------------------------  % Date: Thu, 25 Apr 2002 17:56:55 +0200a- From: Didier Morandi <Didier.Morandi@Free.fr>s Subject: Re: EDT or EVE (EDT)u' Message-ID: <3CC82747.A2D62308@Free.fr>M  O you miss the point. The gentleman who started the post is asking about an E D I 1 T O R (latin: edit or die), not about a language.t   D   jojimbo wrote: > G > Oh you dinosaurs.  Use EVE for goodness sakes. It's the original openeF > source editor, look in sys$examples:*.tpu.  Customize to your hearts& > content, TPU is a powerful language.   ------------------------------    Date: 25 Apr 2002 09:38:48 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen)f% Subject: Re: EDT or EVE (TECO RULES!)l3 Message-ID: <PmrX9ATy7KGI@eisner.encompasserve.org>a  T In article <3CC7B9B3.D20B930C@127.0.0.1>, Nic Clews <sendspamhere@127.0.0.1> writes: > Larry Kilgallen wrote: >> D >> TECO. > I > Rats. Beat me to it. But I feel a demonstration of "my editor is biggere  > than your editor" is required. >  > $ MOUNT/FORE $1$DKA600:f= > %MOUNT-I-MOUNTED, CDVWIP_DATA mounted on _$1$DKA600: (HOST)  > $ EDIT/TECO $1$DKA600: > *  > ) > Don't try this at home folks, it works.- > < > So, if TPU is so 'powerful' how come it can't edit a disk? >  > I rest my case.YG > WARNING: ABSOLUTELY NO RESPONSIBILITY IS ACCEPTED FOR DESTROYING DISK5+ > STRUCTURES IF YOU USE THE COMMANDS ABOVE.  > -- p* > Regards, Nic Clews CSC Computer Sciences > nclews at csc dot come  B Very impressive, although I imagine CSC customers reading this are. hoping DISK$CDVWIP_DATA is not their data :-).  E Success of this method, however, seems to depend on the exact device.a3 I am having trouble with the EW command on DKA400:.    ------------------------------  % Date: Thu, 25 Apr 2002 09:09:23 +0100e( From: Nic Clews <sendspamhere@127.0.0.1> Subject: Re: EDT or EVE (TECO!)E) Message-ID: <3CC7B9B3.D20B930C@127.0.0.1>d   Larry Kilgallen wrote: >  > TECO.   G Rats. Beat me to it. But I feel a demonstration of "my editor is biggeri than your editor" is required.   $ MOUNT/FORE $1$DKA600:l; %MOUNT-I-MOUNTED, CDVWIP_DATA mounted on _$1$DKA600: (HOST)f $ EDIT/TECO $1$DKA600: *a  ' Don't try this at home folks, it works.^  : So, if TPU is so 'powerful' how come it can't edit a disk?   I rest my case.aE WARNING: ABSOLUTELY NO RESPONSIBILITY IS ACCEPTED FOR DESTROYING DISKe) STRUCTURES IF YOU USE THE COMMANDS ABOVE.- --  ( Regards, Nic Clews CSC Computer Sciences nclews at csc dot coms   ------------------------------    Date: 25 Apr 2002 03:28:32 -0700) From: P.Young@unsw.EDU.AU (Patrick Young)e Subject: Re: EDT or EVE (TECO!)t= Message-ID: <55f85d77.0204250228.7002abab@posting.google.com>w  h Kilgallen@SpamCop.net (Larry Kilgallen) wrote in message news:<538etb7C7mzw@eisner.encompasserve.org>...i > In article <uce00t53bua477@corp.supernews.com>, "Sandeep Yelwatkar" <Sandeep_Yelwatkar@bmc.com> writes:t > > HiJ > >     I am new to the OpenVMS system. Just wondering which is the better; > > editor EVE or EDT. Would like to know your suggesstion.u >  > TECO.t  J YES! - you can get a lot done with K C D EX T S HT and the escape key, um,( oops! where did DECterm map that one :-)   ------------------------------    Date: 25 Apr 2002 08:27:52 -0700# From: jgessling@yahoo.com (jojimbo)e Subject: Re: EDT or EVE (TECO!)e= Message-ID: <dc2d8031.0204250727.10cb06fc@posting.google.com>e  n P.Young@unsw.EDU.AU (Patrick Young) wrote in message news:<55f85d77.0204250228.7002abab@posting.google.com>...j > Kilgallen@SpamCop.net (Larry Kilgallen) wrote in message news:<538etb7C7mzw@eisner.encompasserve.org>...k > > In article <uce00t53bua477@corp.supernews.com>, "Sandeep Yelwatkar" <Sandeep_Yelwatkar@bmc.com> writes:  > > > HiL > > >     I am new to the OpenVMS system. Just wondering which is the better= > > > editor EVE or EDT. Would like to know your suggesstion.- > > 	 > > TECO.0 > L > YES! - you can get a lot done with K C D EX T S HT and the escape key, um,* > oops! where did DECterm map that one :-)  E Oh you dinosaurs.  Use EVE for goodness sakes. It's the original openBD source editor, look in sys$examples:*.tpu.  Customize to your hearts$ content, TPU is a powerful language.   Jimm   ------------------------------  % Date: Thu, 25 Apr 2002 22:21:36 +0800w- From: "Mahadhir Shaarani" <det2@pd.jaring.my> I Subject: How to configure DECvms to connect 3Com OfficeConnect ISDN Modem * Message-ID: <aa93es$elu$1@news5.jaring.my>  I i have problem with this vms, i naver use vms before, now my problem came I becouse of they want to use ISDN connection, i try to find support how tos- configure this vms to connect this ISDN Modeme   ------------------------------  # Date: Thu, 25 Apr 2002 14:43:38 GMT  From: system@SendSpamHere.ORGyM Subject: Re: How to configure DECvms to connect 3Com OfficeConnect ISDN Modemu0 Message-ID: <00A0CFCE.7ECADC87@SendSpamHere.ORG>  Z In article <aa93es$elu$1@news5.jaring.my>, "Mahadhir Shaarani" <det2@pd.jaring.my> writes:J >i have problem with this vms, i naver use vms before, now my problem cameJ >becouse of they want to use ISDN connection, i try to find support how to. >configure this vms to connect this ISDN Modem  F Do you have a static IP address or is the address being assigned DHCP?  H If you have a DHCP address, you may find that the easiest thing to do isG obtain a low-priced router which provides NAT and will handle the DHCP  H lease for the IP.  http://www.tmesis.com/internet/  can provide you with% some pointers to appropriate devices.   H Also, if you know which TCP/IP stack you have on the VMS machine, please7 include that information with any future messages here.0     --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COM3            5J   "And of course, I'm a genius, so people are naturally drawn to my fiery I   intellect.  Their admiration overwhelms their envy!" -- Calvin & HobbesS   ------------------------------  # Date: Thu, 25 Apr 2002 14:56:18 GMT01 From: LESLIE@JRLVAX.HOUSTON.RR.COM (Jerry Leslie)mM Subject: Re: How to configure DECvms to connect 3Com OfficeConnect ISDN Modema: Message-ID: <mOUx8.12771$g72.473217@typhoon.austin.rr.com>  , Mahadhir Shaarani (det2@pd.jaring.my) wrote:K : i have problem with this vms, i naver use vms before, now my problem camecK : becouse of they want to use ISDN connection, i try to find support how to / : configure this vms to connect this ISDN Modem  : M The 3COM OfficeConnect ISDN Modem is really an ISDN router with four 10BaseT  G ethernet ports. It can be configured from any system's web browser. ItsED product specifications are available from (URL folded to two lines):  0   http://www.3com.com/products/en_US/detail.jsp?.   tab=prodspec&sku=3C891A-US&pathtype=purchase  I 3COM should have NEVER called it an ISDN modem, since that's the misnomerMH used for ISDN Terminal Adaptors, which are connected to a host computer  via a serial interface.   K Your VMS system will be connected to one of the ethernet ports. To be able u. to provide further assistance, please provide:     o the VMS version8     o the TCP/IP stack:r      - Digital TCP/IP Services    - Process Software's TCPWaree     - Process Software's Multinet  A   o whether the VMS system will have a static IP address, or use       a dynamic IP address    H --Jerry Leslie   leslie@clio.rice.edu  (my opinions are strictly my own)9   Note: leslie@jrlvax.houston.rr.com is invalid for emailv   ------------------------------  % Date: Thu, 25 Apr 2002 10:59:57 -0400o1 From: Michael Austin <maustin@firstdbasource.com> M Subject: Re: How to configure DECvms to connect 3Com OfficeConnect ISDN ModemT2 Message-ID: <3CC819ED.52DD90D4@firstdbasource.com>   Mahadhir Shaarani wrote: > K > i have problem with this vms, i naver use vms before, now my problem cameZK > becouse of they want to use ISDN connection, i try to find support how to0/ > configure this vms to connect this ISDN Modem   F Use an ISDN Router/Hub like the NetGear RH348.  I would suggest DSL ifE it is available.  ISDN is only 128K up/down max where DSL has varyingpH levels - all above 256K. Residential DSL is generally 1.5M down/256 up. - And you can get SDSL at 756/756 or 1.5/1.5.  l     -- y Regards,  7 Michael Austin            Registered Linux User #261163a7 First DBA Source, Inc.    http://www.firstdbasource.coml Sr. Consultant 704-947-1089 (Office)  704-236-4377 (Mobile)T   ------------------------------  % Date: Thu, 25 Apr 2002 08:04:06 -0700k% From: Dean Woodward <deanw@rdrop.com>vM Subject: Re: How to configure DECvms to connect 3Com OfficeConnect ISDN Modem') Message-ID: <3CC81AE6.EF59D26F@rdrop.com>i   Mahadhir Shaarani wrote: > K > i have problem with this vms, i naver use vms before, now my problem came K > becouse of they want to use ISDN connection, i try to find support how tou/ > configure this vms to connect this ISDN Modem   > OS Version? Hardware Platform? UCX? TCP/IP? MultiNet? TCPWare?   ------------------------------  # Date: Thu, 25 Apr 2002 17:29:45 GMTt1 From: LESLIE@JRLVAX.HOUSTON.RR.COM (Jerry Leslie)lM Subject: Re: How to configure DECvms to connect 3Com OfficeConnect ISDN Modems: Message-ID: <d2Xx8.10147$Y03.608947@typhoon.austin.rr.com>  2 Michael Austin (maustin@firstdbasource.com) wrote: : Mahadhir Shaarani wrote: : > M : > i have problem with this vms, i naver use vms before, now my problem camejM : > becouse of they want to use ISDN connection, i try to find support how to 1 : > configure this vms to connect this ISDN Modemy : 2 : Use an ISDN Router/Hub like the NetGear RH348.    9 That's exactly what the 3Com OfficeConnect ISDN Modem is.-  E : I would suggest DSL if it is available.  ISDN is only 128K up/down  H : max where DSL has varying levels - all above 256K. Residential DSL  isK : generally 1.5M down/256 up. And you can get SDSL at 756/756 or 1.5/1.5.    :   C DSL is limited to locations within 15,000 feet of the telco central2B office. ISDN can be used up to 31,000 feet by use of a device such as an Adtran "Total Reach".t    H --Jerry Leslie   leslie@clio.rice.edu  (my opinions are strictly my own)9   Note: leslie@jrlvax.houston.rr.com is invalid for emaili   ------------------------------  % Date: Thu, 25 Apr 2002 10:39:16 -0700d% From: Dean Woodward <deanw@rdrop.com>aM Subject: Re: How to configure DECvms to connect 3Com OfficeConnect ISDN ModemL) Message-ID: <3CC83F44.4FF69600@rdrop.com>o   Jerry Leslie wrote:s > 1 > 3COM should have NEVER called it an ISDN modem,   G Cisco, the company that should know better, refers to the 67x series as 
 "DSL Modems".   F The issue is that Jane Q. Public knows what a "modem" is.  She doesn'tF know what a "terminal adapter" or "router" is, so they use terminologyE that the user can relate to.  (Not to mention some DSL "modems" don'tsE route, either- nor do cable "modems", or none that I've come across.)    ------------------------------  % Date: Thu, 25 Apr 2002 13:39:04 -0400O1 From: Michael Austin <maustin@firstdbasource.com>sM Subject: Re: How to configure DECvms to connect 3Com OfficeConnect ISDN Modemt2 Message-ID: <3CC83F38.B78EBE15@firstdbasource.com>   Jerry Leslie wrote:h > 4 > Michael Austin (maustin@firstdbasource.com) wrote: > : Mahadhir Shaarani wrote: > : >mO > : > i have problem with this vms, i naver use vms before, now my problem camedO > : > becouse of they want to use ISDN connection, i try to find support how tor3 > : > configure this vms to connect this ISDN Modemo > :o2 > : Use an ISDN Router/Hub like the NetGear RH348. > ; > That's exactly what the 3Com OfficeConnect ISDN Modem is.e > F > : I would suggest DSL if it is available.  ISDN is only 128K up/downJ > : max where DSL has varying levels - all above 256K. Residential DSL  isK > : generally 1.5M down/256 up. And you can get SDSL at 756/756 or 1.5/1.5.a > :i > E > DSL is limited to locations within 15,000 feet of the telco centralaD > office. ISDN can be used up to 31,000 feet by use of a device such > as an Adtran "Total Reach".v > J > --Jerry Leslie   leslie@clio.rice.edu  (my opinions are strictly my own); >   Note: leslie@jrlvax.houston.rr.com is invalid for emaile  > and the telcos can also install remote DSLAM +repeaters in theE individual neighborhoods like mine to get within those ranges.  CablerG will provide the same service - and much faster than ISDN.  It is greatlB to download 5MB+ files in less time than it takes to pour a cup of coffee.o -- t Regards,  7 Michael Austin            Registered Linux User #261163e7 First DBA Source, Inc.    http://www.firstdbasource.comp Sr. Consultant 704-947-1089 (Office)  704-236-4377 (Mobile)w   ------------------------------  % Date: Thu, 25 Apr 2002 13:45:19 +0100nT From: Andrew Harrison SUNUK Consultancy <andrew_nospam.remove_this.harrison@sun.com>6 Subject: Re: I was right!  Alpha will live in Itanium!& Message-ID: <3CC7FA5F.1050208@sun.com>   Terry C. Shannon wrote:H  % > "Andrew Harrison SUNUK Consultancy"T? > <andrew_nospam.remove_this.harrison@sun.com> wrote in messageD" > news:3CC6B616.8060107@sun.com... >  > <snip> > @ >>Incedentally Sun has better throughput and per CPU performanceC >>on the three other large commercial benchmarks you have publishedM? >>for the GS, SAP, Oracle Apps and TPC-H. Perhaps thats why you   >>only want to talk about TPC-C. >> > J > Andrew, when you make statements such as the one above, would you pleaseL > provide links to supporting information. I'm sure there are a lot of us in8 > this newsgroup who would be interested in the numbers. > 	 > Thanks,t >      I did in detail.    & Oracle applications standard benchmark  A Sun F6800 24 x 750 Mhz CPU's ~25% faster than a 32 CPU GS320 withv 731 Mhz CPU's.  B Both Sun and Compaq now have faster CPU's and given the respectiveB performance increases the GS with more CPU's is still not going toB overhaul the F6800. The IBM P680 with 24 CPU's is also faster thanB the GS. So the F6800 has better throughput and considerably better per CPU performance.  
 SAP SD 2 Tiers  + Sun F15000 76 CPU's SAP  vsn 4.6 4100 usersc, Compaq GS320 32 CPU's SAP vsn 4.0 2720 users  . SAP vsn 4.6 uses 1.7 x the resources of 4.0 so0 the F15000 is actually doing more throughput per3 CPU than the GS320 it also has a higher throughput.a   TPC-Ho  % Sun F15000 72 CPU's 18802 QphH (1 TB)u( Compaq GS320 32 CPU's 4951 QphH (300 GB)# Sun F6800 24 CPU's 4735 QphH (1 TB)m  9 Why Compaq havn't done the 1 TB number I don't know since < all the other larger DW contenders concentrate on 1 or 3 TB.  > Now these are the only large commercial public benchmarks that< Compaq appear to have published for the GS320's note that in> all of them the total throughput and the per CPU throughput is lower than Sun's.u   Regardse Andrew Harrisono   ------------------------------    Date: 25 Apr 2002 10:57:02 -0700( From: bob@instantwhip.com (Bob Ceculski)6 Subject: Re: I was right!  Alpha will live in Itanium!< Message-ID: <d7791aa1.0204250957.5f89704@posting.google.com>   Andrew Harrison SUNUK Consultancy <andrew_nospam.remove_this.harrison@sun.com> wrote in message news:<3CC7FA5F.1050208@sun.com>... > Terry C. Shannon wrote:  >  > ; > Why Compaq havn't done the 1 TB number I don't know since-> > all the other larger DW contenders concentrate on 1 or 3 TB. > @ > Now these are the only large commercial public benchmarks that> > Compaq appear to have published for the GS320's note that in@ > all of them the total throughput and the per CPU throughput is > lower than Sun's.n > 	 > Regardsd > Andrew Harrisonk  B oh they have Andrew, and I hear it is going to be M A R V E L ous!   ------------------------------  % Date: Thu, 25 Apr 2002 09:44:58 +0100 % From: Alan Greig <a.greig@virgin.net>.9 Subject: Re: Inquirer even says Alpha in Itanium in 2005!a8 Message-ID: <sqffcuk2jak93ou44vhqakvlk14o9qarm5@4ax.com>  A On 24 Apr 2002 05:37:28 -0700, bob@instantwhip.com (Bob Ceculski)5 wrote:     >PE >if that is true, then the above is true, Intel will offer two 64 bit B >platforms, an x86 junk one and use Alpha EV8 for the high end ...  E The AMD 64 bit chip was designed primarily by ex- Alpha engineers and-F apparently blows away the Intel 64 bit chip in terms of performance. IE have seen speculation that Intel has *some* of the Alpha engineers itaC acquired working on 64 bit AMD compatible extensions to the Pentium-C (Yamhill) instead of improving  the IA64 chip as we are told.. OnlyrC Intel employees could confirm or deny that directly and I doubt anye will.D  B If Microsoft really is leaning on Intel to follow the AMD line andA downplay IA64 then the future of IA64 really is in serious doubt. @ Note the following email (which is apparently genuine) from DaveD Cutler quoted on the Inquirer. I note Gordon Bell is also on the To: list.i   From: David Cutler) Sent: Thursday, February 07, 2002 7:27 PMtC To: Bill Gates; Steve Ballmer; Jim Allchin; Brian Valentine; Vance, C Mike; Brunner, Richard; Weber, Fred; Meretsky, Wayne Cc: AMD 64 bitKE working group; Forrest Foltz; David Cutler; Gordon Bell; Neill Clift;TB Samer Arafeh; Landy Wang; Jeff Havens; Andre Vachon; Bryan Tuttle;C Simon Koeman; Kevin Frei; Bob Davidson; Terry Leeper; BV's Cabinet;e Jim Allchin's StaffC   Subject: A day in history.  F Today at 3pm we received the first AMD64 K8 based hardware system from AMD.  E Today at 5pm we booted and are running the 64-bit versions of Windowso XP on the AMD64 K8 system.  , We are even running 32-bit apps under wow64!  E The hardware system is unbelievably solid for a chip that just poppedo! out of the oven a mere month ago.e  B In addition to running the 64-bit system we can dual boot into theB 32-bit version of XP as well on the very same hardware without any changes!  C Thanks are due to a great many people for their help and support ini making this happen.   + Watch this space for more news coming soon!D   The Inquirer  2002 Breakthrough. Publishing Ltd   All rights reserved.   -- Alan   ------------------------------    Date: 25 Apr 2002 07:11:58 -0700( From: bob@instantwhip.com (Bob Ceculski)9 Subject: Re: Inquirer even says Alpha in Itanium in 2005!s= Message-ID: <d7791aa1.0204250611.4e5a00d1@posting.google.com>g  e Alan Greig <a.greig@virgin.net> wrote in message news:<sqffcuk2jak93ou44vhqakvlk14o9qarm5@4ax.com>...I >eG > The AMD 64 bit chip was designed primarily by ex- Alpha engineers and.H > apparently blows away the Intel 64 bit chip in terms of performance. IG > have seen speculation that Intel has *some* of the Alpha engineers itlE > acquired working on 64 bit AMD compatible extensions to the PentiumdE > (Yamhill) instead of improving  the IA64 chip as we are told.. OnlyPE > Intel employees could confirm or deny that directly and I doubt any- > will.d > D > If Microsoft really is leaning on Intel to follow the AMD line andC > downplay IA64 then the future of IA64 really is in serious doubt.@B > Note the following email (which is apparently genuine) from DaveF > Cutler quoted on the Inquirer. I note Gordon Bell is also on the To: > list.e >  > From: David Cutler+ > Sent: Thursday, February 07, 2002 7:27 PMsE > To: Bill Gates; Steve Ballmer; Jim Allchin; Brian Valentine; Vance,pE > Mike; Brunner, Richard; Weber, Fred; Meretsky, Wayne Cc: AMD 64 bitnG > working group; Forrest Foltz; David Cutler; Gordon Bell; Neill Clift;ED > Samer Arafeh; Landy Wang; Jeff Havens; Andre Vachon; Bryan Tuttle;E > Simon Koeman; Kevin Frei; Bob Davidson; Terry Leeper; BV's Cabinet;  > Jim Allchin's Staff  >  > Subject: A day in history- > H > Today at 3pm we received the first AMD64 K8 based hardware system from > AMD. > G > Today at 5pm we booted and are running the 64-bit versions of Windows0 > XP on the AMD64 K8 system. > . > We are even running 32-bit apps under wow64! > G > The hardware system is unbelievably solid for a chip that just popped.# > out of the oven a mere month ago.e > D > In addition to running the 64-bit system we can dual boot into theD > 32-bit version of XP as well on the very same hardware without any
 > changes! > E > Thanks are due to a great many people for their help and support inh > making this happen.d >   G which is all the more reason why Intel has to dump the Epic boat anchor E and roll out EV8 to keep the market lead ... hammer will only stay inuF the same market pentium covers now ... windows will never run the highC end ... ever!  It can't unless they completely rewrite it again and C make it completely like vms and throw out the x86 garbage, but theyoE can't!  Intels only chance is to offer cheap pentiums, maybe a hammer,G knock off for the low end, and EV8 for the high end ... otherwise, theyy> will go the way of DEC and Q ... I hope they aren't as stupid!   ------------------------------  % Date: Thu, 25 Apr 2002 15:42:08 +0100m% From: Alan Greig <a.greig@virgin.net> 9 Subject: Re: Inquirer even says Alpha in Itanium in 2005!e8 Message-ID: <575gcusdknl1o8fdpfibpv7je2f4ajvqjv@4ax.com>  A On 25 Apr 2002 07:11:58 -0700, bob@instantwhip.com (Bob Ceculski)  wrote:   >0H >which is all the more reason why Intel has to dump the Epic boat anchorF >and roll out EV8 to keep the market lead ... hammer will only stay inG >the same market pentium covers now ... windows will never run the high.  E But what if Solaris is ported to Hammer (allegedly already underway),hE Linux is ported to Hammer (already done), HP-UX is ported (who knows,yD if so it hasn't leaked) and VMS is ported to Hammer (not planned but' likely no harder than the Itanium port)o  D >end ... ever!  It can't unless they completely rewrite it again andD >make it completely like vms and throw out the x86 garbage, but theyF >can't!  Intels only chance is to offer cheap pentiums, maybe a hammerH >knock off for the low end, and EV8 for the high end ... otherwise, they? >will go the way of DEC and Q ... I hope they aren't as stupid!   C Intel have never shown themselves to be stupid. My guess is that iffF they dump EPIC it will be in favour of IA64 not Alpha. I would like to	 be wrong.  -- Alan   ------------------------------  # Date: Thu, 25 Apr 2002 16:37:33 GMT * From: "Bill Todd" <billtodd@metrocast.net>9 Subject: Re: Inquirer even says Alpha in Itanium in 2005!n@ Message-ID: <hhWx8.10499$fg.1185525@bin2.nnrp.aus1.giganews.com>  2 "Alan Greig" <a.greig@virgin.net> wrote in message2 news:575gcusdknl1o8fdpfibpv7je2f4ajvqjv@4ax.com...   ...r   > My guess is that ifO- > they dump EPIC it will be in favour of IA647   I suspect you mean Yamhill..    not Alpha. I would like to@ > be wrong.   L So would I.  But it's unlikely:  it would require that  a) Yamhill be unableL to take up the slack (unlikely unless Yamhill isn't a credible competitor toF Hammer, in which case Intel will be in *real* trouble) and  b) that anB on-chip-glue-enhanced Itanic (likely around 2005) still isn't evenF marginally competitive with its high-end competition (being marginallyK competitive in performance is all Intel needs to grab a decent slice of thetL market).  By contrast, converting to Alpha would require either  1) throwingG away all compatibility with the already-completed hardware and softwarehJ surrounding Itanic (not only a major embarrassment for Intel, but a *real*L stick in the eye to its partners who have committed to it) or  2) convertingK the Alpha core to the Itanic instruction set and physical interfaces (whichpE would push it out to 2006 or so - quite possibly too late to save the-
 platform).  J I suppose it's conceivable that Intel is sufficiently wealthy to have keptH EV8 development going full blast (but in secret) since last June so thatL option 1 above would exist in 2004 just in case Itanic sank spectacularly inL the near term rather than just continued taking on water at its current rateI (which might keep it afloat until the 2005 improvements).  But I wouldn't  advise betting on it.u   - bill   ------------------------------  % Date: Thu, 25 Apr 2002 12:17:54 -0400C( From: David Froble <davef@tsoft-inc.com>9 Subject: Re: Inquirer even says Alpha in Itanium in 2005! , Message-ID: <3CC82C32.3030606@tsoft-inc.com>   Alan Greig wrote:     C > The AMD 64 bit chip was designed primarily by ex- Alpha engineersy     Alan,p  O I've seen this claim before.  Is there any more detailed information available r on the subject?G  N To think that at one time Digital had it all, and then due to not letting the C people do good things, they went elsewhere to do those good things.   P Instead of having kids and raising them, it's more like the parents die gicving 4 birth to the kids.  At least it sure killed Digital.   ------------------------------    Date: 25 Apr 2002 10:28:21 -0700( From: bob@instantwhip.com (Bob Ceculski)4 Subject: Itanium 2 hits ... Itanium 3 will be Alpha!= Message-ID: <d7791aa1.0204250928.149c3eee@posting.google.com>W  B I told you J.F. and Bill "the genius" Todd that Alpha will show upF not to far down the road ... no doubt Itanium 3 will be the beginnings
 of Alpha EV8!a    ! Intel to call McKinley Itanic 2? l   Is it hitting the icebergs? & By Mike Magee, 25/04/2002 15:52:17 BST    F A NUMBER OF ASIAN SITES are claiming that Intel will call the McKinleyD processor the Itanium 2, with a few sites having pictures of the new logo.UC According to the reports, the logo was shown at the Intel Developera Forum in Asia.  = Intel in Germany sent out a release today confirming that thel' processor will be called the Itanium 2.s  F We first saw these reports on VR-Zone, but a number of other sites areC also carrying pictures such as: Prom Alphaloop, Gzeasy, and PC Pop.e  C There's not much more to say - it's what Intel has gone and done to @ the name of the chip that Nico Ernst from tec channel dubbed the
 McMonster.  A The only thought that occurs to us is that as the McKinley is therE Itanic 2, doesn't that imply that the first Itanic hit the iceberg...kB and sank? Intel is now having two goes with this processor family. Will it be third time lucky?   ------------------------------    Date: 25 Apr 2002 10:16:26 +02006 From: Jan Ingvoldstad <jani+news-comp@nntp.ifi.uio.no>H Subject: Re: markup vs wysiwyg (was: Re: learning how to use a computer)0 Message-ID: <oaselh4b3c5.fsf@tyrfing.ifi.uio.no>  J On Thu, 25 Apr 2002 03:00:27 +0200, Michael Joosten <joost@c-lab.de> said:   > JF Mezei wrote:c >> h >> Jan Ingvoldstad wrote:"H >> > This is not a problem in MS Word, either.  MS Word has styles, very8 >> > similar in function to the styles of e.g. PageMaker >> i  G > But you are not required to use them, and... how long did it take for05 > you to actually find out how to use them correctly?:  B For me?  Absolutely no time at all, but then again, that's not the issue.    N >> markup language forces you to use styles, so the mistake of  building large+ >> documents without styles doesn't happen.m  
 > Exactly.  F This depends on the users Getting the Point.  Arguably, it's easier toB make people understand that a section begins with \section{SectionB Title}, etc., but it doesn't mean they won't be making many of theF same mistakes as they would be if they used MS Word.  It depends a lotF more on what they're used to, combined with what you try to teach themF and their general attitude.  But yes, of course, LaTeX does give usersD more of a push in the general direction of thinking about structure.  F I don't have any formal background in this particular material (though@ some minor background with the theoretical parts as a part of myC education), but I've worked supporting students of computer sciencerE for a few years, both with studies and technical issues.  Contrary tosF popular belief, there are quite a lot of computer science students who> have barely touched a computer before, so a lot of interesting@ reactions could be had.  (Some times, I was more frustrated than actually interested. :).  C I'd say that most people will be able to make passable documents in D either MS Word, FrameMaker or LaTeX, depending on how well they wereC taught (or how well they learned) to use their tools.  It's amazingDC how similar \large can be to a small section title to the untrained5F person, and how much easier it is for a lot of people to use \textit{}, or \textbf{} rather than logical structures.  D Really, this doesn't strike me as any more different than most otherD areas of expertise.  People who are interested (either as a hobby, aD tool, or a line of work) in doing things properly, for the necessaryE values of "things", will be better at it than those who don't.  Thosen+ who don't, will mostly be pushing out crap.e   -- nA In the beginning was the Bit, and the Bit was Zero.  Then SomeonesE said, Let there be One, and there was One.  And Someone blessed them,iD and Someone said unto them, Be fruitful, and multiply, and replenishC the Word and subdue it: and have dominion over every thing that is.t   ------------------------------  % Date: Thu, 25 Apr 2002 12:20:44 +02003( From: Bernd Paysan <bernd.paysan@gmx.de>H Subject: Re: markup vs wysiwyg (was: Re: learning how to use a computer)" Message-ID: <s9l8aa.l1v.ln@miriam>   Michael Joosten wrote:F > Exactly. At some point I really would like to force everybody to use? > LaTeX 2e in the next European project. If it were not for the E > problem of adding pictures/diagrams in (La)TeX, perhaps I might get. > a majority sometime.  B Is there a problem? Even programs that are not able to export eps ? (which are a minority), can print to postscript, and there's a oE ps2epsi script in the Ghostscript package that generates the outline , (with the bbox driver).e  D Not being able to properly include EPS (at least WYSIWYG) is one of  my arguments against Word.   -- t Bernd Paysan7 "If you want it done right, you have to do it yourself"o http://www.jwdt.com/~paysan/   ------------------------------  # Date: Thu, 25 Apr 2002 14:24:09 GMTh+ From: Anne & Lynn Wheeler <lynn@garlic.com>aH Subject: Re: markup vs wysiwyg (was: Re: learning how to use a computer)( Message-ID: <elh3n9fk.fsf@earthlink.net>  8 Jan Ingvoldstad <jani+news-comp@nntp.ifi.uio.no> writes:H > I don't have any formal background in this particular material (thoughB > some minor background with the theoretical parts as a part of myE > education), but I've worked supporting students of computer sciencerG > for a few years, both with studies and technical issues.  Contrary to4H > popular belief, there are quite a lot of computer science students who@ > have barely touched a computer before, so a lot of interestingB > reactions could be had.  (Some times, I was more frustrated than > actually interested. :)0  E markup derivative happens to be that one of the guys that added it to ? script in the very early '70s had a last name that started witheB M. script had been developed in the mid-60s and was something of aC "dot" command derivative (like runoff). G, M, & L then came up with D this other structure and added it to script as an alternative to theF dot command stuff. Generalized Markup Language ...  came about becauseB of their initials G, M, & L. GML evolved into ISO standard as SGMLB (standard) in the late '70s (and then later things like HTML, XML,E etc) . Script, GML, SGML at the Cambrigdge Science Center, 4th floor,  545 tech sq.   random refs:` http://www.garlic.com/~lynn/94.html#43 Bloat, elegance, simplicity and other irrelevant concepts[ http://www.garlic.com/~lynn/94.html#55 How Do the Old Mainframes Compare to Today's Micros? 2 http://www.garlic.com/~lynn/96.html#24 old manualsI http://www.garlic.com/~lynn/97.html#9 HELP! Chronology of word-processingc? http://www.garlic.com/~lynn/97.html#26 IA64 Self Virtualizable?eI http://www.garlic.com/~lynn/98.html#16 S/360 operating systems geneaology2Z http://www.garlic.com/~lynn/98.html#21 Reviving the OS/360 thread (Questions about OS/360)h http://www.garlic.com/~lynn/99.html#42 Enter fonts (was Re: Unix case-sensitivity: how did it originate?h http://www.garlic.com/~lynn/99.html#43 Enter fonts (was Re: Unix case-sensitivity: how did it originate?: http://www.garlic.com/~lynn/99.html#91 Documentation queryw http://www.garlic.com/~lynn/99.html#197 Computing As She Really Is. Was: Re: Life-Advancing Work of Timothy Berners-Lees? http://www.garlic.com/~lynn/2000.html#8 Computer of the centuryhB http://www.garlic.com/~lynn/2000.html#34 IBM 360 Manuals on line ?K http://www.garlic.com/~lynn/2000c.html#30 internal corporate network, misc. B http://www.garlic.com/~lynn/2000d.html#30 Secure Operating SystemsU http://www.garlic.com/~lynn/2000e.html#0 What good and old text formatter are there ?sU http://www.garlic.com/~lynn/2000e.html#1 What good and old text formatter are there ?sU http://www.garlic.com/~lynn/2000e.html#23 Is Tim Berners-Lee the inventor of the web? A http://www.garlic.com/~lynn/2001b.html#50 IBM 705 computer manuale9 http://www.garlic.com/~lynn/2001c.html#88 Unix hard linksoB http://www.garlic.com/~lynn/2001d.html#42 IBM was/is: Imitation...L http://www.garlic.com/~lynn/2001e.html#73 CS instruction, when introducted ?I http://www.garlic.com/~lynn/2001g.html#54 DSRunoff; was Re: TECO Critique:A http://www.garlic.com/~lynn/2001h.html#9 VM: checking some myths.:+ http://www.garlic.com/~lynn/2001h.html#34 DlK http://www.garlic.com/~lynn/2001h.html#46 Whom Do Programmers Admire Now???,b http://www.garlic.com/~lynn/2001i.html#1 History of Microsoft Word (and wordprocessing in general): http://www.garlic.com/~lynn/2001i.html#39 IBM OS Timeline?6 http://www.garlic.com/~lynn/2001l.html#16 Disappointed< http://www.garlic.com/~lynn/2001l.html#20 mainframe question< http://www.garlic.com/~lynn/2001l.html#24 mainframe questionV http://www.garlic.com/~lynn/2001m.html#43 FA: Early IBM Software and Reference Manualsa http://www.garlic.com/~lynn/2001n.html#36 Movies with source code (was Re: Movies with DEC minis)ac http://www.garlic.com/~lynn/2001n.html#37 Hercules etc. IBM not just missing a great opportunity...n4 http://www.garlic.com/~lynn/2002.html#53 School Help_ http://www.garlic.com/~lynn/2002b.html#35 bzip2 vs gzip (was Re: PDP-10 Archive migration plan)sX http://www.garlic.com/~lynn/2002b.html#46 ... the need for a Museum of Computer Software| http://www.garlic.com/~lynn/2002c.html#39 VAX, M68K complex instructions (was Re: Did Intel Bite Off More Than It Can Chew?)F http://www.garlic.com/~lynn/2002f.html#29 Computers in Science Fiction     -- mH Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/    ------------------------------  % Date: Thu, 25 Apr 2002 09:36:30 -0400o% From: Jim Agnew <agnew@mail2.vcu.edu>n Subject: Re: Memory Corruption- Message-ID: <3CC8065E.AFAEB4C3@mail2.vcu.edu>e  G no, but i can tell you if your loops are hard-coded so they will not go H out of array bounds, taking /check=bounds off will speed up your programC by about a factor of 3.. (pulling number out of air after a test 15i
 years ago...)w   Jimo   Jason O'Donnell wrote: > ] > Arne Vajhj <arne.vajhoej@gtech.com> wrote in message news:<3CC5564E.511F7CCD@gtech.com>...  > > jlsue wrote:I > > > Here's another thing to try.  When compiling optimized, turn on alloI > > > checks.  For Fortran, this was /check=all iirc.  At the very least,  > > > put in /checks=bounds. > > >sJ > > > What you will then get is an accvio if/when your code tries to stuff > > > too much into a variable.q > >.D > > In very rare cases /CHECK=BOUND is not possible, because it will. > > effect expected behaviour of the program ! > >  > > Arne > E > Can you post an example of good coding practices where /check=bound, > would not be wanted? >  > JMOD   ------------------------------    Date: 25 Apr 2002 09:52:49 -0500 From: briggs@encompasserve.org Subject: Re: Memory Corruption3 Message-ID: <UdnbwJdq8TG$@eisner.encompasserve.org>r  m In article <9059bf6b.0204231009.173f0c6a@posting.google.com>, jodonnell@hrblock.com (Jason O'Donnell) writes:oE > Can you post an example of good coding practices where /check=bound  > would not be wanted?  = Zero length substrings can be a problem with bounds checking.o          	implicit        none 	options		/extend_source         character *80   word         integer         ll         integer         it  +         read ( 5, '(q,a<l>)' ) l, word(1:l)            i = index ( word, ',' )- 	if ( i .ge. 1 ) i = i - 1         type *, word(1:i)o           end    ------------------------------  # Date: Thu, 25 Apr 2002 17:32:42 GMT * From: John Reagan <john.reagan@compaq.com>! Subject: Re: Netscape and Mozilla ) Message-ID: <3CC83C90.2030207@compaq.com>t   system@SendSpamHere.ORG wrote: > E > I'm assuming that you're running Mozilla 0.9.9 on the latest MarveleD > hardware when you make that claim.  Sorry, but Mozilla is a dog on > the Alpha. >  > --Q > VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMh  B No Marvel here, but as others have mentioned, I have a reasonable C machine, with reasonable memory, with additional tuning for "large e memory footprint applications"  F My OpenVMS machine is an XP 1000, 667MHZ, 2GB RAM, along with lots of   increased SYSGEN and UAF quotas.   (hiyall)$ show sys/multiK OpenVMS V7.2-1  on node HIYALL  25-APR-2002 13:24:56.46  Uptime 21 14:46:36lH    Pid    Process Name    State  Pri      I/O       CPU       Page flts    Pages G 000000AE VUE$REAGAN_11   HIB      4  3259989   0 00:39:37.14    161600 t 29813 MSD Here's Mozilla running on my system. Note that is using 29813 pages 9 (those are 8Kbyte pages) which is about 232 MB of memory.l   -- s John Reagans' Compaq Pascal/{A|I}MACRO Project Leader    ------------------------------  # Date: Thu, 25 Apr 2002 11:28:05 GMTg* From: no.email.address.entered@none444.yet> Subject: Never lose the data in your Palm Desktop again. 38733' Message-ID: <25040206.2807@none444.yet>   p Palm Desktop Utilities version 1.2  is a Windows application designed to work with the Palm Desktop application.   You can use it to quickly and easily backup and restore the data in your Palm Desktop.  Never lose your data again because of a Hotsync error or data corruption.o   You can use it to copy data directly from one Palm Desktop user to another. Making it very easy to have the same data on multiple handhelds. n  It also provides basic information about your handheld such as the secret hotsync ID number that most users never see. (Which is sometimes the reason why your Palm won't hotsync on two different computers to the same user name. The Hotsync ID on each computer is different.) R  X Give it a try. Download the free 14 day trial version from one of these great web sites.   PalmOSWorld.como) http://www.palmosworld.com/downloads/pdu t   VersionTracker.com5 http://www.versiontracker.com/morei...id=5068&db=palm    PalmGear.comW http://www.palmgear.com/software/showsoftware.cfm?sid=72455920020412121409&prodID=41105n  B To register your copy of Palm Desktop Utilities version 1.2 visit:1 http://www.palmosworld.com/downloads/pdu/register    ------------------------------    Date: 25 Apr 2002 07:08:38 -0700+ From: ghairst@mcvh-vcu.edu (G. W. Hairston)n$ Subject: Re: Openvm 7.3 and DECevent= Message-ID: <1ef525c1.0204250608.57f8421a@posting.google.com>   p ghairst@mcvh-vcu.edu (G. W. Hairston) wrote in message news:<1ef525c1.0204241101.50cea19c@posting.google.com>...E > Please give me insight on installing Decevent under openvms 7.3. WeA' > are new at using openvms.  Thank you.w     4/25/02iO Is it necessary to increase the JTquota and PGLLQUO if it was currently running 
 under v7.2-1?b    < Has any ever installed Compaq Availibity Manager on desktop.   Thank you for your cooperation.    ------------------------------  # Date: Thu, 25 Apr 2002 07:55:27 GMTe From: as@if.com (MG), Subject: Re: Pathworks 6.1 exhausting KNBCBs1 Message-ID: <3cc7b4fe.23108106@news.videotron.ca>n   Rob,C you can quickly check for stale sessions by observing the idle timenE against "$admin show sessions" for each session.  Kill the duplicatesu# that show the highest idle times.  s  E You might want to check with your CSC as that I believe a patch mightwC exist for sessions being exhausted due to session leaks to the PDC.iE This mostly affects Member servers, but a patch does exist to correct- this.  -  F Still do try to raise you session limit though since it will give your sesions more padding.    HTH, Mike G   ------------------------------  + Date: Thu, 25 Apr 2002 17:09:19 +0100 (MET)f9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> Y Subject: put up or shut up (was: RE: AMD To Announce Microsoft Support For New Chip Famil ; Message-ID: <01KGZKRZZ3DI8Y7GCV@sysdev.deutsche-boerse.com>b  K > > >you just gave the evidence ... if EV8 can compete or beat hammer, thentL > > >Intel will have to use EV8 or go out of business ... if Itanic tanks as< > > >you are suggesting above, then there is no alternative! > > I > > You miss the obvious. Intel could use EV8 technology and engineers to-G > > make their own "Hammer" compatible chip and not use them to improveo
 > > the IA64.o > I > why do that when they already have the Alpha engineers in tact, the EV8-I > blueprints in hand ... that is a waste of more time and money!  And didnF > anyone ever wonder why there is a beta 3 release of windows 2000 for	 > Alpha? o  I Even though it has been stated many times by many people who should know lF that ALPHA will not live on in Itanium in any meaningful sense of the I word, neither according to current plans nor to any yet-to-be-made plans lD (i.e. such plans will not be made), you still continue to post such 8 fantasies.  why not put your money where your mouth is?   B For example, why not publicly state that if VMS systems running onF Itanium do not have "ALPHA inside" in any meaningful sense of the termG by, say, 2010, you will buy me such a system of my choice at that time.fF And if one of my current ALPHA images runs on a VMS system running on F Itanium without binary translation, then I will buy you the system of  your choice?  G For anyone from outside the group following this thread, your comments eC make it seem like this group is totally clueless.  I now know what hF Robert Plant feels like when some dazed and confused kid says for the H upteen-thousandth time "if you play `Stairway to Heaven' backwards, one D can hear `my sweet Satan, you're the only one for me'---really, Mr. I Plant, it does---maybe you didn't notice it at the time because you were u guided by a higher force".  :-|s   ------------------------------  # Date: Thu, 25 Apr 2002 11:28:06 GMTf* From: no.email.address.entered@none444.yet6 Subject: Read this if you own a PalmOS handheld. 68655' Message-ID: <25040206.2808@none444.yet>o  p Palm Desktop Utilities version 1.2  is a Windows application designed to work with the Palm Desktop application.   You can use it to quickly and easily backup and restore the data in your Palm Desktop.  Never lose your data again because of a Hotsync error or data corruption.y   You can use it to copy data directly from one Palm Desktop user to another. Making it very easy to have the same data on multiple handhelds.    It also provides basic information about your handheld such as the secret hotsync ID number that most users never see. (Which is sometimes the reason why your Palm won't hotsync on two different computers to the same user name. The Hotsync ID on each computer is different.) n  X Give it a try. Download the free 14 day trial version from one of these great web sites.   PalmOSWorld.com ) http://www.palmosworld.com/downloads/pdu n   VersionTracker.com5 http://www.versiontracker.com/morei...id=5068&db=palm    PalmGear.comW http://www.palmgear.com/software/showsoftware.cfm?sid=72455920020412121409&prodID=41105s  B To register your copy of Palm Desktop Utilities version 1.2 visit:1 http://www.palmosworld.com/downloads/pdu/registere   ------------------------------  + Date: Thu, 25 Apr 2002 15:37:19 +0100 (MET)d9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com>  Subject: SMTP sanity checksc; Message-ID: <01KGZHFG96TQ8Y7GCV@sysdev.deutsche-boerse.com>o  E With TCPIP from Compaq, is it documented somewhere exactly what SMTP aI does when it receives an email in order to decide (assuming /NORELAY for   now) to accept it or reject it?i  E Other applications (TELNET, FTP etc) will work when SMTP won't; SMTP tE apparently does some sort of sanity check.  On my hobbyist system, I iG always managed to get it working, but would like to understand exactly  : which of the various measures (such as "extra" entries in I TCPIP$HOSTS.DAT) are necessary and, since the behaviour might be changed  1 in the future, which are the preferred way to go.e  H As an example, consider machine.foo.subdomain.domain which has an alias 0 DNS entry (in the DNS server for the domain) of 9 alias.bar.subdomain.domain and consider sending email to  G user@machine.foo.subdomain.domain, user@alias.bar.subdomain.domain and tF user@mail.bar.subdomain.domain, where mail.bar.subdomain.domain is an 7 alias DNS entry (in the DNS server for the domain) for hH cluster.foo.subdomain.domain, which in turn points to the IP address of I the cluster alias (the other names pointing to the individual IP address - of the machine).  B On a related note, when is it NECESSARY to have multiple entries, E differing only in case, in TCPIP$HOSTS.DAT?  When is it necessary to nH have, for local machines, both unqualified and fully qualified domains,  in TCPIP$HOSTS.DAT?c  I In TCPIP$CONFIG.COM, is my recollection correct that in the past one was iD prompted for the unqualified host name (and is now prompted for the  fully qualified domain)?   ------------------------------  % Date: Thu, 25 Apr 2002 06:55:09 -0700 # From: "Tom Linden" <tom@kednos.com>p Subject: RE: SMTP sanity checks 9 Message-ID: <CIEJLCMNHNNDLLOOGNJIGEFEENAA.tom@kednos.com>    Did you look at smtp.config?   >   -----Original Message-----D >   From: Phillip Helbig [mailto:HELBPHI@sysdev.deutsche-boerse.com]* >   Sent: Thursday, April 25, 2002 7:37 AM >   To: Info-VAX@Mvb.Saic.Coma >   Subject: SMTP sanity checksi >o >eH >   With TCPIP from Compaq, is it documented somewhere exactly what SMTP? >   does when it receives an email in order to decide (assumingn >   /NORELAY for# >   now) to accept it or reject it?: >aH >   Other applications (TELNET, FTP etc) will work when SMTP won't; SMTPH >   apparently does some sort of sanity check.  On my hobbyist system, IJ >   always managed to get it working, but would like to understand exactly= >   which of the various measures (such as "extra" entries in A >   TCPIP$HOSTS.DAT) are necessary and, since the behaviour mighte >   be changed5 >   in the future, which are the preferred way to go.a >eK >   As an example, consider machine.foo.subdomain.domain which has an aliaso3 >   DNS entry (in the DNS server for the domain) ofe< >   alias.bar.subdomain.domain and consider sending email toJ >   user@machine.foo.subdomain.domain, user@alias.bar.subdomain.domain andI >   user@mail.bar.subdomain.domain, where mail.bar.subdomain.domain is anh: >   alias DNS entry (in the DNS server for the domain) forK >   cluster.foo.subdomain.domain, which in turn points to the IP address ofhA >   the cluster alias (the other names pointing to the individual  >   IP address >   of the machine). >nE >   On a related note, when is it NECESSARY to have multiple entries,MH >   differing only in case, in TCPIP$HOSTS.DAT?  When is it necessary toK >   have, for local machines, both unqualified and fully qualified domains,i >   in TCPIP$HOSTS.DAT?f >r? >   In TCPIP$CONFIG.COM, is my recollection correct that in theh >   past one wasG >   prompted for the unqualified host name (and is now prompted for then >   fully qualified domain)? >r   ------------------------------  % Date: Thu, 25 Apr 2002 10:58:51 -0400d1 From: Michael Austin <maustin@firstdbasource.com>P Subject: Re: SMTP sanity checks 2 Message-ID: <3CC819AB.C44B20BA@firstdbasource.com>   Phillip Helbig wrote:g > F > With TCPIP from Compaq, is it documented somewhere exactly what SMTPJ > does when it receives an email in order to decide (assuming /NORELAY for! > now) to accept it or reject it?h > F > Other applications (TELNET, FTP etc) will work when SMTP won't; SMTPF > apparently does some sort of sanity check.  On my hobbyist system, IH > always managed to get it working, but would like to understand exactly; > which of the various measures (such as "extra" entries inpJ > TCPIP$HOSTS.DAT) are necessary and, since the behaviour might be changed3 > in the future, which are the preferred way to go.  > I > As an example, consider machine.foo.subdomain.domain which has an alias?1 > DNS entry (in the DNS server for the domain) ofl: > alias.bar.subdomain.domain and consider sending email toH > user@machine.foo.subdomain.domain, user@alias.bar.subdomain.domain andG > user@mail.bar.subdomain.domain, where mail.bar.subdomain.domain is anl8 > alias DNS entry (in the DNS server for the domain) forI > cluster.foo.subdomain.domain, which in turn points to the IP address ofnJ > the cluster alias (the other names pointing to the individual IP address > of the machine). > C > On a related note, when is it NECESSARY to have multiple entries,lF > differing only in case, in TCPIP$HOSTS.DAT?  When is it necessary toI > have, for local machines, both unqualified and fully qualified domains,n > in TCPIP$HOSTS.DAT?f > J > In TCPIP$CONFIG.COM, is my recollection correct that in the past one wasE > prompted for the unqualified host name (and is now prompted for theg > fully qualified domain)?  1 I think what you are looking for can be found at:iR >http://www.openvms.compaq.com:8000/73FINAL/6526/6526pro_031.html#smtp_config_file  ' What problem are you trying to resolve?B   -- m Regards,  7 Michael Austin            Registered Linux User #261163/7 First DBA Source, Inc.    http://www.firstdbasource.comt Sr. Consultant 704-947-1089 (Office)  704-236-4377 (Mobile)p   ------------------------------  + Date: Thu, 25 Apr 2002 17:27:13 +0100 (MET)s9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com>r Subject: Re: SMTP sanity checksm; Message-ID: <01KGZLETHXOC8Y7GCV@sysdev.deutsche-boerse.com>c  1 > http://www.openvms.compaq.com:8000/73FINAL/6526/$ > /6526pro_031.html#smtp_config_file  E Thanks for the pointer.  This is mostly concerned with anti-spamming :4 features, which I'm not interested in at the moment.  ) > What problem are you trying to resolve?u  H No particular problem.  I just want to understand things.  It works OK, 8 it's just that I think some of the stuff is unnecessary.  I For example, suppose email is sent to an address which is not the "real" eF name of the machine.  One can make the machine accept it by adding an E entry for the address in the To: header to TCPIP$HOSTS.DAT, by using  C /SUBSTITUTE_DOMAIN, by entries in a file in the login directory of jG TCPIP$FTP and perhaps by MX records in the local database.  What is theeG preferred method, and why?  What exactly causes the mail to be rejectedgH if the necessary information is not present?  Is only the local database3 checked, or are remote DNS lookups done?  Etc etc. i  G Even with a simple system, the number of possibilities is quite large,  I so just testing out all combinations to see what works would take a long C! time---and still not tell me why.    ------------------------------  % Date: Thu, 25 Apr 2002 11:21:49 -0400 1 From: Michael Austin <maustin@firstdbasource.com>cG Subject: Re: somewhat off-topic: is there a DSL router with a console?!y2 Message-ID: <3CC81F0D.5DB26DA4@firstdbasource.com>   Jerry Leslie wrote:u > 4 > Michael Austin (maustin@firstdbasource.com) wrote: > : Phillip Helbig wrote:u > : >iN > : > For my hobbyist setup, I have a LAN behind an ISDN router.  I have fixedG > : > IP addresses.  I am thinking of adding DSL because flat rates areoK > : > available.  I don't know of anyone who offers fixed IP addresses with0I > : > DSL.  DHCP is not an option because, again, making the names of thepK > : > machines known to the world is an extra service which is probably not94 > : > available with a standard cheap flat-rate DSL. > : >cF > : > The ISDN connection times out after 2 minutes.  If I initiate anM > : > outgoing connection, I pay for it as part of the normal phone bill.  IfnM > : > the ISP creates a connection to me, he charges me at the standard phonetM > : > rate.  (My telephone bill is such that I don't have to pay for outgoing M > : > calls on Sundays and holidays, so I initiate an outgoing connection and / > : > keep it alive from midnight to midnight.)i > : >1N > : > It seems that, at least for now, I will have to keep the ISDN router forI > : > incoming stuff.  For outgoing stuff, however, it might be worth theaN > : > trouble to go through a DSL router, since I wouldn't have to pay for it.I > : > (I think that this will save a lot of costs, but there WILL be somenN > : > additional cost since under the new scheme all incoming connections willJ > : > be initiated by the ISP---in the past, if an outgoing connection was, > : > already there, no new one was needed). > : >(N > : > I suppose it is no problem to have two routers in the same LAN---one theM > : > present ISDN router and one a new DSL router.  I assume that by settingoG > : > the DSL router as the default gateway (and, perhaps, changing the,J > : > ALTERNATE GATEWAY in SMTP as well as the bind resolver---so that allM > : > external machines I need to access are available via DSL (at the moment6N > : > they are at the ISP and I'm not sure if he will allow just anyone to useA > : > them)), outgoing stuff will not use the ISDN router at all.. > : >6L > : > So far, so good, unless I've missed something, which hopefully someone > : > here could point out.a > : >iK > : > It would be nice to be able to connect to my LAN from the outside viaOM > : > DSL as well---without a fixed IP address.  I could initiate connections:M > : > periodically automatically to where I want to connect from, see what IP M > : > address (that of the DSL router) they come from, and then TELNET to the M > : > router.  However, I want to get from the router to machines on the LAN.r > : >tF > : > I've heard of routers which have a "TELNET console": once one isE > : > connected to it (via telnet, dial-up, whatever), one can TELNETaI > : > elsewhere from it (sort of like the console on a terminal server oryF > : > whatever).  Are there DSL routers which have this functionality? > :h > :hI > : If you want to do it right, get an external ethernet modem and a SOHOaJ > : router (I use the Linksys BEFSR41.  I have SMTP server and News ServerH > : (downloads from my ISP) that works well.. .they don't "allow" ftp orI > : http servers so I have those ports blocked at the router and open thes > : ones I need. > :iF > : I also have a PERL script called ddclient (from ZONEEDIT.COM) thatH > : updates my dynamic IP address (every 5 minutes) -- but it runs on myK > : Linux box.  I have not ported it to run as a detached process on my VMS I > : box yet. I have extracted the usable parts and get "@getip.pl" to get5K > : the current. My VMS box has a static non-routable 192.168.. IP address.t5 > : DNS points to local domain and uses DNS from ISP.l > :pI > : 2 VMS boxes (Alpha 2100's - one is currently powered off due to heat)8* > : 2 Linux PC's - (133 and 166Mhz Pent 1) > : 2 Windoze (98 and ME)v > : 1 Laptop (WNT4.0)s > : an additional 8-port hub > :1H > That's similar to my cable modem router setup. The router is a Linksys= > BEFSR41 router/4-port switch that does NAT on the LAN side.. > 6 >  o The VMS box has a static IP address: 192.168.1.96  >  o One Windows 98 PC uses DHCP< >  o Another Windows 98 has a fixed IP address: 192.168.1.90< >    It's running the Mocha LPD package used by the VMS box. > J > It appears that my ISP assigns a static IP address to their cable modem,I > so I registered the VMS box at www.dyndns.org with a static IP address.h > L > I had to define a subsititute domain in UCX to make the VMS system capable > of sending & receiving email.b > J > --Jerry Leslie   leslie@clio.rice.edu  (my opinions are strictly my own); >   Note: leslie@jrlvax.houston.rr.com is invalid for email"  G Configuring mail to be a pop server and avoid being a relay is a pretty H good balancing act.  I finally got it working (including adding a numberG of net blocks for Asia/Pacific to my BAD-CLIENTS list.  this has reduce G spam significantly.  I still need to test this box as to whether or nothG I have created an open relay, I don't think so, but I need to be sure. vE To use Good-Clients with POP, I had to set the RELAY option on with a-* good clients list. That apparently works.   E I like www.zoneedit.com for my dynamicDNS provider. I am on DSL which A uses PPPoE, I also have a linux box running a perl script every 5oC minutes to log into the router and check the IP address.  If it hashD changed, it then updates zoneedit, if not, it goes back to sleep.  I@ have some pieces of this ported to VMS, but it is not completely transportable yet.   --   Regards,  7 Michael Austin            Registered Linux User #261163!7 First DBA Source, Inc.    http://www.firstdbasource.comh Sr. Consultant 704-947-1089 (Office)r 704-236-4377 (Mobile)e   ------------------------------  % Date: Thu, 25 Apr 2002 08:46:04 -0700t# From: "Tom Linden" <tom@kednos.com>iG Subject: RE: somewhat off-topic: is there a DSL router with a console?!n9 Message-ID: <CIEJLCMNHNNDLLOOGNJICEFPENAA.tom@kednos.com>a   >   -----Original Message-----< >   From: Michael Austin [mailto:maustin@firstdbasource.com]* >   Sent: Thursday, April 25, 2002 8:22 AM >   To: Info-VAX@Mvb.Saic.ComnK >   Subject: Re: somewhat off-topic: is there a DSL router with a console?!- >b >a >   Jerry Leslie wrote:a >   > 8 >   > Michael Austin (maustin@firstdbasource.com) wrote: >   > : Phillip Helbig wrote:y	 >   > : >s< >   > : > For my hobbyist setup, I have a LAN behind an ISDN >   router.  I have fixedhK >   > : > IP addresses.  I am thinking of adding DSL because flat rates arel@ >   > : > available.  I don't know of anyone who offers fixed IP >   addresses with@ >   > : > DSL.  DHCP is not an option because, again, making the >   names of theB >   > : > machines known to the world is an extra service which is >   probably not8 >   > : > available with a standard cheap flat-rate DSL.	 >   > : > J >   > : > The ISDN connection times out after 2 minutes.  If I initiate anA >   > : > outgoing connection, I pay for it as part of the normalr >   phone bill.  IfrB >   > : > the ISP creates a connection to me, he charges me at the >   standard phone@ >   > : > rate.  (My telephone bill is such that I don't have to >   pay for outgoingB >   > : > calls on Sundays and holidays, so I initiate an outgoing >   connection and3 >   > : > keep it alive from midnight to midnight.)R	 >   > : > B >   > : > It seems that, at least for now, I will have to keep the >   ISDN router forpC >   > : > incoming stuff.  For outgoing stuff, however, it might beh
 >   worth theoC >   > : > trouble to go through a DSL router, since I wouldn't havet >   to pay for it.@ >   > : > (I think that this will save a lot of costs, but there >   WILL be someA >   > : > additional cost since under the new scheme all incomingp >   connections will? >   > : > be initiated by the ISP---in the past, if an outgoing >   connection was0 >   > : > already there, no new one was needed).	 >   > : > ? >   > : > I suppose it is no problem to have two routers in theo >   same LAN---one theA >   > : > present ISDN router and one a new DSL router.  I assume  >   that by settingsK >   > : > the DSL router as the default gateway (and, perhaps, changing theo7 >   > : > ALTERNATE GATEWAY in SMTP as well as the bindw >   resolver---so that allB >   > : > external machines I need to access are available via DSL >   (at the moment? >   > : > they are at the ISP and I'm not sure if he will allowi >   just anyone to useE >   > : > them)), outgoing stuff will not use the ISDN router at all. 	 >   > : > > >   > : > So far, so good, unless I've missed something, which >   hopefully someone  >   > : > here could point out.0	 >   > : > C >   > : > It would be nice to be able to connect to my LAN from theS >   outside viac< >   > : > DSL as well---without a fixed IP address.  I could >   initiate connections? >   > : > periodically automatically to where I want to connectc >   from, see what IP C >   > : > address (that of the DSL router) they come from, and then  >   TELNET to thes< >   > : > router.  However, I want to get from the router to >   machines on the LAN.	 >   > : >nJ >   > : > I've heard of routers which have a "TELNET console": once one isI >   > : > connected to it (via telnet, dial-up, whatever), one can TELNETiC >   > : > elsewhere from it (sort of like the console on a terminal 
 >   server ornJ >   > : > whatever).  Are there DSL routers which have this functionality? >   > :o >   > :iB >   > : If you want to do it right, get an external ethernet modem >   and a SOHOB >   > : router (I use the Linksys BEFSR41.  I have SMTP server and >   News ServermL >   > : (downloads from my ISP) that works well.. .they don't "allow" ftp or@ >   > : http servers so I have those ports blocked at the router >   and open the >   > : ones I need. >   > :iJ >   > : I also have a PERL script called ddclient (from ZONEEDIT.COM) thatL >   > : updates my dynamic IP address (every 5 minutes) -- but it runs on my= >   > : Linux box.  I have not ported it to run as a detached  >   process on my VMS : >   > : box yet. I have extracted the usable parts and get >   "@getip.pl" to getC >   > : the current. My VMS box has a static non-routable 192.168..P >   IP address.y9 >   > : DNS points to local domain and uses DNS from ISP.f >   > :q@ >   > : 2 VMS boxes (Alpha 2100's - one is currently powered off >   due to heat). >   > : 2 Linux PC's - (133 and 166Mhz Pent 1) >   > : 2 Windoze (98 and ME)L >   > : 1 Laptop (WNT4.0)u  >   > : an additional 8-port hub >   > : L >   > That's similar to my cable modem router setup. The router is a LinksysA >   > BEFSR41 router/4-port switch that does NAT on the LAN side.o >   >I: >   >  o The VMS box has a static IP address: 192.168.1.96$ >   >  o One Windows 98 PC uses DHCP@ >   >  o Another Windows 98 has a fixed IP address: 192.168.1.90@ >   >    It's running the Mocha LPD package used by the VMS box. >   >tA >   > It appears that my ISP assigns a static IP address to their  >   cable modem,A >   > so I registered the VMS box at www.dyndns.org with a staticl >   IP address.w >   >MA >   > I had to define a subsititute domain in UCX to make the VMSs >   system capable# >   > of sending & receiving email.  >   > = >   > --Jerry Leslie   leslie@clio.rice.edu  (my opinions are" >   strictly my own)? >   >   Note: leslie@jrlvax.houston.rr.com is invalid for email  > K >   Configuring mail to be a pop server and avoid being a relay is a prettysL >   good balancing act.  I finally got it working (including adding a numberK >   of net blocks for Asia/Pacific to my BAD-CLIENTS list.  this has reducenK >   spam significantly.  I still need to test this box as to whether or notlJ >   I have created an open relay, I don't think so, but I need to be sure.I >   To use Good-Clients with POP, I had to set the RELAY option on with ae- >   good clients list. That apparently works.   K Works for me too, but I have only my local IPs in the good client list, butmK on a7.1-2 system I had UCX4.2 running which was running smtp wide open.  So 4 I had to shut that down.  So check all your systems.   Why do you need dynamicDNS?y > I >   I like www.zoneedit.com for my dynamicDNS provider. I am on DSL whichhE >   uses PPPoE, I also have a linux box running a perl script every 5SG >   minutes to log into the router and check the IP address.  If it hashH >   changed, it then updates zoneedit, if not, it goes back to sleep.  ID >   have some pieces of this ported to VMS, but it is not completely >   transportable yet. >  >   -- >   Regards, > ; >   Michael Austin            Registered Linux User #261163u; >   First DBA Source, Inc.    http://www.firstdbasource.comA >   Sr. Consultant >   704-947-1089 (Office)  >   704-236-4377 (Mobile)q >s   ------------------------------  % Date: Thu, 25 Apr 2002 12:59:25 -0400l1 From: Michael Austin <maustin@firstdbasource.com>tG Subject: Re: somewhat off-topic: is there a DSL router with a console?!l2 Message-ID: <3CC835ED.25812DF6@firstdbasource.com>   Tom Linden wrote:. >   > >   -----Original Message-----> > >   From: Michael Austin [mailto:maustin@firstdbasource.com] <snip> f > Why do you need dynamicDNS?"  E With out it, I would need to know the current IP address assigned viaiE PPPoE so I can now SSH/ftp to my box from any internet connection.  IoF travel(ed) quite a bit and because the WAN IP address changes every soC often due to PPPoE, (no static IP address) I can get to it using myaD domain name and not need to have someone tell me what the IP address> is.  Also, it is my email and news server so mail addressed to9 <whatever>@firstdbasource.com will be accepted by my box.    --   Regards,  7 Michael Austin            Registered Linux User #261163o7 First DBA Source, Inc.    http://www.firstdbasource.coms Sr. Consultant 704-947-1089 (Office)m 704-236-4377 (Mobile)v   ------------------------------   Date: 25 Apr 02 05:11:13 PST From: mckinneyj@cpva.saic.comp) Subject: Re: Suggestion: SET DEVICE/RESETT( Message-ID: <Mchp0B3fjK1J@cpva.saic.com>  ( In article <3CC77402.E91255B5@fsi.net>, 3 "David J. Dachtera" <djesys.nospam@fsi.net> writes:f  > mckinneyj@cpva.saic.com wrote: >> l [...snip...] >> >J >> > I recently discovered - these drives and their younger cousins (TZ88,I >> > etc.) do this to the HSJs, as well. Ya gotta delete them and add 'emg8 >> > back before youy can use 'em again after a failure. >> > >> > :-( >> > >> > --# >> > David J. Dachtera >> > dba DJE Systems >> > http://www.djesys.com/d >> >- >> > Unofficial Affordable OpenVMS Home Page: $ >> > http://www.djesys.com/vms/soho/ >> wI >> It's been a while so I forget all the details, but, in the instance of7G >> a tape drive hung off an HSJ you can frequently recover use of it by:I >> issueing a SET SECURITY command targeting the device. Try changing themI >> the protection mask to something else and back again. A side effect ofvI >> this is that a bit in some status mask gets cleared. I don't recall if/A >> the bit is within one of the UCB fields or the ORB. Anyway, myTF >> recollection is that the bit represented "damaged" hardware. Sorry,F >> I can't remember all the details (it's been a while); don't know if& >> this would help with JF's TK drive. > H > My problem was that the HSJ reported the device as "misconfigured - noI > device at PTL" or something. VMS simply reported it as off-line. So, noi4 > matter what you did, the HSJ still saw it as AWOL. >   G Yep, the HSJ can't serve it if it doesn't know what it is. The instanceoE I described was one where VMS was the confused party. My recollectioniD is that possesion of the BYPASS privilege also allowed access to the% drive when the "damaged" bit was set.e   - Jim    ------------------------------    Date: 25 Apr 2002 10:27:42 -0700+ From: ghairst@mcvh-vcu.edu (G. W. Hairston)o$ Subject: System slowdowns with RWSCS= Message-ID: <1ef525c1.0204250927.1d6e65f8@posting.google.com>n  E We are currently running VMS7.3 and DSM 7.2-1 cluster environment andt# experencing the following problems;oA The VMS cluster went into slowdown last night.  It looks like the-D slowdown started later in the backups than previously, I crashed theD system shortly after that with out attempting to shutdown DSM.  BothD nodes were crashed and rebooted at the same time to get better dumps  C Of the dozen processes in RWSCS state, all were "queries" The RWSCS@F processes in AP-Node were suspended, while those in Y2K-Node continued to get minimal CPU time.  B We have been having these problems since we cluster the systems inE Janary 2002, then we upgraded to v7.3, but the same problem reappear.oD Of course we are working with the Vendors, but has anyone experienceE any of these symptons and how to pinpoint the problem. Thanks for anyh suggestions.   ------------------------------    Date: 25 Apr 2002 05:39:12 -0700( From: bob@instantwhip.com (Bob Ceculski)  Subject: Re: TCPWare Vs MultiNet= Message-ID: <d7791aa1.0204250439.65abeabc@posting.google.com>t  \ carl@gerg.tamu.edu (Carl Perkins) wrote in message news:<24APR200219061084@gerg.tamu.edu>...S > In article <3CC69517.366D@vcn.bc.ca>, Thomas Dzubin <dzubint@vcn.bc.ca> writes...t > }issinoho wrote:K > }> I've been plowing through Process Software's site trying to decide (a)gI > }> why do they produce 2 high-end IP stacks which look to be doing much = > }> the same thing, and (b) which one should I purchase/use.I > } F > }In 1997, Process software (makers of TCPware), bought Multinet from' > }Cisco who bought it from TGV in 1996 K > }I'm glad they're separate products since I've been a Multinet user sincewD > }the early 1990s, but yeah...good question...I'm not sure why they  > }haven't merged them together. > } 	 > }Thomasn > F > They are, very slowly, merging them together. They are starting with< > various things that run in separate processes. If you do a! > $ SEARCH MULTINET:*.COM TCPWARE > > you'll find that some of them check to see if it is MultinetD > or TCPware that you are running (it's two of them in V4.3A, it may > be more in 4.4). > E > TCPware now uses the SMTP server that was originally from Mutlinet. G > They also apparently share the SSH stuff as of 18-Jul-2000 (accordingd$ > to the comment in START_SSH.COM).  > 
 > --- Carl  G except the strength of TCPware is that because it is still based on the  VMS kernel, it runs crisper ...    ------------------------------  % Date: Thu, 25 Apr 2002 09:14:19 +0100e% From: Alan Greig <a.greig@virgin.net>oX Subject: Re: The 'tone' of c.o.v. (was Re: Prediction:  VMS lives, merger or no merger!)8 Message-ID: <aeefcu4emtnr3r3f06842tjbush0vk402t@4ax.com>  F On Wed, 24 Apr 2002 18:19:06 GMT, "Bill Todd" <billtodd@metrocast.net> wrote:     >hL >Whoops - ISTR that the C run-time library (assuming that this is what DavidJ >was performing his measurements with) may in fact emulate rather than useL >RMS, and underneath that emulation use the file system directly.  If so, my  E I'm fairly sure the C RTLis still layered on top of RMS. However workvE is underway to allow it to talk directly to the filesystem as part of B the Unix integration work. Possibly ths is even available with theF latest C RTL that's part of the GNV work but I haven't looked closely.   I could be wrong though.  L >statement above was incorrect:  while the path-length to a RAM disk in suchF >use is noticeably longer than that to the Unix system cache, it's not >because of RMS. >oE >The VIOC should provide a somewhat shorter path length, but may havehL >limitations (ISTR some 35-sector max cache size being mentioned) that wouldM >prevent the caching efficiencies that the Unix cache (or RMS global buffers)eK >can attain.  And while XFC should improve this situation as well, it's notuK >quite there yet.  However, I also STR that the C run-time library emulatespH >RMS sufficiently faithfully that it can use the buffer-size informationL >settable externally via DCL:  if David did not bump up that size, he wasn'tG >doing everything he could to minimize *average* operation path length.w >d >- billt >e >8   -- Alan   ------------------------------  % Date: Thu, 25 Apr 2002 13:59:45 +0100(T From: Andrew Harrison SUNUK Consultancy <andrew_nospam.remove_this.harrison@sun.com>X Subject: Re: The 'tone' of c.o.v. (was Re: Prediction:  VMS lives, merger or no merger!)& Message-ID: <3CC7FDC1.8050207@sun.com>   Rob Young wrote:   > In article <3CC6C1AD.9030007@sun.com>, Andrew Harrison SUNUK Consultancy <andrew_nospam.remove_this.harrison@sun.com> writes:  >  > J >>I am sure that altering the default settings helps, but if David MathogsI >>benchmark results for OpenVMS and other OS's were correct and if I readrG >>them correctly changing the default settings improved performance butiE >>not by enough to make OpenVMS filesystem I/O comparable with modern 	 >>UNIX's.r >>F >>Its also worth pointing out that David did not use any of the fasterH >>UNIX filesystem options. He did not run UFS+ on Solaris in direct modeF >>or use VxFS or QFS all of which could have dramatically improved theC >>performance when compared with default UFS+ performance. Most bigkE >>DBMS servers use either VxFS or UFS+ with Direct I/O (standard partiG >>of UFS). These both bypass the UFS buffer cache for filesystem writese? >>in order to improve performance and reduce memory contention.o >>G >>The OpenVMS filesystem also does not seem to support a mechanism likeiC >>read clustering which can improve read performance without makingd >>changes to apps. >> >> > D > 	Support "read clustering?"  No... and a good point as sequential E > 	pre-fetch at the controller still means you have to go out to the aF > 	controller.  However, XFC (eXtended File Cache) futures (2.x) show D > 	sequential pre-fetch in the cache which will gain parity there.  * > 	Likewise, RDBMSes will pre-fetch today. >     ; However XFC still doesn't work reliably so this is a future % Solaris UFS has had it since day one.        >  >>I >>I think I covered this in the rest of my posting. Compaq has squandered-J >>most of the leads it had in areas like clustering. Even more gastly theyJ >>actually gave OpenVMS clustering IP to Microsoft which fortunately being >>MS they squandered.C >> >> > A > 	Why?  Like many , MS suffers from NIH* syndrome.  And actually B > 	occurs within the same company across OS platforms, (surprise!) >     8 As I said fortunately MS sqandered what they were given.   Regardsj Andrew Harrison    ------------------------------  % Date: Thu, 25 Apr 2002 14:04:36 +0100eT From: Andrew Harrison SUNUK Consultancy <andrew_nospam.remove_this.harrison@sun.com>X Subject: Re: The 'tone' of c.o.v. (was Re: Prediction:  VMS lives, merger or no merger!)$ Message-ID: <3CC7FEE4.60306@sun.com>   Bill Todd wrote:  % > "Andrew Harrison SUNUK Consultancy" ? > <andrew_nospam.remove_this.harrison@sun.com> wrote in messager" > news:3CC6C1AD.9030007@sun.com... > N > Without knowing exactly what you mean by 'read clustering', I will note thatJ > VMS  1) supports facilities that make files contiguous, or optionally asE > contiguous as possible, on disk (and even finer 'placement control' N > mechanisms, but they aren't relevant to this particular point),  2) supportsE > (via specifiable buffer sizes) variable access granularity to files J > (contiguous or not), such that each disk request brings in a specifiableF > amount of data (and this can be controlled on a per-file basis by anB > application or by setting values outside the application - i.e.,M > transparently), and  3) supports asynchronous multi-buffering for both readpI > and write access (I'm not certain whether this option can be controlled,H > outside the application, but it is easy to do inside the application). >     = Clustering allows pre-fetching without having to code for it.      regards    Andrew Harrisono   ------------------------------    Date: 25 Apr 2002 09:35:47 -0500+ From: young_r@encompasserve.org (Rob Young),X Subject: Re: The 'tone' of c.o.v. (was Re: Prediction:  VMS lives, merger or no merger!)3 Message-ID: <lTC9sqkpZYsM@eisner.encompasserve.org>b  } In article <3CC7FDC1.8050207@sun.com>, Andrew Harrison SUNUK Consultancy <andrew_nospam.remove_this.harrison@sun.com> writes:  >  >  > Rob Young wrote: >  >> In article <3CC6C1AD.9030007@sun.com>, Andrew Harrison SUNUK Consultancy <andrew_nospam.remove_this.harrison@sun.com> writes: >> t >>  K >>>I am sure that altering the default settings helps, but if David Mathogs J >>>benchmark results for OpenVMS and other OS's were correct and if I readH >>>them correctly changing the default settings improved performance butF >>>not by enough to make OpenVMS filesystem I/O comparable with modern
 >>>UNIX's. >>> G >>>Its also worth pointing out that David did not use any of the faster I >>>UNIX filesystem options. He did not run UFS+ on Solaris in direct modesG >>>or use VxFS or QFS all of which could have dramatically improved theoD >>>performance when compared with default UFS+ performance. Most bigF >>>DBMS servers use either VxFS or UFS+ with Direct I/O (standard partH >>>of UFS). These both bypass the UFS buffer cache for filesystem writes@ >>>in order to improve performance and reduce memory contention. >>> H >>>The OpenVMS filesystem also does not seem to support a mechanism likeD >>>read clustering which can improve read performance without making >>>changes to apps.h >>>. >>>  >> nE >> 	Support "read clustering?"  No... and a good point as sequential nF >> 	pre-fetch at the controller still means you have to go out to the G >> 	controller.  However, XFC (eXtended File Cache) futures (2.x) show iE >> 	sequential pre-fetch in the cache which will gain parity there.   + >> 	Likewise, RDBMSes will pre-fetch today.  >>   >  > = > However XFC still doesn't work reliably so this is a future ' > Solaris UFS has had it since day one.E >   ? 	However, VMS is considerably more sophisticated/complicated ino> 	several regards.  XFC must contend with shared everything and? 	be poised for the future when it migrates into Galactic Shared > 	Memory.  Is XFC "broken" today?  Yes, and admittedly so.  ButG 	as has been noted in this forum extended field testing is taking placed 	and it should ship RSN.  @ 	But not to belabor the point but I acknowledge above: "Support : 	"read clustering?"  No... and a good point as sequential C 	pre-fetch at the controller still means you have to go out to the o
 	controller."h  C 	But don't get too cheery XFC is broken.  VMS is very sophisticatede@ 	and complicated and well engineered, caching is in the critical> 	path and is a very difficult piece but they will get it right? 	and move on.  XFC is a nice value add and the SDA info is veryM
 	powerful.  : 	Solaris had it from day one... sure.. much simpler model.  D 	When will Solaris have a Distributed Lock Manager in Cluster-commonE 	shared memory?  When will Solaris have Cluster members communicatingpC 	through shared memory?  When will Solaris have a common filesystemb' 	cache in Cluster-common shared memory?   = 	But I understand .... with Unix you do the best you can with  	what you have to work with.   				Rob    ------------------------------  # Date: Thu, 25 Apr 2002 16:16:53 GMT * From: "Bill Todd" <billtodd@metrocast.net>X Subject: Re: The 'tone' of c.o.v. (was Re: Prediction:  VMS lives, merger or no merger!)C Message-ID: <VZVx8.125644$XV5.11200239@bin4.nnrp.aus1.giganews.com>s  # "Andrew Harrison SUNUK Consultancy" = <andrew_nospam.remove_this.harrison@sun.com> wrote in message   news:3CC7FDC1.8050207@sun.com... >r >. > Rob Young wrote:   ...   D > > Support "read clustering?"  No... and a good point as sequentialE > > pre-fetch at the controller still means you have to go out to the F > > controller.  However, XFC (eXtended File Cache) futures (2.x) showC > > sequential pre-fetch in the cache which will gain parity there.B+ > > Likewise, RDBMSes will pre-fetch today.s > >i >t > = > However XFC still doesn't work reliably so this is a futuree' > Solaris UFS has had it since day one.   I Of course, 'day one' for Solaris really isn't that long ago, since Sun OSLI definitely did *not* have adequate pre-fetch mechanisms (vanilla BerkeleykE FFS looking 'adequate' only in comparison to the abomination that them original Unix filesystem was).  L IIRC it wasn't until after 1990 that Larry McVoy et al. started grouping (ifF this is what you mean by 'read clustering' 4 KB filesystem blocks intoG contiguous 56 KB chunks (still rather pathetically small, especially byPG today's standards) to avoid the worst aspects of fragmentation that FFSoG allocation groups failed to address ('near' rather than 'adjacent' justiH doesn't cut the mustard in disk access).  By contrast, extent-based fileL systems like VMS's have provided as much file contiguity as the disk allowedH since their inception, and by virtue of its support for user-specifiableL buffer sizes (and/or unbuffered access to an application buffer of arbitraryD size) RMS has always allowed users to take as much advantage of said contiguity as they cared to.  L And, as I mentioned earlier, RMS definitely *does* support (and has for overI two decades) multi-buffered read-ahead/write-behind (if *that's* that you K mean by 'read clustering'), and more controllably than Solaris (in that youoJ can specify both the sizes and number of said buffers, then let RMS manage them transparently).  I So once again, the only 'deficiency' VMS suffers from in this area is thehK fact that the user (either in the application or, in some cases, optionallyIE via externally-settable parameters) must specify what s/he wants, the)J defaults being ridiculously conservative in today's environment.  XFC will< fix even that, but the fix is without question long overdue.   - bill   ------------------------------  # Date: Thu, 25 Apr 2002 16:21:45 GMT * From: "Bill Todd" <billtodd@metrocast.net>X Subject: Re: The 'tone' of c.o.v. (was Re: Prediction:  VMS lives, merger or no merger!)@ Message-ID: <t2Wx8.10351$fg.1170804@bin2.nnrp.aus1.giganews.com>  # "Andrew Harrison SUNUK Consultancy" = <andrew_nospam.remove_this.harrison@sun.com> wrote in message  news:3CC7FEE4.60306@sun.com... >  >e > Bill Todd wrote: >T' > > "Andrew Harrison SUNUK Consultancy":A > > <andrew_nospam.remove_this.harrison@sun.com> wrote in messages$ > > news:3CC6C1AD.9030007@sun.com... > >sK > > Without knowing exactly what you mean by 'read clustering', I will notep thatL > > VMS  1) supports facilities that make files contiguous, or optionally asG > > contiguous as possible, on disk (and even finer 'placement control'wG > > mechanisms, but they aren't relevant to this particular point),  2)  supportsG > > (via specifiable buffer sizes) variable access granularity to files>L > > (contiguous or not), such that each disk request brings in a specifiableH > > amount of data (and this can be controlled on a per-file basis by anD > > application or by setting values outside the application - i.e.,J > > transparently), and  3) supports asynchronous multi-buffering for both readK > > and write access (I'm not certain whether this option can be controlledoJ > > outside the application, but it is easy to do inside the application). > >  >i >n? > Clustering allows pre-fetching without having to code for it.   J Ah.  Well, while I think I've now addressed this adequately elsewhere, forK completeness I'll reiterate here that you don't have to write *code* to geteJ RMS to do this (assuming you mean asynchronous, multi-buffered pre-fetch),L you just specify the size and number of buffers it should use and ask it to.F And if you just mean reading data in larger chunks than you explicitlyK requested, RMS has *always* done that for small requests - the only problemeL being that its default interpretation of 'large' (8 KB) dates back 25 years.   - bill   ------------------------------  # Date: Thu, 25 Apr 2002 06:50:43 GMTe3 From: Carl Nelson <carl.nelson@mcmail.maricopa.edu>h/ Subject: Re: VAX decommissioning/secure HD wipe-3 Message-ID: <3CC7A748.90D725C4@mcmail.maricopa.edu>2  F   I am a very firm believer in the policy of making the punishment fit the crime, so...  H   After all the writing/erasing/writing/erasing/initializing is done....  H   Write a command procedure to fill the disk with a bazillion files withC the complete lyrics of "It's a small world after all" in them. Then.G delete them. Do NOT erase or overwrite them. Anybody that wants to looklB at the residual data on your disks will have something to look at.  =   How many blocks of this would YOU look at before giving up?i   --Carl   ------------------------------  % Date: Thu, 25 Apr 2002 03:17:05 -0400h- From: JF Mezei <jfmezei.spamnot@videotron.ca>w/ Subject: Re: VAX decommissioning/secure HD wipeC, Message-ID: <3CC7AD6E.B26A8164@videotron.ca>   Carl Nelson wrote:J >   Write a command procedure to fill the disk with a bazillion files withE > the complete lyrics of "It's a small world after all" in them. Theni. > delete them. Do NOT erase or overwrite them.  L If you're going to do that, you might has well put in something interesting.C 	"Your boss is having an affair with your wife" would be better :-)l  N However, if you are going to be doing that, shouldn't the text be large enoughB to fill the minimum file size so that all blocks are overwritten ?  M If the cluster size is 9 blocks and you create a file that has just 40 bytes,.L will the remainder of the first block as well as the last 8 blocks be filledM with zeros on the disk, or will the disk contain indeterminate data, with RMSw9 just faking zeros beyond end of file when you access it ?i   ------------------------------  % Date: Thu, 25 Apr 2002 09:13:00 +0100 ( From: Nic Clews <sendspamhere@127.0.0.1>/ Subject: Re: VAX decommissioning/secure HD wipeu) Message-ID: <3CC7BA8C.B02D132B@127.0.0.1>r   Larry Kilgallen wrote: > H > INITIALIZE/ERASE is good enough for most purposes other than defendingH > against some national intelligence agency.  If the latter is a problemE > with which you must deal, you should be talking to your accreditingnA > agency rather than to us furriners from south of the border :-)a  G Larry, following on from the demonstration of TECO's ability, how about  a TECO macro whichF walks all over the files 11 structure? Then, before you start erasing,G you kick off with something that looks nothing like a files 11 disk :-)r   -- y( Regards, Nic Clews CSC Computer Sciences nclews at csc dot comf   ------------------------------    Date: 25 Apr 2002 21:46:31 +0800, From: Paul Repacholi <prep@prep.synonet.com>/ Subject: Re: VAX decommissioning/secure HD wipel- Message-ID: <87adrrlwlk.fsf@prep.synonet.com>-  , WILLIAM WEBB <WWEBB1@email.usps.gov> writes:  @ > 5) Disassemble drive, sand platter surfaces, dissolve in acid;  cC >    (don't laugh, I've worked with somebody who had to do #5 while- >     in the military)  C One person to destroy platters, one to certify the #1 had destroyedd! platters, third as safty officer.    --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.y@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------    Date: 25 Apr 2002 08:52:14 -0700" From: cstranslations@msn.com (Joe)/ Subject: Re: VAX decommissioning/secure HD wipea= Message-ID: <d56d1c2d.0204250752.21085707@posting.google.com>l  a JF Mezei <jfmezei.spamnot@videotron.ca> wrote in message news:<3CC71753.813C5509@videotron.ca>...u > Hoff Hoffman wrote: M > >   If you do not have a bootable environment nor current distribution, youhK > >   can build a standalone BACKUP root on one of the local disk spindles,RM > >   and BACKUP/IMAGE/INIT restore some random saveset onto the system disk.g >fL > I suspect that there were a few files left on, but from a proprietary dataO > perspective, that data would have been Digital proprietary data (the very lowrM > level basic VMS files needed to keep a minimally booted VMS system running,y? > with the DELETE.EXE file which could haven't deleted itself).y  D Is that necessarily true? sys$common:[sysexe]delete.exe isn't a veryB big file. Assuming the necessary privileges and that it's entirelyF memory resident when it goes to delete itself (safe bets) why wouldn't it?p  ? I can delete the executable (resulting from) the BASIC program:d  ! there: print "hello" \ goto thereo  B out from under itself while it's running round in circles and it's? just like the "Energizer Bunny." It keeps going, and going, and  going...   ------------------------------   Date: 25 Apr 2002 16:45:05 GMT1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon)n/ Subject: Re: VAX decommissioning/secure HD wiper, Message-ID: <aa9bqh$11v5$1@info.cs.uofs.edu>  , In article <3CC71559.7C1B3795@videotron.ca>,0  JF Mezei <jfmezei.spamnot@videotron.ca> writes: |> Michael Austin wrote:M |> > more secure do you need it to be?  Once the bit pattern has been writtenOH |> > it is virtually impossible to tell what the bits were before it was
 |> > written.t |> aO |> "virtually" is the keyword here. There are experts who are _supposedly_ ablerO |> to extrapolate data that was written over once. Hence my suggestion to use a P |> few different methods to ensure that enough data was written sufficient times3 |> that the original data is no longer recoverable.p  G According to my Ghost manual 7 passes is the security standard for DOD.e6 However, the program itself supports up to 100 passes.   bill   -- >J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   t   ------------------------------  % Date: Thu, 25 Apr 2002 08:12:03 -0700a# From: "Tom Linden" <tom@kednos.com>d' Subject: VAXCLUSTER license alternativeq9 Message-ID: <CIEJLCMNHNNDLLOOGNJIKEFMENAA.tom@kednos.com>.  : The license fee is almost $8,778 with the CSA discount !!!  @ Now the basic NAS license (which I have )will allow me to accessB the other cluster members from the VAX, but not vice versa.  So...  > can I move one of the VAX drives, say dka200 logical disk$foo,@ that has data on it to an LD container on one of the alphas and  have it appear as disk$foo?I   ------------------------------  % Date: Thu, 25 Apr 2002 13:58:44 +0200 % From: "IdrEASY" <IdrEASY@bigfoot.com>H2 Subject: Re: VT520 and VT420 terminal and PC mouse* Message-ID: <aa8r2p$31ul$1@as201.hinet.hr>  G Maybe somebody are capable to write a little subroutine which will readjF mouse inputs from VT420 or VT520 RS232 port and then recalculate it in1 escape sequences for cursor movement and panning.   D I have draft idea, but not knowledge and good tools to realize this.  0 "IdrEASY" <IdrEASY@bigfoot.com> wrote in message# news:aa21qt$f4c$1@brown.net4u.hr...3 > Hi!. >0H > Does it exist some "chemistry" for using PC mouse on VT terminals (I'd like  > to use it with SMG$ routines)? >o >w   ------------------------------  % Date: Thu, 25 Apr 2002 09:20:23 -0400 5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> 2 Subject: Re: VT520 and VT420 terminal and PC mouse, Message-ID: <aa9042$7dsf$1@lead.zk3.dec.com>  L Try this (these were from the VWS emulator, which went into DECterm I think,' and which should be in some terminals).    Enable Locator Reports (DECELR)f  I CSI Ps1;Ps2 'z  (that is an "escape [" with 2 parameters seperated by a ; . ending with a wingle quote and a lowercase z).  D This enables/disabled locator reports.  The first parameter can be 0I (disable), 1 (enable) and 2 (enable for a single report "one shot").  ThepL second parameter can be 1 (physical units, i.e. pixels), 2 (Character cells) and 3 (Regis user coordinates).t  L So, $[1;2'z  should enable reports, and the values will be in character cell coordinates.   Select Locator Events (DECSLE)  	 CSI Ps '{e  L Defines the type of events the locator reports.  0 (only explicit requests),F 1 (report button down events), 2 (no button down reports) 3 (button up" reports) 4 (no button up reports).  F So, $[1;4'{  means give me button downs, but not button ups.   And the6 reports will come asynch (without a explicit request).  " Request Locator Position (DECRQLP)  	 CSI Ps '|c  K The only value for Ps is 0 or 1 (which means the same thing) and requests ac locator report.   7 So $[1'|  will cause a position report to be generated.@   The locator report (DELRP)   CSI Pe ; Pb ; Pr ; Pc ; Pp &wv  F Pe is the event code - 0 (explict request - but no mouse available), 1I (explicit request), 2 left button down, 3 left button up, 4 middle buttonOJ down, 5 middle button up, 6 right button down, 7 right button up, 8 fourthE button down, 9 fourth button up, 10 explicit request - but locator ise  outside of the filter rectangle.  ? Pb is a mask of the buttons 1 right, 2 middle, 4 left, 8 fourth   L Pr is the ROW address, Pc is the Column address, Pp is the Page address (for multi-page terminals).  J So, $[2;4;10;7;1&w  means the left button was pressed, only that button is= down, and the cursor location is at (10,7) on the first page.e  L The "filter rectangle" mentioned above was a way to restrict reports only toL a specific area on the screen - CSI Pt ; Pl ; Pb ; Pr 'w    -- where Pt is tJ he top, Pl is left, Pb is bottom, and Pr is right - in coordinates definedG by the DECELR.  This is a one-shot thing, and is disabled as soon as anrE event takes place.  So $[0;0;20;20'w  would set the box from (0,0) toaJ (20,20) as sensitive for a event (events outside the area don't get sent).  - Lastly, Select Locator Cursor Style (DECSLCS)e  	 CSI Ps )ut  ? 0 = Crosshair, 1 - Box corner  (So $[0)u  turns on a crosshair.         Terry Marosites wrote in messageC <14987C0DA438094D91DAA6CED58D060A6E7760@dalntexch2.unitedad.com>...e > I >The VT330's Plus also had a mouse , To access it you used Regis commands > >and/or escape commands , at least I think it has been awhile. > -----Original Message------ >From: Jim Agnew [mailto:agnew@mail2.vcu.edu]c) >Sent: Wednesday, April 24, 2002 12:00 PMC >To: Info-VAX@Mvb.Saic.Com3 >Subject: Re: VT520 and VT420 terminal and PC mouset >iI >Actually, we have vt340's with mice holes in them.  I think they're used  >with GKS apps...e >nH >My homebrew GKS (Graphics Kernel System) fired up mouse crosshairs whenF >used with a Kermit pc running windows and therefore had a mouse.  theG >VMS s/w detected the mouse, and the display automagically compensated.f8 >I suppose the vt430 with a dec mouse would do the same. >o
 >Jim Agnew >h >IdrEASY wrote:t >>K >> Yes, I know. I thought that it maybe could work like mouse in Clipper or  >> Norton Commander under DOS. >>3 >> "Ade" <adrian.birkett@ic24.net> wrote in message = >> news:z01x8.2748$Cp4.253301@news-binary.blueyonder.co.uk...AI >> > Off the top of my head, probably NO. VT terminals are character celliK >> > terminals, not windows based. Look for a DecWindows terminal (VXT?) or-J >> > something similar. I'm sure someone else can confirm or give a better >> > explanation.  >> > >> > Ade >> >G >> > === If the world is a stage and we are merely actore, I WANT A NEW  >> > SCRIPTWRITER !!! ===t >> >5 >> > "IdrEASY" <IdrEASY@bigfoot.com> wrote in messagew( >> > news:aa21qt$f4c$1@brown.net4u.hr...
 >> > > Hi! >> > >H >> > > Does it exist some "chemistry" for using PC mouse on VT terminals (I'd	 >> > like'% >> > > to use it with SMG$ routines)?M >> > > >> > > >> > >> >   ------------------------------  % Date: Thu, 25 Apr 2002 08:04:24 -0700 ' From: David Mathog <mathog@caltech.edu>d, Subject: Web based/encrypted printing (O.T.)+ Message-ID: <3CC81AF8.E96C488A@caltech.edu>.  A (Somewhat O.T. but some OpenVMS admins may be dealing with this.)"  D Pretty much every printer with a network connection these days comesG with a web server for administration purposes.  Yet I've never seen onet: that would accept print jobs through that web interface.    6 Furthermore, some sites have a need to print documents? securely.  Never mind the DOD and spooks, the medical communityt" needs this badly.  There was a law: passed a year or two ago that requires very high levels of. security on patients' medical information.  SoC how is an HMO supposed to print anything from the central server tom= a printer near an employee?  All print protocols I know of gom: across the net unencrypted.  I've heard that one major HMOI is "solving" this with, essentially, 100% local printers - either hanging @ off the end users' machines or in some cases running off a smallK servers which encrypted links for print jobs.  That's an incredibly painful K way to do things since it's much easier to drop network printers in than ton configure local printers.p  = Which brings up the big question - why doesn't anybody ship a > printer with a job upload (POST) function on the web interface9 and the same function through https?  The drivers for thea> PC/Mac/unix/VMS side of such an interface should be completely> trivial to write and the "server" (printer) side only requires> Apache/openssl (ie free) and a tiny amount of CGI code.  Sure,@ it would need more memory in the printer, but given the price ofF memory these days it certainly looks feasible.  And all they'd need toF do to make such a printer compatible with the higher security standard) is to disable http and only accept https.s  + Or does this already exist on some product?n   Regards,   David Mathog mathog@caltech.edu   ------------------------------  % Date: Thu, 25 Apr 2002 11:49:54 -0400i1 From: Michael Austin <maustin@firstdbasource.com>r0 Subject: Re: Web based/encrypted printing (O.T.)2 Message-ID: <3CC825A2.A923E559@firstdbasource.com>   David Mathog wrote:  > C > (Somewhat O.T. but some OpenVMS admins may be dealing with this.)i > F > Pretty much every printer with a network connection these days comesI > with a web server for administration purposes.  Yet I've never seen one : > that would accept print jobs through that web interface. > 8 > Furthermore, some sites have a need to print documentsA > securely.  Never mind the DOD and spooks, the medical community1$ > needs this badly.  There was a law< > passed a year or two ago that requires very high levels of0 > security on patients' medical information.  SoE > how is an HMO supposed to print anything from the central server to ? > a printer near an employee?  All print protocols I know of goi< > across the net unencrypted.  I've heard that one major HMOK > is "solving" this with, essentially, 100% local printers - either hanging B > off the end users' machines or in some cases running off a smallM > servers which encrypted links for print jobs.  That's an incredibly painfuleM > way to do things since it's much easier to drop network printers in than tou > configure local printers.  > ? > Which brings up the big question - why doesn't anybody ship a>@ > printer with a job upload (POST) function on the web interface; > and the same function through https?  The drivers for theo@ > PC/Mac/unix/VMS side of such an interface should be completely@ > trivial to write and the "server" (printer) side only requires@ > Apache/openssl (ie free) and a tiny amount of CGI code.  Sure,B > it would need more memory in the printer, but given the price ofH > memory these days it certainly looks feasible.  And all they'd need toH > do to make such a printer compatible with the higher security standard+ > is to disable http and only accept https.h > - > Or does this already exist on some product?t > 
 > Regards, >  > David Mathog > mathog@caltech.edu  > How many people have access to the network with a sniffer in aH hospital?  Even at goverment "spook" facilities, printers were generallyE hung off of DEC terminal servers which spoke either LAT or TCP/IP. No E encryption.  If the physical plant is inaccessible, then it is a moott point.  F Most network printers don't speak http.  They speak tcpip and the dataB is either PCL or PS or plain text.  They have an http interface to9 configure them, but the data is not transmitted via http.n -- i Regards,  7 Michael Austin            Registered Linux User #261163e7 First DBA Source, Inc.    http://www.firstdbasource.com  Sr. Consultant 704-947-1089 (Office)f 704-236-4377 (Mobile)f   ------------------------------  % Date: Wed, 24 Apr 2002 15:33:22 +0100t4 From: "Chris Sharman" <chris.sharman@ccagroup.co.uk> Subject: web hostingB Message-ID: <1019727593.17846.0.nnrp-10.9e989e7e@news.demon.co.uk>  G We're looking at web hosting in-house, to offer web access to ordering,y tracking, etc to customers.r( The existing data & apps are mostly VMS.L We've used http_server on VMS for our intranet, and done a certain amount of proof of concept there.-  ; Anyone care to offer any opinions, advice, case histories ?hL We need to decide whether we're hosting on VMS, linux, windows, or whatever,1 and if vms, whether to use apache or http_server.4  E Security through obscurity is an obvious attraction of VMS, as is thelI reduced learning curve. Uncertainty over VMS's future is a (fairly minor)a( con. Any other pros & cons (tools etc) ?  L What sort of horsepower would it need ? Would an old 2100A 5/300 do the job,- or should we look for a modern Alphastation ?r   Thanks,v
 Chris Sharmani   ------------------------------  # Date: Thu, 25 Apr 2002 10:27:20 GMTnL From: winston@SSRL.SLAC.STANFORD.EDU ("Alan Winston - SSRL Admin Cmptg Mgr") Subject: Re: web hosting8 Message-ID: <00A0CF91.926A4499@SSRL04.SLAC.STANFORD.EDU>  C In article <1019727593.17846.0.nnrp-10.9e989e7e@news.demon.co.uk>, s  6 "Chris Sharman" <chris.sharman@ccagroup.co.uk> writes:  H >We're looking at web hosting in-house, to offer web access to ordering, >tracking, etc to customers.  ) >The existing data & apps are mostly VMS.-M >We've used http_server on VMS for our intranet, and done a certain amount of. >proof of concept there. >i< >Anyone care to offer any opinions, advice, case histories ?M >We need to decide whether we're hosting on VMS, linux, windows, or whatever,t2 >and if vms, whether to use apache or http_server.   >nF >Security through obscurity is an obvious attraction of VMS, as is theJ >reduced learning curve. Uncertainty over VMS's future is a (fairly minor)) >con. Any other pros & cons (tools etc) ?a  P Linux/Unix/xBSD, then Windows, probably have the most tools available, but a lotF of key tools (Java, Perl, PHP) are there on VMS already, and if you'reK developing your own stuff using people who know VMS, VMS is a very sensible-H platform.  (If you wanted to hire a bunch of contractors to develop yourJ applications, you might get pushed to use something they already know and  are productive on.)r  L If you think of running IIS on Windows you really need to devote an admin toN keeping IIS and Windows up to date; it seems to be a full-time job.  If you'reJ thinking about Linux, you might well be better off with FreeBSD or NetBSD,K which have a lot more history and stability behind them - but none of these M systems are secure without an administrator who knows how to manage them, so  ; you'd be hiring somebody new or training somebody you have.R  K There are people running complex sites on OSU, on WASD, and on CSWS/Apache. M The CSWS people keep porting more tools over; in some cases, you can use themW from any of the web servers.     >lM >What sort of horsepower would it need ? Would an old 2100A 5/300 do the job,b. >or should we look for a modern Alphastation ?  M That last question is really, really hard to answer without having an idea ofuJ your expected load.  That said, my site got along reasonably well serving K 15,000 mostly-static pages a day using OSU with a dual-processor 2100 5/300 L that was also running database, mail, and a few dozen interactive users - weN might see slowish response on stuff like Rdbweb CGI programs (where SQL storedL procedures generate the pages, so there's lots of processing before anythingK gets out), but perfectly adequate response on other CGIs. We like our DS20E  better, though.u   -- Alant  O ===============================================================================a0  Alan Winston --- WINSTON@SSRL.SLAC.STANFORD.EDUM  Disclaimer: I speak only for myself, not SLAC or SSRL   Phone:  650/926-3056nM  Physical mail to: SSRL -- SLAC BIN 69, PO BOX 4349, STANFORD, CA  94309-0210tO ===============================================================================i   ------------------------------  % Date: Thu, 25 Apr 2002 08:46:37 -0400 1 From: Michael Austin <maustin@firstdbasource.com>p Subject: Re: web hosting2 Message-ID: <3CC7FAAD.45C5B820@firstdbasource.com>   Chris Sharman wrote: > I > We're looking at web hosting in-house, to offer web access to ordering,t > tracking, etc to customers. * > The existing data & apps are mostly VMS.N > We've used http_server on VMS for our intranet, and done a certain amount of > proof of concept there.a > = > Anyone care to offer any opinions, advice, case histories ?DN > We need to decide whether we're hosting on VMS, linux, windows, or whatever,3 > and if vms, whether to use apache or http_server.  > G > Security through obscurity is an obvious attraction of VMS, as is thetK > reduced learning curve. Uncertainty over VMS's future is a (fairly minor)b* > con. Any other pros & cons (tools etc) ? > N > What sort of horsepower would it need ? Would an old 2100A 5/300 do the job,/ > or should we look for a modern Alphastation ?. > 	 > Thanks,s > Chris Sharman>  H It will all depend on what your workload will look like, do you need Rdb? access, Oracle access, internet connection bandwidth, #hits perdE day/week/month etc...?  I have a 2100A/200 that works reasonably welle# for very small number of hits/day. o  E BTW, If you are looking for a contractor that can handle web-enablingiC Rdb, give me a call - Both OSU and Apache-(preferred).  I have alsoeF taken a canned application that generates a packing list and created aG web page to use a Compaq wirless IPAQ as a picking system. Works great.- -- - Regards,  7 Michael Austin            Registered Linux User #261163"7 First DBA Source, Inc.    http://www.firstdbasource.comv Sr. Consultant 704-947-1089 (Office)F 704-236-4377 (Mobile)    ------------------------------    Date: 25 Apr 2002 07:04:39 -0700( From: bob@instantwhip.com (Bob Ceculski) Subject: Re: web hosting= Message-ID: <d7791aa1.0204250604.24500b38@posting.google.com>a  ~ "Chris Sharman" <chris.sharman@ccagroup.co.uk> wrote in message news:<1019727593.17846.0.nnrp-10.9e989e7e@news.demon.co.uk>...I > We're looking at web hosting in-house, to offer web access to ordering,2 > tracking, etc to customers.m* > The existing data & apps are mostly VMS.N > We've used http_server on VMS for our intranet, and done a certain amount of > proof of concept there.e > = > Anyone care to offer any opinions, advice, case histories ?sN > We need to decide whether we're hosting on VMS, linux, windows, or whatever,3 > and if vms, whether to use apache or http_server.b > G > Security through obscurity is an obvious attraction of VMS, as is thetK > reduced learning curve. Uncertainty over VMS's future is a (fairly minor)o* > con. Any other pros & cons (tools etc) ? > N > What sort of horsepower would it need ? Would an old 2100A 5/300 do the job,/ > or should we look for a modern Alphastation ?0 > 	 > Thanks,  > Chris Sharmans  D we are doing exactly that ... vms is not security thru obscurity ...C it is security, period.  VMS future is assured mainly thanks to the E defense departments use thru at least 2012, but w/itanium port coming,E I project VMS will live long past that ... as far as a web server, itED depends what you want to do ... Apache is good, WASD don't know, OSUD is ok, but also check out Purveyor, although it is not free like the& others ... as far as a good box, go to   www.islandco.com  E and take a gander at either the pws500au ... for $1000 you can't beat.B it ... or if you expect really heavy traffic for $2000 look at theF xp1000 ... it screams!  If you need a lot of disk space than for $3000A look at the DS20, it has a 7 bay storage works built right in ...aF if you need to stay under $1000, then try the pws433au, for medium andH small traffic it will do nicely ... also soon they will be getting in a B bunch of Alphaserver 800's for cheap, which is also a nice box ... you have plenty of options ...   ------------------------------  + Date: Thu, 25 Apr 2002 14:45:38 +0000 (UTC)l1 From: sssslewis@spyder.mitre.org (Keith A. Lewis)s Subject: Re: web hosting. Message-ID: <aa94qi$97s$1@newslocal.mitre.org>   "Chris Sharman" <chris.sharman@ccagroup.co.uk> writes in article <1019727593.17846.0.nnrp-10.9e989e7e@news.demon.co.uk> dated Wed, 24 Apr 2002 15:33:22 +0100:H >We're looking at web hosting in-house, to offer web access to ordering, >tracking, etc to customers.) >The existing data & apps are mostly VMS.eM >We've used http_server on VMS for our intranet, and done a certain amount of7 >proof of concept there. >u< >Anyone care to offer any opinions, advice, case histories ?M >We need to decide whether we're hosting on VMS, linux, windows, or whatever,,2 >and if vms, whether to use apache or http_server.  J If you have VMS systems available, I wouldn't consider anything else for aJ web server.  But that could be my own bias; I've never set one up on Unix.  - Windows is full of security holes.  Avoid it.o  E I have set up secure VMS webservers using both OSU HTTPD (a.k.a. "the ; Decthreads http server") and Apache (CSWS).  Both are free.   J Both offer https, however, with CSWS it's ready out-of-the-box.  For OSU IG had to find a specific version of SSLEAY and build it.  Now that I havet? upgraded my C compiler, I don't think I could repeat the build.0  H Apache also seems to be more versatile and scaleable.  OSU is based on aJ single multithreaded process; whereas Apache can be configured to run withL several server processes.  Password protection is more straightforward under Apache.a  J My company's penetration squad tried out some windows hacks to break in toG my OSU servers.  They did not penetrate, but their evil URLs did createhJ several dozen infinite looping process (one per attempt), which slowed theI systems to a crawl until I intervened.  I have not tested their attack on: Apache.w  F >Security through obscurity is an obvious attraction of VMS, as is theJ >reduced learning curve. Uncertainty over VMS's future is a (fairly minor)) >con. Any other pros & cons (tools etc) ?   H With the current (buyer's) market for used alphas, you should be able toJ stockpile enough parts to keep your systems running for a decade or more. E As long as SCSI lives, you will even be able to upgrade your disks.  d  K What's your favorite language for CGI scripting?  VMS has DCL, Perl, C(++),v3 and if you're willing to write a DCL wrapper, Java.n  M >What sort of horsepower would it need ? Would an old 2100A 5/300 do the job,l. >or should we look for a modern Alphastation ?  J Whether the 2100A will handle your traffic depends on how much traffic youJ have, obviously.  My opinion:  If you've got money for a DS20E go for it. I You can swap disks with the machine up and running, just like your 2100A,iD but you can get 15,000 RPM disks off-the-shelf that fit the DS20E.    + --Keith Lewis              klewis$mitre.orgg> The above may not (yet) represent the opinions of my employer.   ------------------------------  % Date: Thu, 25 Apr 2002 15:43:29 +0100P% From: Alan Greig <a.greig@virgin.net>< Subject: Re: web hosting8 Message-ID: <je5gcucql12ltc086e5i9sie2nd4kufnj6@4ax.com>  A On 25 Apr 2002 07:04:39 -0700, bob@instantwhip.com (Bob Ceculski)m wrote:  E >we are doing exactly that ... vms is not security thru obscurity ...>D >it is security, period.  VMS future is assured mainly thanks to theF >defense departments use thru at least 2012, but w/itanium port coming  = Err, didn't you just predict that Intel will drop Itanium and1 resurrect Alpha?  F >I project VMS will live long past that ... as far as a web server, itE >depends what you want to do ... Apache is good, WASD don't know, OSUcE >is ok, but also check out Purveyor, although it is not free like thee' >others ... as far as a good box, go tou >t >www.islandco.com  > F >and take a gander at either the pws500au ... for $1000 you can't beatC >it ... or if you expect really heavy traffic for $2000 look at theTG >xp1000 ... it screams!  If you need a lot of disk space than for $3000fB >look at the DS20, it has a 7 bay storage works built right in ...G >if you need to stay under $1000, then try the pws433au, for medium andlI >small traffic it will do nicely ... also soon they will be getting in a  C >bunch of Alphaserver 800's for cheap, which is also a nice box ...o >you have plenty of options ...f   -- Alan   ------------------------------  + Date: Thu, 25 Apr 2002 15:46:37 +0000 (UTC)  From: david20@alpha1.mdx.ac.uk Subject: Re: web hosting+ Message-ID: <aa98ct$bav$1@aquila.mdx.ac.uk>   b In article <aa94qi$97s$1@newslocal.mitre.org>, sssslewis@spyder.mitre.org (Keith A. Lewis) writes: >"Chris Sharman" <chris.sharman@ccagroup.co.uk> writes in article <1019727593.17846.0.nnrp-10.9e989e7e@news.demon.co.uk> dated Wed, 24 Apr 2002 15:33:22 +0100: I >>We're looking at web hosting in-house, to offer web access to ordering,p >>tracking, etc to customers.c* >>The existing data & apps are mostly VMS.N >>We've used http_server on VMS for our intranet, and done a certain amount of >>proof of concept there.a >>= >>Anyone care to offer any opinions, advice, case histories ?nN >>We need to decide whether we're hosting on VMS, linux, windows, or whatever,3 >>and if vms, whether to use apache or http_server.o >tK >If you have VMS systems available, I wouldn't consider anything else for a-K >web server.  But that could be my own bias; I've never set one up on Unix.i >i. >Windows is full of security holes.  Avoid it. >dF >I have set up secure VMS webservers using both OSU HTTPD (a.k.a. "the< >Decthreads http server") and Apache (CSWS).  Both are free. >sK >Both offer https, however, with CSWS it's ready out-of-the-box.  For OSU IcH >had to find a specific version of SSLEAY and build it.  Now that I have@ >upgraded my C compiler, I don't think I could repeat the build. >t  I See http://www.ourservers.net/openvms_ports/openssl/openssl_contents.htmln  J for information on Openssl (which replaced SSLEAY quite some time ago) and< instructions on building SSL support into the OSU webserver.  I >Apache also seems to be more versatile and scaleable.  OSU is based on a K >single multithreaded process; whereas Apache can be configured to run with M >several server processes.  Password protection is more straightforward undert >Apache.  I The current version of Apache provided by COMPAQ in CSWS does not support0G threads. A new version of Apache was produced earlier this month which oN at long last supports threads. Although Unix is better in this regard than VMSL creating large numbers of child processes to handle incoming requests is not good for performance.hL Generally multi-threading is the way to go for high scalability on multi-cpuJ systems. For single CPU systems WASD's AST's (which might be considered to, be lightweight threads) seems to be better.   g  
 David Webb VMS and Unix team leader CCSS Middlesex University   ------------------------------  # Date: Thu, 25 Apr 2002 13:22:51 GMTr& From: "Ed Vogel" <ed.vogel@compaq.com>) Subject: Re: where can I find older DECC?r9 Message-ID: <LqTx8.11$bX5.435322@cacnews.cac.cpqcorp.net>h  . "Tom Linden" <tom@kednos.com> wrote in messageL | and they weren't able to rebuild the C programs using later version of the | compiler.s       Hi Tom,d  C     There may be several easy workarounds using the newer compiler.aK     First, you don't explain exactly what the problem is.  I can guess thate theuK     the newer compiler is emitting more messages than the old one.  If they G     don't want to take the time to look at these messages, just compileo
 /NOWARING.  K     Second, there's a chance that the newer optimizer is making assumptionsp thatH     the older one does not.  We've seen bad code break when upgrading to newertH     compilers.  You can try compilining /NOOPT/ASSUME=NOCLEAN_PARAMETERSA     if the programs are compiling O.K. but not running correctly.r  L     Can you give me any more details about the problems they are having with     the newer compiler?B  @                                                         Ed Vogel@                                                         Compaq C Engineering.   ------------------------------    Date: 25 Apr 2002 06:53:03 -0700 From: wingwong@witty.com (wing):6 Subject: will memory leak program caused system crash?= Message-ID: <873e96d6.0204250553.15fcf454@posting.google.com>s   Hi,f  ; Will a memroy leak program cause system crash in openvms?     T Is there any user account setting/run option to set the max memory use by a program?   Wing   ------------------------------  + Date: Thu, 25 Apr 2002 17:17:18 +0000 (UTC),1 From: sssslewis@spyder.mitre.org (Keith A. Lewis)e: Subject: Re: will memory leak program caused system crash?. Message-ID: <aa9dmu$9n8$2@newslocal.mitre.org>   wingwong@witty.com (wing) writes in article <873e96d6.0204250553.15fcf454@posting.google.com> dated 25 Apr 2002 06:53:03 -0700:(< >Will a memroy leak program cause system crash in openvms?    L >Is there any user account setting/run option to set the max memory use by a	 >program?o  K If you use more memory than is available on a VMS system, it tends to hang.-  E However, if the user's "paging file quota" is less than the amount ofSK virtual memory on the machine, a single process cannot exhaust memory.  Yous* will get a "process quota exceeded" error.  + Use AUTHORIZE to set the paging file quota.8  + --Keith Lewis              klewis$mitre.orgC> The above may not (yet) represent the opinions of my employer.   ------------------------------    Date: 25 Apr 2002 10:53:09 -0700( From: bob@instantwhip.com (Bob Ceculski): Subject: Re: will memory leak program caused system crash?= Message-ID: <d7791aa1.0204250953.4c963e35@posting.google.com>c  d wingwong@witty.com (wing) wrote in message news:<873e96d6.0204250553.15fcf454@posting.google.com>... > Hi,h > < > Will a memroy leak program cause system crash in openvms?    NO!   c > V > Is there any user account setting/run option to set the max memory use by a program?  9 Yes, but only if I give you the priviledges to set it ...i   >  > Wing   ------------------------------   End of INFO-VAX 2002.228 ************************