0 INFO-VAX	Sat, 12 Jan 2002	Volume 2002 : Issue 21      Contents:A Re: 17 years and still waiting for my first VMS crash (yawn!) ... A RE: 17 years and still waiting for my first VMS crash (yawn!) ... A Re: 17 years and still waiting for my first VMS crash (yawn!) ... A RE: 17 years and still waiting for my first VMS crash (yawn!) ... A Re: 17 years and still waiting for my first VMS crash (yawn!) ... A Re: 17 years and still waiting for my first VMS crash (yawn!) ... : Re: 2 new Compaq Accredited Professional exams for OpenVMSP 5744           Would you like to lose weight while you sleep?                  2) Re: Accessing tapes in DLT4500 tape drive  Re: Amazing, and a good laugh C Re: Bugs and Optimization (was: Re: VMS on more common hardware...)   Re: Can SMTP relay be non-evil ?" Re: COMPAQ - Spin off VMS Business" Re: COMPAQ - Spin off VMS Business" Re: COMPAQ - Spin off VMS Business" Re: COMPAQ - Spin off VMS Business" Re: COMPAQ - Spin off VMS Business2 Re: Compaq still tries to spin Alphacide both ways& Re: CPQ Board Member Missing in Action* Re: CSWS: Cancel or Expire Authentication? CSWS: POST Problem: Re: DCL day of the minute: Inconsistency in date handling?, Re: follow up on ATI Radeon 7500 for OpenVMS Re: FTP Success/Failure? Re: FTP Success/Failure? Re: FTP Success/Failure? Re: FTP Success/Failure? RE: FTP Success/Failure? Re: FTP Success/Failure?  Re: Getting DEC ADA 3.5A to work1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 1 Re: historical evidence of what went wrong at DEC 5 Re: How many Fibre Channel - Multiple Hosts/4 Fabrics 0 Re: HP admits it will kill VMS if merger suceeds0 RE: HP admits it will kill VMS if merger suceeds0 Re: HP admits it will kill VMS if merger suceeds. RE: Huge EV7 Alpha Galaxy Cluster being built!. Re: Huge EV7 Alpha Galaxy Cluster being built!. Re: Huge EV7 Alpha Galaxy Cluster being built!& Re: Inquirer take on Ayr Alpha monster& Re: Inquirer take on Ayr Alpha monster Invitation to Compaq webcast( Re: OpenVMS Disk Services for Windows NT Re: Re[2]: DCL whish list ; Re: Steve Jobs not affraid to sell his "different" products N Re: Sv: Sv: Sv: Younger recruits versus experienced veterans ( was Re: The dem Re: Unix filesystems on VMS  Re: Unix filesystems on VMS  Re: Unix filesystems on VMS  Re: VMS ECOs? Re: VMS on more common hardware (was Re: VMS as a career move ) ? Re: VMS on more common hardware (was Re: VMS as a career move ) ? Re: VMS on more common hardware (was Re: VMS as a career move ) ? Re: VMS on more common hardware (was Re: VMS as a career move )  Re: VMSBackup to unix  Re: VMSBackup to unix # Re: what is future of VMS and alpha # Re: what is future of VMS and alpha # Re: what is future of VMS and alpha " Re: who else got the VMS brochure?" Re: who else got the VMS brochure?" Re: who else got the VMS brochure? Re: xml and OpenVMS  Re: xml and OpenVMS P Re: Younger recruits versus experienced veterans  ( was Re: The demise of compaqF Re: Younger recruits versus experienced veterans ( was Re:     The demF Re: Younger recruits versus experienced veterans ( was Re:     The demF Re: Younger recruits versus experienced veterans ( was Re:     The demF Re: Younger recruits versus experienced veterans ( was Re:     The dem< RE: Younger recruits versus experienced veterans ( was Re: T< RE: Younger recruits versus experienced veterans ( was Re: TB Re: Younger recruits versus experienced veterans ( was Re: The demB Re: Younger recruits versus experienced veterans ( was Re: The demP Re: Younger recruits versus experienced veterans ( was Re: Thedemise of compaq )P Re: Younger recruits versus experienced veterans ( was Re: Thedemise of compaq )P Re: Younger recruits versus experienced veterans ( was Re: Thedemise of compaq )  F ----------------------------------------------------------------------  % Date: Fri, 11 Jan 2002 19:00:53 -0000 ' From: "Lanzecki" <lanzecki@heptite.org> J Subject: Re: 17 years and still waiting for my first VMS crash (yawn!) .... Message-ID: <a1ncp6$5in$1@news7.svr.pol.co.uk>  > "John McLean" <mcleanj@swissonline.delete.ch> wrote in message/ news:3C3F257D.5810331A@swissonline.delete.ch...  >  <snip> >  > G > I heard of a nice system crash in Melbourne, Australia one day.  They H > were using a crane to get one of the 8000 series machines in through a > window on the second floor.  > G > Apparently things were shaken but not stirred (too much) - just a few J > parts needed.  The customer got a new machine but I think this one lived= > a useful operational life, albeit with a few dented panels.  >  > 
 > John McLean   D Isn't that one mentioned in the DEC History pdf on the OVMS website?  & It's definately not apocryphal anyway!   Lanzecki   ------------------------------  % Date: Fri, 11 Jan 2002 19:21:17 -0000 8 From: John Macallister <J.Macallister1@physics.ox.ac.uk>J Subject: RE: 17 years and still waiting for my first VMS crash (yawn!) ...N Message-ID: <35666012DF4CD411BE940090279FA240010BF1A9@ppnt41.physics.ox.ac.uk>  I If you never change anything, use anything or upgrade anything just about J the only thing which will cause a crash is a hardware failure. In the realL world VMS systems crash just as frequently as other comparable systems. I'veI had crashes with DECwindows, UCX, DECnet/OSI,etc. Some of our VMS systems K have never crashed while some of them have had periods when they've crashed D every other day. Some of our Windows PCs have also never crashed ...   John   ------------------------------  # Date: Fri, 11 Jan 2002 19:29:56 GMT " From: Alfred Falk <falk@arc.ab.ca>J Subject: Re: 17 years and still waiting for my first VMS crash (yawn!) ...9 Message-ID: <Xns91937F24DD696falkarcabca@205.233.108.180>   , bob@instantwhip.com (Bob Ceculski) wrote in 5 news:d7791aa1.0201101747.69ba9331@posting.google.com:   E > It been almost 17 years now I have been on VMS without an OS crash! H > Started out on a PDP 11/83 running RSTS/E briefly without a crash alsoE > then to Vax VMS 4.7 then to Vax VMS 6.0 then to Vax VMS 6.2 then to E > Alpha VMS 7.1 ... I dare anyone in windoze or unix or linux land to E > claim the same!  I think this once again proves what OS is the most G > reliable ... read this and weep windoze and unix users, oh, I forgot, $ > you already are weeping ... sorry!  F Wow! some people have amazing records!  While VMS is a lot more crash-G resistant than some other OS's, I'm quite sure I've had crashes (or at  5 least hangs) with every version from 4.7 through 7.3.   G Just last night I was re-booting an Alphaserver running OVMS 7.3.  The  G other nodes (another Alpha at 7.3 and a VAX at 7.2) decided to join it  @ in re-booting too.  (LOCKMGRERR, Error detected by Lock Manager)   --  @ ----------------------------------------------------------------A   A L B E R T A         Alfred Falk               falk@arc.ab.ca  @ R E S E A R C H         Information Systems Dept   (780)450-5185+   C O U N C I L         250 Karl Clark Road 1                         Edmonton, Alberta, Canada  http://www.arc.ab.ca/   T6N 1E4   http://www.arc.ab.ca/staff/falk/   ------------------------------  % Date: Fri, 11 Jan 2002 14:24:49 -0500  From: William_Bochnik@acml.comJ Subject: RE: 17 years and still waiting for my first VMS crash (yawn!) ...> Message-ID: <OFE5CE6BF9.9ABADA0D-ON85256B3E.006A7159@acml.com>  6 You said "comperable systems" then mentioned windows - hahahahahahaha. T Anyway, in my 8 years of VMS experience, and 11 of Windows and / or NT, VMS was less> prone to crashed - changes or not.  All MS os's - not so much.      8 >     John Macallister <J.Macallister1@physics.ox.ac.uk>J >If you never change anything, use anything or upgrade anything just aboutK >the only thing which will cause a crash is a hardware failure. In the real M >world VMS systems crash just as frequently as other comparable systems. I've J >had crashes with DECwindows, UCX, DECnet/OSI,etc. Some of our VMS systemsL >have never crashed while some of them have had periods when they've crashedE >every other day. Some of our Windows PCs have also never crashed ...  >  >John       F ______________________________________________________________________    The information contained in this transmission may contain privileged and confidential information and is intended only for the use of the person(s) name above.  If you are not the intended recipient, or an employee or agent responsible for delivering this{ message to the intended recipient, any review, dissemination, distribution or duplication of this communication is strictly  prohibited.  If you are not the intended recipient, please contact the sender immediately by reply e-mail and destroy all copies of the original message.   ------------------------------  % Date: Fri, 11 Jan 2002 18:05:18 -0500 1 From: Michael Austin <maustin@firstdbasource.com> J Subject: Re: 17 years and still waiting for my first VMS crash (yawn!) ...2 Message-ID: <3C3F6FAE.3B13B0D8@firstdbasource.com>   John McLean wrote: <snip> > G > I heard of a nice system crash in Melbourne, Australia one day.  They H > were using a crane to get one of the 8000 series machines in through a > window on the second floor.  > G > Apparently things were shaken but not stirred (too much) - just a few J > parts needed.  The customer got a new machine but I think this one lived= > a useful operational life, albeit with a few dented panels.  > 
 > John McLean    H There were a number of stories like this when the big earthquake hit San1 Francisco a few years ago (late 80's early 90's).   G " microvaxes that fell through the floor only needed to have the boards H re-seated and booted the first try".  "Disk cabinets bouncing 4-5 inches" off the floor and kept running"...   --   Regards,   Michael Austin7 First DBA Source, Inc. -- http://www.firstdbasource.com  President/Sr. DBA Consultant   ------------------------------  % Date: Sat, 12 Jan 2002 00:13:48 +0100 1 From: John McLean <mcleanj@swissonline.delete.ch> J Subject: Re: 17 years and still waiting for my first VMS crash (yawn!) ...5 Message-ID: <3C3F71AC.8582A254@swissonline.delete.ch>    Michael Austin wrote:  >  > John McLean wrote: > <snip> > > I > > I heard of a nice system crash in Melbourne, Australia one day.  They J > > were using a crane to get one of the 8000 series machines in through a > > window on the second floor.  > > I > > Apparently things were shaken but not stirred (too much) - just a few L > > parts needed.  The customer got a new machine but I think this one lived? > > a useful operational life, albeit with a few dented panels.  > >  > > John McLean  > J > There were a number of stories like this when the big earthquake hit San3 > Francisco a few years ago (late 80's early 90's).  > I > " microvaxes that fell through the floor only needed to have the boards J > re-seated and booted the first try".  "Disk cabinets bouncing 4-5 inches$ > off the floor and kept running"... >   E I recall hearing at DECUS 1988 or 89 in Annaheim about something like  this.   B The system manager said that the Vax actually fell over during theH earthquake.  About a minute or two after it happened the phone rang.  It4 was a user.  The conversation went something like .. "Is the Vax down ?"  "Um, yes..."% "Will it be up again in 10 minutes ?"      John   ------------------------------  % Date: Fri, 11 Jan 2002 17:19:13 -0500 2 From: "Sue Skonetski" <susan.skonetski@compaq.com>C Subject: Re: 2 new Compaq Accredited Professional exams for OpenVMS 2 Message-ID: <lwJ%7.593$5Y4.15922@news.cpqcorp.net>  H North America folks who maintain the web pages have yet to post the betaJ exams on their beta page. The 436 and 450 beta exams are live at Prometric9 Test Centers as indicated on the Testing Information page L (http://www.compaq.com/training/2023.html), which also provides links to theH Exam Preparation Guides. See bottom of table on Testing Information page  = "Sue Skonetski" <susan.skonetski@compaq.com> wrote in message , news:7en_7.369$5Y4.10466@news.cpqcorp.net... > Newsgroup, > C > I just received this, you will have to go to the web site for the G > attachments that are mentioned in this email.  I have no idea why its  double: > spaced again.  This will be on web sites later this week >  > Sue  > @ > ______________________________________________________________ > ( > To: OpenVMS Professionals and Managers > 9 > From: Compaq Accredited Professional Program - HPS Team  > 1 > Re: Recruitment of OpenVMS Beta Exam Candidates  >  > Important: > K > Please act on this request immediately to ensure April, 2002 availability  of
 > live exams! J > Please distribute this announcement to OpenVMS professionals everywhere. >  > Greetings, > K > The Compaq High Performance Server group (HPS) has launched betas for two F > new Compaq Accredited Professional exams that support the OpenVMS V7 Systems I > Engineer and AlphaServer/OpenVMS V7 ASE certification designations. The  > exams are: > H > OpenVMS V7 Advanced Administration, Performance, and Support (011-450) > - > OpenVMS V7 Network Administration (011-436)  > L > The beta exams are available, now, at Prometric Testing Centers worldwide.H > The final "live" exams will be available at CAT Global Test Centers in EMEA, > > as well as Prometric, following completion of the beta test. > 5 > Take a free beta exam. win an iPAQ color pocket PC!  > K > All individuals who complete (pass or fail) one or both of the above beta L > exams before February 15, 2002 will qualify to win an iPAQ color pocket PCG > (model H3650). At the close of the beta period for each exam, the HPS L > certification program manager will randomly draw one name from each of theL > two beta test groups and award the lucky participants an iPAQ (see lotteryF > terms and conditions below). So take the test(s) for free and take a chanceL > at winning an iPAQ. If you do not pass, then you will receive a voucher toH > take the live exam for free. EMEA candidates will receive a CAT Global< > voucher. All other candidates receive a Prometric voucher. >  > Time frame > I > The beta exams are available now and will run to approximately February  15, J > 2002. The target end date, which may change based on enrollment figures,K > marks the end of the data collection period for the beta exam. Candidates I > may continue to take the exam beyond the end date, since beta exams are  not J > retired until the live exams are available. We're expecting the exams to go > live in April 2002.  > @ > Register now for one or both of the betas, or identify severalK > colleagues/employees in your organization who are willing to participate. K > Please encourage candidates to take the exam before February 15. The only L > individuals who will qualify for the iPAQ lotteries are those who take the) > exam(s) on or before February 15, 2002.c >a! > Who should take the beta exams?- >-H > The beta exams are intended for OpenVMS Systems Administrators, SystemI > Engineers, Service Engineers, and Pre-Sales Technical Support EngineerstG > within Compaq, Compaq partner, and Compaq customer organizations. The I > expertise and experience levels of beta candidates may vary, from those6 who @ > have advanced OpenVMS system skills to those who have basic orF > intermediate-level skills. See the Exam Preparation Guides (EPGs) atK > http://www.compaq.com/certification/na/betaexam.html for a description of.  > the competencies being tested. >OJ > OpenVMS professionals within Compaq, Compaq partner, and Compaq customerI > organizations should be recruited, as well as OpenVMS instructors. This1 mixWG > of candidates is critical to evaluate the exam questions and define a  > passing score. >.L > Because a range of expertise is required for the beta, it is expected thatK > some candidates will not pass. However, candidates who take the beta willnK > become familiar with the areas where they are skilled and the areas wheres > they need development. > % > How to prepare for the beta exam(s)d >aJ > Before you register and take an exam, review the content of the exam and therL > applicable training or documentation. This information is available in theH > Exam Preparation Guide for each exam. The beta exams are "closed book", > exams. No reference materials are allowed. >_ > Exam Preparation GuidesT >iK > The beta Exam Preparation Guides (EPGs) are attached and are also locatedt atI > http://www.compaq.com/certification/na/betaexam.html. It's important to ' > review the EPG before taking an exam.  >e > What is a beta exam? >eL > A beta exam allows us to test the integrity of questions before we develop anK > final exam. After candidates take a beta exam, we analyze the results and B > the candidates' comments to select the best questions, construct
 equivalent3 > forms of the exam, and establish a passing score.  >s% > Benefits to candidates and managers  >a# > There is no cost for a beta exam!  >mL > Candidates who do not pass a beta exam have a better chance of passing theJ > live exam, since they see all the questions in the exam pool. During theB > live exam, candidates see only selected questions from the pool. >nJ > Candidates become familiar with the areas where they are skilled and the > areas where they need work.s >cJ > A candidate who does not pass a beta exam can retake the live exam at no > additional cost. >rG > The passing score for an exam is defined several weeks after the betahJ > concludes. Candidates who pass a beta exam are recorded as having passed thenD > live exam and do not have to take it. (The test center may provideI > candidates with a record of their beta results, which the certification0= > program office may require, so retain any record provided.)l >e > Time commitmenta > L > The exam (and administration time) can take up to 4 hours depending on theK > number of questions being tested. Travel time to and from the test centerB is
 > additional.t >/
 > Registeringu >.L > To register for the exam, call the Prometric Testing Center nearest to youL > and request the Compaq 450 and/or 436 exam (see titles above), or register' > online at http://2test.com/index.jsp.a >nH > Please register as soon as possible. Since beta exams tend to be long,H > Prometric may have difficulty scheduling blocks of time for individualB > candidates. You will need the following information to register: >a( > Compaq badge number (Compaq employees) >eK > Social Security Number (If you do not have a Social Security Number, thenr anG > ID number will be assigned. (Note: You must use this same number when $ > registering for all future exams.) >o > Company Name > 
 > Tutorial >eI > Before you begin an exam, you can take a tutorial at the test center to L > familiarize yourself with the details of taking online Prometric exams. IfI > you have not taken an online Prometric exam, you are encouraged to take  thisL > tutorial. It will help you take the exam more efficiently. Time for takingJ > the tutorial (approximately 10 to 15 minutes) is not counted against the! > time you have to take the exam.  > ) > Comments on the exam / Candidate surveyn >eH > After you complete the exam, you can make comments on questions: theirK > accuracy, appropriateness to audience, etc. You also see survey questionspJ > before the exam. The survey is designed to help the exam team define theH > final exam forms for the target audiences, so please answer the surveyI > questions accurately. Taking the survey does not count against the time1 you1 > have to take the exam. >i, > iPAQ color PC lottery terms and conditions >w> > The following terms and conditions have been established for administration > of the iPAQ lotteries. >t5 > To qualify for the iPAQ lottery, beta participants:p > K > Must have gained experience selling, supporting, or using OpenVMS systemso asG > a system administrator, systems engineer, service engineer, pre-saleswK > technical support engineer, instructor, student, or other OpenVMS-relateda > role.i >.B > Must possess at least basic OpenVMS system skills and knowledge. >sI > Must complete the 450 and/or 436 beta exam(s) on or before February 15,p 2002) > at an authorized Prometric Test Center.  >sA > Must attempt to answer all questions presented during the exam.a >mG > Must complete the candidate survey that is presented before the exam.  >iK > Must provide appropriate contact information during the exam registration_
 > process. >_I > Candidates can take one or both beta exams, but cannot take either exam  more# > than once within the beta period.o >mD > Candidates who take both betas are eligible to participate in both+ > lotteries, but cannot win more than once.q >tD > Prizes are iPAQ color pocket PCs (model H3650). If unavailable, anJ > equivalent value prize will be substituted. No cash substitution will be
 > offered. >,H > If at least 50 participants have not taken a beta exam by February 15, 2002, L > the lottery deadline for the exam may be extended at the discretion of theK > program manager. The two lottery winners will be notified by phone and/orh byK > mail within 5 business days following each lottery, which will occur whenoE > Prometric makes the beta participant names available (probably late K > February). Winners have 10 business days from the date of notification to K > claim their prize, otherwise they will forfeit their prize and a new name L > will be selected. There is no implied obligation to notify lottery losers. >eL > Compaq is not responsible for beta participant names omitted by Prometric,I > or related Prometric errors/omissions which may eliminate a participant  from > a lottery. >nF > Compaq is not responsible if Prometric cannot schedule test time forD > individual candidates within the beta period, so candidates should register > early. >eA > Winners are responsible for any applicable country or local tax  obligations. >eH > These terms and conditions may be amended or elaborated by the programC > manager to comply with additional corporate requirements or locale > regulations. > J > Please distribute this announcement to OpenVMS professionals everywhere. >e> > Thank you in advance for your cooperation and participation. >a3 > Compaq Accredited Professional Program - HPS Teamu >o >p >i   ------------------------------   Date: Thu, 10 Jan 2002 22:48:22 ) From: 4822718travelincentives2002@aol.comnY Subject: 5744           Would you like to lose weight while you sleep?                  2e+ Message-ID: <200201102254232.SM00952@plain>   6  As seen on NBC, CBS, CNN, and even Oprah! The health : discovery that actually reverses aging while burning fat, < without dieting or exercise! This proven discovery has even 9 been reported on by the New England Journal of Medicine. n: Forget  aging and dieting forever! And it's Guaranteed!      Click here:e http://ultimatehgh.81832.com  . Would you like to lose weight while you sleep! No dieting!l No hunger pains! No Cravings! No strenuous exercise! Change your life forever!    100% GUARANTEED!  + 1.Body Fat Loss            82% improvement. ( 2.Wrinkle Reduction     61% improvement.- 3.Energy Level               84% improvement.!* 4.Muscle Strength         88% improvement.+ 5.Sexual Potency           75% improvement.a* 6.Emotional Stability     67% improvement.0 7.Memory                        62% improvement.  ; ***********************************************************t  . Click here to see another weight loss product: http://weighout.81832.comv  , You are receiving this email as a subscriber$ to the Opt-In America Mailing List. . To remove yourself from all related maillists, just click here:/ mailto:pac2server@btamail.net.cn?Subject=REMOVE    ------------------------------  # Date: Sat, 12 Jan 2002 03:30:28 GMTu1 From: "David J. Dachtera" <djesys.nospam@fsi.net>p2 Subject: Re: Accessing tapes in DLT4500 tape drive' Message-ID: <3C3FAEAE.C828EB02@fsi.net>    Christoph Gartmann wrote:o > f > In article <3c3e5b7d.371578081@news.wlg.netlink.net.nz>, martin.hunt@inl.co.nz (Martin Hunt) writes:G > >I am currently playing around with a Quantum DLT4500 tape drive with G > >loader. If I dismount a tape, the next tape in the loader is loaded.rH > >But, is there any way of loading a specific tape, such as the tape inB > >slot 4? Or, as I suspect, would I have to purchase a product to > >support this feature?4 > >It would be nice to be able to do something like:! > >$ MOUNT/FOREIGN MKA500 /SLOT=4w > O > I wrote a program to control a loader. It works fine with a Spectralogic tapeeP > library, an overland DLT library and a Pioneer CD-ROM changer. If there is any- > interest in it I'll put it up for download.   ( Yes! Please post it! (one vote in favor)   --   David J. Dachteral dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/a   ------------------------------  # Date: Sat, 12 Jan 2002 03:23:57 GMT 1 From: "David J. Dachtera" <djesys.nospam@fsi.net> & Subject: Re: Amazing, and a good laugh' Message-ID: <3C3FAD27.A5C977AE@fsi.net>a   Bill Todd wrote: > L > Truly impressive - makes any silliness that has gone on around here (in myM > memory, anyway) look tame, and all our participants look positively erudite 
 > and urbane.e  G I spent the bulk of last evening reading this (after I finished workingaC on our new kitchen) - I was so taken with disbelief I couldn't stoprE reading! That bad news is that this Bernard character is a Chicagoan.a9 I'm a suburbanite, but he still gives us all a black eye.w   UFB...   --   David J. Dachtera  dba DJE Systemsg http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/o   ------------------------------  # Date: Fri, 11 Jan 2002 21:37:51 GMTr2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)L Subject: Re: Bugs and Optimization (was: Re: VMS on more common hardware...)2 Message-ID: <PWI%7.591$5Y4.15525@news.cpqcorp.net>  a JF Mezei <jfmezei.spamnot@videotron.ca> wrote in message news:<3C3A146D.9BF412F9@videotron.ca>...SG : When writing kernel code or deep down stuff for the OS, does using a  H : compiler with lots of optimization for the harware ever cause serious  : problems ?    J   Ever?  Sure.  But that is not particularly meaningful answer, of course.  K   I've debugged bugs that were latent, and that that were uncovered by the  L   optimizer -- in entirely user-mode code.  The failures do tend to be more M   dramatic within kernel-mode code of course, but the potential for problems hH   is not specific to the operating system nor to privileged-mode code.  K   (Most any optimizer tends to be a useful tool for uncovering latent code sH   bugs, after all.  More than a few programmers have blamed the compilerG   or the optimizer for these latent bugs, too -- the first time I codedhJ   one of these bugs, I blamed the optimizer.  This is not to say that the B   optimizer can't ever introduce bugs, I've seen that happen too.)  H : (I am thinking of instruction ordering, and variable "elimination" by > : keeping them in registers, and possible many other factors).  L   The Alpha real-time or shared-memory programmer must be familiar with the J   sequence points and with the trap and memory constructs barriers -- not K   only the compiler, but the microprocessor can also re-order and can also xK   cache stuff.  This is nothing particularly new -- various VAX boxes have u   similar considerations.s  J   For an introduction into some of the usual sorts of programming mistakesM   and considerations, please review topics (1661), (2681), (4647) and (6099) dI   over in the Ask The Wizard area.  These are an introduction, of course.   I   There are language keywords around that will keep a variable in memory rH   (and not in a register; eg: "volatile") -- volatile-like keywords are I   particularly useful for I/O space references.  (You must be careful to nG   have the correct widget marked as volatile, too.  You probably don't eF   want to end up with a memory-based pointer to an object stored in a F   register, of course.  :-)  Volatile doesn't help with memory caching,   and memory access resequencing, however...  I   There is a list of the hardware instructions that are available in I/O -F   space listed in the documentation set -- related information in the I   manuals can provide details, as can the usual discussions around sharedVF   memory and synchronization techniques.  The circa-1982 VAX hardware H   manual had such a list, and the second edition of the VAX ArchitectureK   Manual (1987, 1991) has I/O space reference details on pages 321 and 322.c  K   As a start, please read and heed the OpenVMS Programming Concepts manual.   N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------   Date: 11 Jan 2002 21:38 CST ' From: carl@gerg.tamu.edu (Carl Perkins)l) Subject: Re: Can SMTP relay be non-evil ?-- Message-ID: <11JAN200221385670@gerg.tamu.edu>0  1 JF Mezei <jfmezei.spamnot@videotron.ca> writes...oA }I have often heard of SMTP relay being evil, and understand why.c } N }However, is it possible with TCPIP services (5.0) to enable the relaying only }for local hosts ? } N }For instance a mac on the same lan as the vax could send messages to the vax,J }and if they don't belong on the vax, the vax would then relay them to the6 }ISP's SMTP server to be delivered onto the internet ? } ) }Or is relaying an all or nothing thing ?n  G I do not use TCPIP Services, but I would have to say "of course this is I possible". But I can't prove it - a quick look at the on-line help insidesG of the TCPIP command (V5.0A - ECO 2) revaled nothing but the apparently-G all or nothing "TCPIP> SET CONFIGURATION/OPTIONS=" with he alternativescK of "RELAY" and "NORELAY". If there is any way to do it, it is not mentioned.G any place I could find locally. I didn't check for documentation at then$ usual place: www.openvms.compaq.com.  C It is, however, exactly how the an SMTP server is supposed to work.tC It is fairly uncommon for the PCs and Macs etc. to directly connect A to the final destination's SMTP server - in fact the usual e-mailtF software on PCs, at least, has a place to specify a single SMTP serverC through which all the outgoing mail is sent, and I don't recall anyiB setting to do anything else (in Outlook, Express or otherwise, andF Eudora anyway) unless, perhaps, not specifying the SMTP server addressE makes it try to directly connect to each message's destination (whiche seems a bit unlikely).  H In Multinet, for example, this is controlled via the SMTP_SERVER_REJECT.G file which is checked on a "do what the first match says" basis, so theuF you can put in all "allow relaying from X" type records (which usuallyH specify the entire local subnet or net) and follow them with a "deny allJ relaying" record which catches anything that has not already been allowed.  I Using PMDF there is a mechanism (in the MAPPINGS. file) to indicate whichiH addresses are internal and via the channel selection mechanism these areC then processed via a channel that allows relaying whereas the othert channels do not allow relaying.t  K Somewhere I would hope that there is a method to do this in TCPIP Services.eH If not, the thing is useless for SMTP as anything other than a client to a real SMTP server.e   --- Carl   ------------------------------  + Date: Fri, 11 Jan 2002 19:15:22 +0000 (UTC)r From: david20@alpha2.mdx.ac.uk+ Subject: Re: COMPAQ - Spin off VMS Businessn+ Message-ID: <a1ndka$97a$1@aquila.mdx.ac.uk>   U In article <3C3F2BBC.75F08991@pacbell.net>, Don Sykes <anonymous@pacbell.net> writes:W > $ >I restate part of a previous post :D >VMS could still be saved, but it would take some group with $$$ and >vision and even someuG >acceptance from the VMS community that the OS needs to evolve and thatH4 >might be more difficult than getting the investors. >l  F VMS has always been evolving. In which particular areas do you want it6 to evolve further and in which particular directions ?  
 David Webb VMS and Unix team leader CCSS Middlesex University   >Rick Nickles wrote: >> cH >> It really seems that the best thing for VMS is to spin it off as it'sG >> own separate business (company).  That way there isn't the politicalxI >> agenda with Microsoft, and then they are free to market VMS like crazyiI >> and really build up a following as a single company.  Compaq stands toeH >> gain, The users stand to gain, Customers stand to gain.  The only oneG >> that might lose is Microsoft, and it isn't like they'll be hurt thate >> bad anyway. >> a >> What do you think folks?r   ------------------------------  % Date: Fri, 11 Jan 2002 15:28:54 -0500 % From: JF Mezei <jfmezei@videotron.ca> + Subject: Re: COMPAQ - Spin off VMS Businesss, Message-ID: <3C3F4B04.FD56B966@videotron.ca>   "Terry C. Shannon" wrote:sD > boxes.) In most cases the place for Windows and the place for trueJ > enterprise OSes are different places indeed and Compaq should be able to, > segment the market and target accordingly.  J I take issue with the "segment the market". That is exactly what they haveL done. And as Windows gains in enterpriseness, they reduce the markets of the other to make way for Windows.  I The issue here is that Bill Gates has no problems targetting Unix VMS andhE eventually NSK. (heck, guess what runs the NASDAQ web servers ?). WhylM shouldn't Compaq be allowed to fight back just as agressively as Microsoft isb fighting to gain market share ?   L I say let the VMS folks loose and let them attack whatever market they want.L IF VMS engineers were to go to some heavy duty rendering outfit and convinceH them to port their unix software to VMS, shouldn't the VMS group be thenM allowed to push/price VMS so that its workstations would compete against MACsiN and SGI in the graphics markets ? And now that they would have such pricing atN the business workstation market, why not get Star Office on VMS in a supportedM fashion and then push it ? And as you go along, you develop more and more foreI workstations and before you know it, you have an interesting platform foraB office productivity that is not infectable with microsoft viruses.  M If Compaq continues to segment the market and decide that workstations belongiJ to NT, then VMS doesn't stand a chance. It must have the freedom to exlore opportunities as they arise.  L Look at Linux. Why did it take off in business ? Because it presented itselfK as a viable alternative to Microsoft especially in light of the possibility N that MS might implode due to anti trust. Had VMS been allowed to be agressive,L it could have made serious inroads by presenting itself as a SOLID AND FULLYR DOCUMENTED AND FULLY SUPPORTED alternative to Microsoft, nipping Linux in the bud.  J VMS must have the freedom to explore opportunities as they arise and be asP agressive as it wants. Microsoft will lose its monopoly within the next 5 years.   ------------------------------  # Date: Sat, 12 Jan 2002 01:59:48 GMTo' From: Don Sykes <anonymous@pacbell.net>n+ Subject: Re: COMPAQ - Spin off VMS Businessk+ Message-ID: <3C3F9941.607A6DB7@pacbell.net>n  H First, no one using Unix or NT is going to consider VMS a better OS with@ case INsensitive filenames. I know there are those cov folks whoH disagree, but I think it would be a BIG plus if at least filenames couldF convert one-to-one. Also, Unix style symbolic links must be supported.F Hard links a-la "set file/entry=xxx.dat yyy.dat" are not sufficient. IE could live without these things, but I know non vms people are alwaysiE touting these deficiencies as examples of how old and out of date vmsm is. F In fact if I had my way I'd create an OS that was a super set of VMS &F Unix, where DCL & csh & ksh would all work just fine on the underlying system.v   david20@alpha2.mdx.ac.uk wrote:t > W > In article <3C3F2BBC.75F08991@pacbell.net>, Don Sykes <anonymous@pacbell.net> writes:e > >r& > >I restate part of a previous post :F > >VMS could still be saved, but it would take some group with $$$ and > >vision and even someMI > >acceptance from the VMS community that the OS needs to evolve and that 6 > >might be more difficult than getting the investors. > >P > H > VMS has always been evolving. In which particular areas do you want it8 > to evolve further and in which particular directions ? >  > David Webb > VMS and Unix team leader > CCSS > Middlesex University >  > >Rick Nickles wrote: > >>J > >> It really seems that the best thing for VMS is to spin it off as it'sI > >> own separate business (company).  That way there isn't the politicalaK > >> agenda with Microsoft, and then they are free to market VMS like crazy,K > >> and really build up a following as a single company.  Compaq stands toeJ > >> gain, The users stand to gain, Customers stand to gain.  The only oneI > >> that might lose is Microsoft, and it isn't like they'll be hurt that  > >> bad anyway. > >> > >> What do you think folks?h   -- r   Have VMS. Will Travel. Wire Paladin @alphase.com 
 San Franciscor   ------------------------------  # Date: Sat, 12 Jan 2002 03:59:45 GMT 1 From: "David J. Dachtera" <djesys.nospam@fsi.net>t+ Subject: Re: COMPAQ - Spin off VMS Businesse' Message-ID: <3C3FB58A.E6FEDB09@fsi.net>r   Don Sykes wrote: > J > First, no one using Unix or NT is going to consider VMS a better OS withB > case INsensitive filenames. I know there are those cov folks whoJ > disagree, but I think it would be a BIG plus if at least filenames could > convert one-to-one.   C I can see cases where it might be 100% necessary, especially if thetG middle-ware cannot be made case-blind, or if "sins" have been committed H ("myfile" and "MyFile" being two separate, but needed files for example)@ such that case-blindness would seriously "break" code that works otherwise. u  H I guess it also depends a lot on what you "grew up" with. If you grew upB with VMS, then UN*X's case sensitivity is a potential problem, and vice-versa.   4 > Also, Unix style symbolic links must be supported.F > Hard links a-la "set file/entry=xxx.dat yyy.dat" are not sufficient.  F Well, technically, "SET FILE/ENTER" produces the equivalent of a "softD link" - a new directory entry that point to the original file headerD ("inode", in UN*X-speak). No new file header ("inode") is allocated.   > I-G > could live without these things, but I know non vms people are alwaysiG > touting these deficiencies as examples of how old and out of date vmsi > is.l   The gripe is two-fold in that:  E 1. In ODS, there is no linked list connecting all of the "aliases" to,G the actual file itself. So, there's no way to find, for any given file,e+ all the aliases on a specific volume(-set).T  G 2. In UN*X-land, the "inode" - equivalent to a file header - does *NOT* H contain the filespec. Thus, if the directory entry is lost for the "realB file", it becomes very difficult to find the "beginning" of a fileE chain. That's really only an issue for soft-links, though. Hard-linksgG use an additional inode; so, at least a file can be called by an alias,!A even if the original directory or directory entry for it is lost.o  @ Reconciling that within the current limits of ODS poses rather a@ challenge unless there have been revisions in ODS to provide forF UN*X-like "linkage" of file chains to the their aliases (soft and hard@ links). SET FILE/ENTRY really only constitutes a loose form of aG soft-link. For a file ("chain") to have more than one "primary" header,eE well, I believe that this currently would be considered a corruption.   & I could very well be wrong, of course.  H > In fact if I had my way I'd create an OS that was a super set of VMS &H > Unix, where DCL & csh & ksh would all work just fine on the underlying	 > system.-   I can see that.    --   David J. DachteraM dba DJE SystemsJ http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------    Date: 11 Jan 2002 22:09:09 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen).+ Subject: Re: COMPAQ - Spin off VMS Businesse3 Message-ID: <$tlYXmlVZp9a@eisner.encompasserve.org>L  U In article <3C3F9941.607A6DB7@pacbell.net>, Don Sykes <anonymous@pacbell.net> writes:o > J > First, no one using Unix or NT is going to consider VMS a better OS with > case INsensitive filenames.w  E NT is case preserving, like /PARSE=EXTENDED , so anybody who is happyl? with NT in that regard should be happy with VMS in that regard.A  A As for those using Unix, never say never.  Nobody who _likes_ thei@ case-sensitive filenames on Unix would like /PARSE=EXTENDED, but> some who are _using_ Unix would prefer VMS.  (Granted, none of< them are the sort to be here in comp.os.vms defending Unix.)  & > I know there are those cov folks whoJ > disagree, but I think it would be a BIG plus if at least filenames couldH > convert one-to-one. Also, Unix style symbolic links must be supported.H > Hard links a-la "set file/entry=xxx.dat yyy.dat" are not sufficient. IG > could live without these things, but I know non vms people are always3G > touting these deficiencies as examples of how old and out of date vmsm > is.y  G Since Unix predated VMS, "out of date" is hardly a correct description.,H NT is newer than VMS, so it would appear that case preserving is the way; to be up-to-date.  VMS (on Alpha) got there before Unix :-)o  H > In fact if I had my way I'd create an OS that was a super set of VMS &H > Unix, where DCL & csh & ksh would all work just fine on the underlying	 > system.b  > Fred said that is where they are headed with the DII COE work,@ including the capability to choose a case-sensitive file system.   ------------------------------    Date: 11 Jan 2002 15:24:24 -0500. From: Joel Gallun <root@localhost.localdomain>; Subject: Re: Compaq still tries to spin Alphacide both wayss3 Message-ID: <ygapu4gd4av.fsf@severn.office.aol.com>n  ( peter@abbnm.com (Peter da Silva) writes:  I > But, yes, this is a useful technique. The TenDRA compiler that came outoN > of the UK defense industry used what they called ANDF, Architechture NeutralD > Distributiuon Format, as the first stage output from the compiler.  @ ANDF!?! Shouldn't this thread move to alt.folklore.computers? :)   -- Joel Gallun $ Compaq support team @ America Online http://www.tux.org/~joel   ------------------------------  # Date: Sat, 12 Jan 2002 03:21:46 GMTr4 From: "Terry C. Shannon" <terryshannon@mediaone.net>/ Subject: Re: CPQ Board Member Missing in Actionr> Message-ID: <eZN%7.67610$Sj1.27691417@typhoon.ne.mediaone.net>  ? "Terry C. Shannon" <terryshannon@mediaone.net> wrote in messagee8 news:r5E%7.66347$Sj1.27306368@typhoon.ne.mediaone.net... >t6 > "David Mathog" <mathog@caltech.edu> wrote in message' > news:3C3F0DA4.810A97D7@caltech.edu...  > > "Terry C. Shannon" wrote:r > >eD > > > Who is willing to fill the seat apparently vacated by MIA Lay. > > >h > > > -- > >h > > Walter Hewlett ;-) > H > That was a statement, not a question. Somehow I doubt that Mr. Hewlett wouldnL > be welcome on the Compaq board. He's alway in the HWP Directors' DogHouse. > ;-}t >,  L Perhaps prompted by the MIA Report in this forum, Compaq today 'fessed up toL the fact that His Kenness "stepped down" from the Compaq Board at the end ofF 2001. Guess he has other pressing matters he must pursue (Henry Waxman= apparently is in hot pursuit of Mr. Lay at this very moment).   K Lay also abdicated Board seats on two other corporate Boards in October ands	 December.n   ------------------------------  % Date: Fri, 11 Jan 2002 15:53:41 -0500o% From: JF Mezei <jfmezei@videotron.ca>13 Subject: Re: CSWS: Cancel or Expire Authentication? , Message-ID: <3C3F50D2.DBB5DDF3@videotron.ca>  * "j.lance wilkinson, (814) 865-1818" wrote:M > I'm looking for a way for a document or cgi script (perhaps NOT in the same"H > protected directory) to EXPIRE or CANCEL that authentication so that aH > subsequent attempt to access a script or document IN the /myprotected/L > directory will force a new browser-based login panel (perhaps responded to: > with a different userid/password pair?) to be presented.  H There is no way to do this. Once the client has the authentication for aK domain, it will always send it out in the HTTP request. And the server willlG authenticate each HTTP request individually since there is no "state ofn! connection" in the HTTP protocol.a  J One way around this is for the server to send an authentication request inE response to a HTTP request even though the browser had sent the rightnM username/password combo. But this is not failsafe since the user might simplycM cancel that request and go back to a previous URL, at which point the browserdL would send the "old" username/password, and your server would think that the1 user has completed the re-authentication request.   N And in the above case, the problem is triggering the server into including theM WWW-Authenticate: <realm> header i response to a HTTP request after a certainfK amount of inactivity. The server would have to be modified to keep track of-M the time of last transaction for each user/password combination, and wheneveruG a new transaction comes in, verify that the elapsed time sicne previous-- transaction was less than some defined value.z  N If you're really paranoid, I guess you could just pop up some page that uses aN plug-in that is known to crash the machine, at which point you are pretty sureK of having zapped the authentication from the browser :-) In other words, ifaK you force users to use Windows, then you don't have to worry about expiringnN their authentication since you know that Windows will do that for you whenever* it crashes every 15-20 minutes :-) ;-) :-)   ------------------------------  % Date: Fri, 11 Jan 2002 17:49:24 -0500I1 From: Michael Austin <maustin@firstdbasource.com>B Subject: CSWS: POST Problem 2 Message-ID: <3C3F6BF4.52DD839B@firstdbasource.com>  F I know this is OT for this group, but I am hoping that someone has run across this  issue.N  + I have a form generated by a CGI script at:d  / http://www.firstdbasource.com/cgi-bin/test_form   @ Which submits to a version of the TEST-GCI-DCL.COM procedure in  APACHE$ROOT:[CGI-BIN]   C When you Submit the form, it should give you all of the environmentT= variables including QUERY_STRING and display the values beingCH submitted.  If I use GET or PUT, these do show up, but not during a POSTF and the variables from the form are not parsed.  I cannot get these toD display in either Netscape 4.7 or IE5.x. Did I miss something in the> command procedure (typed out at the end of the result page).    H I have a C program that reads STDIN and does parse these out correctly.  Seeg  6 http://www.firstdbasource.com/cgi-bin/TEST_FORM_C_PROG  
 Any Ideas?   -- s   Regards,   Michael Austin7 First DBA Source, Inc. -- http://www.firstdbasource.comd President/Sr. DBA Consultant 704-947-1089 (Office)n 704-236-4377 (Mobile)a   ------------------------------    Date: 11 Jan 2002 15:29:57 -0800. From: SPAMSINK2001@YAHOO.COM (Alan E. Feldman)C Subject: Re: DCL day of the minute: Inconsistency in date handling?d= Message-ID: <343f30ae.0201111529.15bb5d52@posting.google.com>s   alphaman-nixspam@hsv.sungardtrust.com (Aaron Sakovich) wrote in message news:<8af17fe1.0112121211.43f06f3d@posting.google.com>...iC > (I'm posting this on 12 Dec, so let's use that as a basis for thea > following examples.) >  > Consider the following:h > E >  "Dir /Before=Yesterday" results in files dated 10 Dec and earlier.n > A >  "Dir /Before=Today" results in files dated 11 Dec and earlier.  >  SE >  "Dir /Since=Today" results in files created on the 12th.  Okay, so, > far so good.  But... > : >  "Submit /After=Today" starts a job immediately?  And... > E >  "Submit /After=Tomorrow" starts it as soon as it becomes the 13th,n > not after the 13th.8 > F > Shouldn't "After=Today" submit (or print) a job after ''Today' is noH > longer ''Today', but ''Tomorrow'?  Shouldn't "After=Tomorrow" submit a? > job after Tomorrow+23:59:59.99 instead of Tomorrow+00:00:00? tF > Basically, my opinion is "Today" is 12 Dec 2001, from 00:00:00 until> > 23:59:59.99; "Tomorrow" is 13 Dec 2001, from 00:00:00: untilE > 23:59:59.99.  Submitting something "after tomorrow" shouldn't startnE > the job as soon as tomorrow starts, but rather after tomorrow ends.d > H > Not that this behavior should be changed at this point in time -- thisE > would probably be more catastrophic than Y2K ever thought of being!t > G > Just making a point for discussion.  If someone can provide me with acG > better perspective for interpreting this behavior, I'd appreciate it,u) > because my view is certainly confusing!m >  > Aarone  E Actually, today I realized a better answer to your question. Think ofeD TODAY and TOMORROW as supplying *only* the date portion of the time.B Then consider that when you specify only a date, and specify it inC dd-mmm-yyyy format, the time defaults to 00:00:00.00. Now, when youmF specify TODAY or TOMORROW, you are still only specifying the date. TheE time then defaults to 00:00:00.00 because you specified only the daten% portion and everything is consistent.h  ? So, think of TODAY, TOMORROW, etc., as specifying only the datei portion of the date-time.o  C (Sorry about my previous answer not being up to snuff. I think this $ answer should be much more helpful.)   Disclaimer: JMHO Alan E. Feldman  afeldman!~~/\~~!gfigroup.com;n   ------------------------------  % Date: Fri, 11 Jan 2002 22:06:35 -0600., From: "Rich Jordan" <rjordan@mindspring.com>5 Subject: Re: follow up on ATI Radeon 7500 for OpenVMS,2 Message-ID: <a1ocng$32h$1@slb1.atl.mindspring.net>   Fred,oL      is this going to be EV6 only, like the Oxygen card?  Or can we hope forC a nice fast easier to find than a 4D20 card for PWS's and the like?a   Rich Jordanm  $ Fred Kleinsorge wrote in message ... >g > The ATI 7500K >is planned to be in a release in late fall.  We'll probably have somethings >to Beta test this summer. >    ------------------------------    Date: 11 Jan 2002 12:30:22 -0800( From: bob@instantwhip.com (Bob Ceculski)! Subject: Re: FTP Success/Failure? = Message-ID: <d7791aa1.0201111230.37267c3a@posting.google.com>f  k david_toolan@hotmail.com (Dvaid) wrote in message news:<de9f97c7.0201110625.6a5da7ba@posting.google.com>...k > Hi,s > E > We have a FTP job that transfers files from a VMS machine to an IBMuA > machine. The job can take up to 6 hours to complete. Recently a D > firewall was placed between the two machines. The rule base on theE > firewall is quite stringent regarding the types of packets/requestsoF > that are allowed through. Sometines the ftp process appeared to hangC > (no data going through the local port), which resulted in the jobM > having to be resent. > F > Currently the only way we have of knowing that the job has completed? > successfully is by multiplying the file block size by 512 andbF > comparing this with the number of bytes that have passed through theG > socket. We are not getting any response from the IBM machine. We have E > split the file into 6 files and are planning to send these files toi > the other machine. > 7 > We are running TCP/IP V4.2 - ECO 4 on OpenVMS V7.1-2.- > E > Is there any process/system logical that we can use to determine ifaD > the file transfer is currently taking place or if the transfer has > completed successfully.D  L If the IBM machine supports Decnet Phase IV, you could use Tcpwares Phase IVE over IP w/packet filters and use dcl copy and get a reliable copy ...lG or try another IP stack (Tcpware) ... otherwise there is no sure way tos  guarantee FTP file transfers ...   ------------------------------  % Date: Fri, 11 Jan 2002 16:00:12 -0500r% From: JF Mezei <jfmezei@videotron.ca>I! Subject: Re: FTP Success/Failure?b, Message-ID: <3C3F5259.6BD300BC@videotron.ca>  4 re: finding out if 6 hour FTP process worked or not.  G would COPY/FTP provide you with a $STATUS symbol that would be usable ?   K Have you considered using kermit as transfer protocol over TCPIP link ? itsg+ scripting does have "if failure"  handling.t   ------------------------------    Date: 11 Jan 2002 13:01:38 -0800) From: john@ossc.net (John Gemignani, Jr.)e! Subject: Re: FTP Success/Failure?c= Message-ID: <35b06b78.0201111301.3a106b4a@posting.google.com>   > Use the "hash" command.  It should print a '#' for each bufferE read/written across the network.  In the next release (V5.3) there ist. a ^T function which will display the progress.   -Johni    Y Nic Clews <sendspamhere@127.0.0.1> wrote in message news:<3C3EFCD7.FF1D100D@127.0.0.1>...t > Dvaid wrote: > > G > > We have a FTP job that transfers files from a VMS machine to an IBMcC > > machine. The job can take up to 6 hours to complete. Recently aVF > > firewall was placed between the two machines. The rule base on theG > > firewall is quite stringent regarding the types of packets/requestseH > > that are allowed through. Sometines the ftp process appeared to hangE > > (no data going through the local port), which resulted in the job  > > having to be resent. >   G > > Is there any process/system logical that we can use to determine iflF > > the file transfer is currently taking place or if the transfer has > > completed successfully.e >  > Been there, with problems. > H > TCP/IP is at its best, very simple. You do not get the safety net that? > (say) a DECnet copy gives you or Decnet/SNa. In effect TCP/IPj? > compromises what VMS can offer you with its native protocols.6 > J > Formerly on this list I described how I transfer files to my file serverC > from a VMS system. In summary, I generate a file with the current I > OpenVMS date and time, push it to the server, push the files, then pullIG > back the date/time file, and compare it with a copy of what I sent. IaA > then know I have a fair chance the link did not fail during therH > transfer. To verify the files, you need to pull back the copy you sentG > and do a DIFFERENCES. I had a case where a router was responsible for 4 > corrupting odd packets between a VAX and an AS400.C > Using PUSH-PULL it put the error back under VMS's control and the   > transfer could be reattempted. > G > However you have a 6 hour transfer time, so I don't think that methodoC > will suit. Try using the method of sending a date timefile at thetJ > beginning, then pull that back for comparison. Do a search on google andE > you should find my post. "My OpenVMS web procedures" from around 14  > September 01.t > I > Perhaps another think you might try is an LS when you've sent the file,eI > then examining the output for the filesize. It's a little while since I G > worked with mainframe FTP, it may give you date and size in a way youl% > can compare it to your source file.r > I > Don't believe the status codes sent by your FTP package, the remote endf > may give a false signal.   ------------------------------  % Date: Fri, 11 Jan 2002 22:08:41 +0100l9 From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com>"! Subject: Re: FTP Success/Failure?r' Message-ID: <3C3F5459.956A1EB9@aaa.com>s  > I'v a VMS server that sends aprox 1000 files/day to an IBM/MVS
 mainframe.F I check the FTP log for "250 Transfer completed", and if found, thinksF it all whent OK. Based on this thread, I'm not sure anymore if this isB a *safe* method. Could I get the "250..." message even if the file! wasn't "...Transfer completed" ??t  1 I don't *think* I'v lost a file in 2-3 years now.D   Jan-Erik Sderholm.e   ------------------------------  % Date: Fri, 11 Jan 2002 15:45:53 -0600a/ From: Terry Marosites <TMarosites@unitedad.com> ! Subject: RE: FTP Success/Failure? L Message-ID: <14987C0DA438094D91DAA6CED58D060A0703B7@dalntexch2.unitedad.com>   I use:   $  OPEN /WRITE OFIL FTP.TMP=20 $  WRITE OFIL "$ set noonl $  WRITE OFIL "$, $  WRITE OFIL "$ define sys$output FTP.TMP2") $  WRITE OFIL "$ UNAME =3D ""''uname'"" "rE $  WRITE OFIL "FTP SEAVAX /USERNAME=3D''UNAME' /PASSWORD=3D ''PASSW'" $ $  WRITE OFIL "SET TYPE ''THE_TYPE'" $  WRITE OFIL "VERBOSE" + $  WRITE OFIL "PUT ''WORK_FILE' ''TO_THAT'"J) $  WRITE OFIL "$  ftp_STAT=3D=3D $STATUS"i& $  WRITE OFIL "$  deassign sys$output"9 $  WRITE OFIL "$  write sys$output ""stat =3D ""+ftp_STAT-5 $  WRITE OFIL "$  if ftp_STAT.nes.""%X180180E9"" "=20@ $  WRITE OFIL "$    thenF $  WRITE OFIL "$     mail FTP.TMP /sub =3D ""ftp_err on ''WORK_FILE'""
 ""''uname'=20c $  WRITE OFIL "$  endift
 $  CLOSE OFIL6 $  @FTP.TMPl* $  write sys$output "stat =3D ''ftp_STAT'"  G I have found this very reliable over the years , as long as the "put" =t is the% last this before the "$ status" line.T     Terry Me   -----Original Message-----/ From: Jan-Erik S=F6derholm [mailto:aaa@aaa.com]s& Sent: Friday, January 11, 2002 1:09 PM To: Info-VAX@Mvb.Saic.Com,! Subject: Re: FTP Success/Failure?s    > I'v a VMS server that sends aprox 1000 files/day to an IBM/MVS
 mainframe.F I check the FTP log for "250 Transfer completed", and if found, thinksF it all whent OK. Based on this thread, I'm not sure anymore if this isB a *safe* method. Could I get the "250..." message even if the file! wasn't "...Transfer completed" ??   1 I don't *think* I'v lost a file in 2-3 years now.u   Jan-Erik S=F6derholm.l   ------------------------------  # Date: Sat, 12 Jan 2002 05:57:13 GMTt- From: "John E. Malmberg" <wb8tyw@qsl.network>e! Subject: Re: FTP Success/Failure?i* Message-ID: <3C3FD4D5.5000302@qsl.network>   Terry Marosites wrote:   > I use: >  > $  OPEN /WRITE OFIL FTP.TMP  > $  WRITE OFIL "$ set noonn > $  WRITE OFIL "$1 $  WRITE OFIL "$ define/user sys$output FTP.TMP2"a) > $  WRITE OFIL "$ UNAME = ""''uname'"" "-C > $  WRITE OFIL "FTP SEAVAX /USERNAME=''UNAME' /PASSWORD= ''PASSW'" & > $  WRITE OFIL "SET TYPE ''THE_TYPE'" > $  WRITE OFIL "VERBOSE" 0  $  WRITE OFIL "PUT ''WORK_FILE' ''TO_THAT'.TMP"  2   $  WRITE OFIL "RENAME ''TO_THAT'.TMP ''TO_THAT'"  ' > $  WRITE OFIL "$  ftp_STAT== $STATUS"i; $!  WRITE OFIL "$  deassign sys$output" ! No longer needed.29 > $  WRITE OFIL "$  write sys$output ""stat = ""+ftp_STATd5 > $  WRITE OFIL "$  if ftp_STAT.nes.""%X180180E9"" " d   > $  WRITE OFIL "$    thenF > $  WRITE OFIL "$     mail FTP.TMP /sub = ""ftp_err on ''WORK_FILE'""
 > ""''uname' u > $  WRITE OFIL "$  endifc > $  CLOSE OFILA
 > $  @FTP.TMPc* > $  write sys$output "stat = ''ftp_STAT'" > N > I have found this very reliable over the years , as long as the "put" is the' > last this before the "$ status" line.     G And with a small change, you will not have the receiving end trying to  & interpret a partially transfered file.  H Some programs have trouble dealing with files that end in the middle of 	 a record.1  F Testing for a successful rename in the FTP.TMP2 file then becomes the @ most reliable method of making sure that the transfer succeeded.     -John@ wb8tyw@qsl.network Personal Opinion Onlyo   ------------------------------  # Date: Fri, 11 Jan 2002 23:44:51 GMT32 From: "Zane H. Healy" <healyzh@shell1.aracnet.com>) Subject: Re: Getting DEC ADA 3.5A to work > Message-ID: <TNK%7.519$WM6.112291@e420r-atl1.usenetserver.com>  H Simon Clubley <simon_clubley@remove_me.altavista.co.uk-earth.ufp> wrote:  @ >>> >         ftp://ftp.cs.nyu.edu/pub/gnat/private/old/openvms/  J Many thanks for reposting the link!  I'd been hunting around on the serverK yesterday and hadn't been able to find anything since the private directory-4 itself isn't readable.  I've got it downloading now.   		Zane   ------------------------------   Date: 11 Jan 2002 18:46:18 GMT( From: peter@taronga.com (Peter da Silva): Subject: Re: historical evidence of what went wrong at DEC2 Message-ID: <a1nbtq$22vk$1@citadel.in.taronga.com>  2 In article <bZl%7.533$5Y4.14700@news.cpqcorp.net>,4 Fred Kleinsorge <kleinsorge@star.zko.dec.com> wrote:K >Now you are into the interesting stuff.  It's pretty easy for English, butrK >case *changing* becomes harder in some other languages.  The question herehL >isn't so much what is the uppercase derivitive of "/home/rne" but if thereI >*is* an uppercase derivitive of it, will it match the lowercase version.c  H Indeed. And we're talking about file names, so if I'm accessing the fileE from Houston and you're accessing it from Paris we get the same file.r  I >Unfortunately, outside of say the simple ASCII or ISO-Latin 1 sets, caseg- >rules are not a simple bit set or bit clear.e  < They're not even that simple for Latin-1. Look at DF and FF.  G >Your last example really points out some interesting areas of thought.aK >Ideally, I suppose  should match SS - if your system language was German.T  M Which is why things like case-insensitive matching rules are a user interfacetG problem. I believe they're inherently too hard to implement at the file F system level once you have to deal with anything beyond 7 bit ASCII or
 8 bit EBCDIC.r   -- s@ Rev. Peter da Silva, ULC.	                                 WWFD?  F "Be conservative in what you generate, and liberal in what you accept" 	-- Matthew 10:16 (l.trans)u   ------------------------------  % Date: Fri, 11 Jan 2002 19:03:00 -0000t/ From: "Adam Price" <adam+usenet@pappnase.co.uk>e: Subject: Re: historical evidence of what went wrong at DECC Message-ID: <FFG%7.17927$1s6.2583536@news6-win.server.ntlworld.com>e  e "Jan C. Vorbrggen" <jvorbrueggen@mediasec.de> wrote in message news:3C3EA01D.7B7D19D1@mediasec.de...aP > Is there actually an alive human language where case differences (if they haveN > the concept of case at all) routinely connote semantic differences? How many" > speakers do such languages have? >eC > I do not consider English, for instance, to belong in this class,l > the Lord notwithstanding.t >m3 In English polish to shine or Polish From Poland or A            reading and Reading both of which are instances of the I change in capitalisation showing a change in meanining and pronunciation.n  B And in German, for example polen to polarise and Polen the country  7 And in usenet common usage ALL UPPER CASE for shouting.-  E These are real examples that took only a few moments, but I can thinkr of others I'm sure.a   Adam   ------------------------------  % Date: Fri, 11 Jan 2002 14:10:32 -0500D5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>g: Subject: Re: historical evidence of what went wrong at DEC2 Message-ID: <3QG%7.581$5Y4.15329@news.cpqcorp.net>  # Peter da Silva wrote in message ...i3 >In article <aDl%7.529$5Y4.14743@news.cpqcorp.net>,T5 >Fred Kleinsorge <kleinsorge@star.zko.dec.com> wrote:hK >>Case preservation is fine, but once created lookup should be insensitive. L >>But to do this correctly from the start, you also need to then prevent the= >>creation of both Foo.Bar and foo.bar in the same directory.b >  >>This can't be fixed for UNIX.c >oG >Um, yeh, sure it can. It's easy. In fact, it's already been done. UNIXhH >doesn't care, it lets the filesystem worry about it, and there are bothI >case-bashing and case-insensitive filesystems available on UNIX systems.  >e( >Why would you think this was a problem? >E    F Because tons of things would fall over and die.  I don't know how manyH conflicting .h files I've found over the years that differ only in case.  L There is no *technical* reason why you can't do it, just the accumulation ofK a few decades of practice.  The *same* reason that we didn't just "make VMSn$ case sensitive" and be done with it.   ------------------------------   Date: 11 Jan 2002 19:11:20 GMT( From: peter@taronga.com (Peter da Silva): Subject: Re: historical evidence of what went wrong at DEC2 Message-ID: <a1ndco$23l1$1@citadel.in.taronga.com>  + In article <3C3EA1F9.7AC09FC2@mediasec.de>, 4 Jan C. Vorbrggen  <jvorbrueggen@mediasec.de> wrote:M >Really? The problem I see with your case (c) is that somebody creates a filesI >with the language set to one value, then another person with a differentsL >language setting tries to use a different-case version to access this file,M >but fails because the case-mapping rules of the two languages are different.fJ >Of course, she can still use the correct-case version to access the file.  L Based on my experience in technical support, there are many variants of this) scenario depending on the user interface:   D 	1. The second user uses a GUI to select the file. She uses the sameA 	   filename with the same case settings, no problem. This person G 	   doesn't care if the file system is case sensitive, case insensitivefD 	   for ASCII-only, case-insensitive for Latin-1, or completely i18n 	   compliant.  C 	2. The second user uses an environment where she provides the name D 	   without realtime feedback. Traditional UNIX/VMS, or she cuts and$ 	   pastes the name in a GUI, etc...  D 	   2a. The second user decides she got the name wrong, looks it up,A 	       and enters the right one. This person may gripe about theoC 	       extra step, but she's still going to end up where she needse 	       to be.  A 	   2b. The second user just goes ahead and uses the name anyway.c  B THIS is the real problem. Now we have two files that both have theB same case-insensitive representation, AND we have users who "know"A the IS is case-insensitive. They continue to update the same file  out of sync.  I And don't tell me this won't happen. The program may present her the filewH name as a default, and create it if it doesn't exist. Even if it pops up, a dialog box users frequently just say "OK".  J >I would be interested in an example where two languages have a lower-caseH >letter - say "" - and one upper-cases it to "" and the other to "AE",8 >or whatever. These are the only ones that fit the bill.   The German "SS" is one example.w  K In Turkish, "I" doesn't lowercase to "i", and "i" doesn't uppercase to "I".n  E >> Once you make the difficult choice to be case-sensitive, you don't & >> have to deal with these problems.     >...which almost never occur.u  H RISKS Digest is full of problems caused by situations which almost never occur.  D For example, almost everyone who recieved radiation therapy from the Therac-25 was unharmed.w  C >> You have appropriate and predictable behavior that is completelyh >> portable.    2 >...and surprised the vast majority of your users.  I I think that statement is too strong. Most users never even notice if thehI file system is case sensitive or not, because they simply copy file names-H verbatim. As user interfaces improve, most of them never enter the names of existing files at all.S  L >So you're trading surprising a small subset of your users in rare cases for< >surprising a majority of your users often. How kind of you.  G And they are extremely unlikely to be surprised twice. People who can't D remember that filenames are simple literal strings don't tend to use7 command lines anyway, there's way too much to remember.    -- a@ Rev. Peter da Silva, ULC.	                                 WWFD?  F "Be conservative in what you generate, and liberal in what you accept" 	-- Matthew 10:16 (l.trans)e   ------------------------------  # Date: Fri, 11 Jan 2002 19:15:47 GMTr3 From: Christopher Stacy <cstacy@spacy.Boston.MA.US>.: Subject: Re: historical evidence of what went wrong at DEC. Message-ID: <ur8owy9zw.fsf@spacy.Boston.MA.US>  C >>>>> On 11 Jan 2002 17:50:17 GMT, Peter da Silva ("Peter") writes:   :  Peter> In article <aDl%7.529$5Y4.14743@news.cpqcorp.net>,<  Peter> Fred Kleinsorge <kleinsorge@star.zko.dec.com> wrote:M  >> Case preservation is fine, but once created lookup should be insensitive. N  >> But to do this correctly from the start, you also need to then prevent the?  >> creation of both Foo.Bar and foo.bar in the same directory.i  !  >> This can't be fixed for UNIX.   N  Peter> Um, yeh, sure it can. It's easy. In fact, it's already been done. UNIXO  Peter> doesn't care, it lets the filesystem worry about it, and there are botheP  Peter> case-bashing and case-insensitive filesystems available on UNIX systems.  /  Peter> Why would you think this was a problem?t  E I missed the beginning of this thread, so forgive me if I'm restating1O what you already know: he's not describing case-bashing, nor case-insensitive. FN Case preserving means that the file system remembers what the name looked likeL when you created it, but is case-insensitive on reference.  So when you listL your directory, you'll see "FooBAR", but you can refer to it using any case.M That is probably what you did mean -- it's what Windows NT does -- but I just ' wanted to clarify what he's asking for.1  B For the kernel/filesystem itself, I don't know why it would be anyE harder to change in Unix than any other operating system that already1F supports mixed-case file names.  But Unix, especially, is a collectionD of programs, and maybe they have some kind of assumptions that wouldG require them all to be modified, too. Maybe application programs, also. E There might also be problems transitioning from one kind of system toCF the other, or interoperating (eg. over NFS) between different systems.F I don't know exactly what all the problems might be, or how they mightG be easily or not so solvable.   Perhaps Fred or someone will elaborate.a  C Probably he is unaware of Unix filesystem which do this.  I've usedtC lots of flavors of Unix systems since 1979, and I am unaware of it,l? but I haven't really followed Unix research things like Plan 9. H So which ones do this, and are there problems like the ones I suggested?   ------------------------------  # Date: Fri, 11 Jan 2002 19:25:47 GMT 3 From: Christopher Stacy <cstacy@spacy.Boston.MA.US>t: Subject: Re: historical evidence of what went wrong at DEC. Message-ID: <uofk0y9j8.fsf@spacy.Boston.MA.US>  C >>>>> On 11 Jan 2002 18:04:18 GMT, Peter da Silva ("Peter") writes:t  =  Peter> In article <1020110134809.7689A-100000@Ives.egh.com>,j*  Peter> John Santos  <JOHN@egh.com> wrote:K  >> So you are arguing that the filesystem should allow files that the userr(  >> interface won't allow you to access?  N  Peter> I think it would be better to say that if that happens, it is the user#  Peter> interface that is at fault.h  A Speaking of which, does anybody know how the hell I'm supposed toa? delete this random file I created via "tar" on Windows NT, thata= neither File Explorer nor Emacs are willing to let me delete?l It's apparently named "sail.".  ; File Explorer gets me the following operating system error:o3     "Cannot delete file: File system error (1026)."0  @ I can see the General Properties (0 byte file, unknown creator).9 But trying to access the Security Properties for the filea: always gives: "The system cannot find the file specified".  @ I wrote a program that lists the directory and gets the filename) from there, and tries to delete the file:i     "Access is denied (5)."d  A I had one of these a long time ago and managed to get rid of it, 1% but I have since forgotten the trick.t   ------------------------------   Date: 11 Jan 2002 19:24:03 GMT( From: peter@taronga.com (Peter da Silva): Subject: Re: historical evidence of what went wrong at DEC2 Message-ID: <a1ne4j$240q$1@citadel.in.taronga.com>  K In article <francini1026-ya02408000R1001020302190001@news.ne.mediaone.net>, - John J Francini <francini1026@mac.com> wrote:eL >Suppose someone -- a financial analyst -- works on a spreadsheet file namedJ >"April Sales Data" on MacOS X, which is UNIX under the hood, but uses theJ >HFS+ filesystem as the default. (It also has UFS available as an option.) >MJ >He saves it to disk.  He then later makes a word-processing document thatA >references that spreadsheet (using Microsoft-like object linkingaJ >facilities), and, instead of using a GUI-oriented file selector, he typesK >the filename in directly.  He makes an error typing it in -- enters "aprile >sales data" instead.o  H That's a user-interface problem. The user interface should at this pointH provide a list of file names that may match the one he entered and allowI him to select one. That would also deal with the situation where one nameiA was "may sales forecast" and the other was "may sales forecasts".c  & A file system is not a user interface.  G >Writing C (or its descendents), on the other hand, subjects you to theOC >tyranny. Perish forbid that you might forget that, say, the OpenGLaK >"glTexImage3D()" routine begins with a lower-case "gl", and you code it as'H >"GLTexImage3D()".  The linker will say "Duh?  I can't find that routine >anywhere!".  I No routine should ever be called glTexImage3D. Any API that requires moreEJ than 30 or 40 basic calls (or methods) is broken. With that few names, youL can make them readable. That routine, and all variants of it, should have an- interface that corresponds to something like:T  & 	mygldisplay at: (x, y, z) put: image.  G >The calls in the UNIX libraries, taken together, comprise an amazingly * >varied set of different naming standards.  M I don't know any library that you could call one of "the UNIX libraries" that L includes a name like "glTexImage3D". I believe you're assigning blame to the wrong place.   -- S@ Rev. Peter da Silva, ULC.	                                 WWFD?  F "Be conservative in what you generate, and liberal in what you accept" 	-- Matthew 10:16 (l.trans)    ------------------------------   Date: 11 Jan 2002 19:29:40 GMT( From: peter@taronga.com (Peter da Silva): Subject: Re: historical evidence of what went wrong at DEC2 Message-ID: <a1nef4$249t$1@citadel.in.taronga.com>  B In article <b4G%7.614867$8q.49627269@bin2.nnrp.aus1.giganews.com>,) Bill Todd <billtodd@metrocast.net> wrote:3F >I may indeed have gotten a tad carried away on this particular point,H >especially as mixed-language file names should be fairly unusual.  I'llC >stand by the assertion that multiple representations of *the same*DG >presentation character (e.g., single-character and composed) should beO* >treated as identical on look-up, however.   In the user interface? Sure.  H In the file system? That's putting a heck of a lot of code into a commonH path for operations that are supposed to be fast. Have you looked at the! Unicode conversion tables lately?T   -- '@ Rev. Peter da Silva, ULC.	                                 WWFD?  F "Be conservative in what you generate, and liberal in what you accept" 	-- Matthew 10:16 (l.trans)_   ------------------------------  % Date: Fri, 11 Jan 2002 14:37:18 -0500d5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>p: Subject: Re: historical evidence of what went wrong at DEC2 Message-ID: <adH%7.586$5Y4.15379@news.cpqcorp.net>  # Peter da Silva wrote in message ...' >KJ >No routine should ever be called glTexImage3D. Any API that requires moreK >than 30 or 40 basic calls (or methods) is broken. With that few names, youiJ >can make them readable. That routine, and all variants of it, should have an. >interface that corresponds to something like: >,' > mygldisplay at: (x, y, z) put: image.n >oH >>The calls in the UNIX libraries, taken together, comprise an amazingly+ >>varied set of different naming standards.  >eI >I don't know any library that you could call one of "the UNIX libraries"  thatI >includes a name like "glTexImage3D". I believe you're assigning blame to@ thet
 >wrong place.  >s    L glTextImage3D is from the OpenGL libraries, originated by SGI for UNIX.  ButL that isn't much different than the X11 and Motif libraries.  Heck, they evenH insist on uppercasing the first character of the .h file.  You might notL call these "UNIX" libraries, but they originated on UNIX, for UNIX, and have become the standards on UNIX.   E Because there ARE NO "norms" for UNIX naming, there is no good way toeB prevent namespace collision in libraries.  So you get crazy namingL conventions that assume that the combination of a couple characters, and theB case - will minimize conflicts with user code and other libraries.   ------------------------------   Date: 11 Jan 2002 19:33:06 GMT( From: peter@taronga.com (Peter da Silva): Subject: Re: historical evidence of what went wrong at DEC2 Message-ID: <a1neli$24d0$1@citadel.in.taronga.com>  2 In article <3QG%7.581$5Y4.15329@news.cpqcorp.net>,4 Fred Kleinsorge <kleinsorge@star.zko.dec.com> wrote:) >>Why would you think this was a problem?y  G >Because tons of things would fall over and die.  I don't know how many.I >conflicting .h files I've found over the years that differ only in case.h  H In the same subsystem from the same vendor? I find that hard to believe.  F If not, well, I've run into a whole bunch of conflicting .h files thatH don't even differ in case. Heck, I found myself binary patching a kernelK 5 minutes before an important presentation because two different networkingIE vendors had picked the name "np_tab" as a global symbol in their API.    -- r@ Rev. Peter da Silva, ULC.	                                 WWFD?  F "Be conservative in what you generate, and liberal in what you accept" 	-- Matthew 10:16 (l.trans)    ------------------------------   Date: 11 Jan 2002 19:41:54 GMT( From: peter@taronga.com (Peter da Silva): Subject: Re: historical evidence of what went wrong at DEC2 Message-ID: <a1nf62$24mm$1@citadel.in.taronga.com>  . In article <ur8owy9zw.fsf@spacy.Boston.MA.US>,5 Christopher Stacy  <cstacy@spacy.Boston.MA.US> wrote: F >I missed the beginning of this thread, so forgive me if I'm restatingP >what you already know: he's not describing case-bashing, nor case-insensitive. O >Case preserving means that the file system remembers what the name looked likeu; >when you created it, but is case-insensitive on reference.h  C That's what I was referring to as "case-insensitive" in contrast torI "case-bashing". Otherwise there's no point to making that distinction, isi there.  C >For the kernel/filesystem itself, I don't know why it would be anyhF >harder to change in Unix than any other operating system that already  >supports mixed-case file names.  E You don't need to. It's already been done. Right now I have access ton8 both kinds of filesystems from UNIX. It's not a problem.  D >Probably he is unaware of Unix filesystem which do this.  I've usedD >lots of flavors of Unix systems since 1979, and I am unaware of it,@ >but I haven't really followed Unix research things like Plan 9.I >So which ones do this, and are there problems like the ones I suggested?a  H You mentioned one. NTFS. Unix systems that support NTFS and UNIX systemsK that support CIFS access to NTFS volumes already experience this behaviour.rI There's also the various case-preserving extensions to FAT and FAT32. I'mrK pretty sure you can get a Linux version right now that will happily installdK into your C: FAT partition on right next to your Windows or WinNT directory-L and operate under FAT/FAT32 case-preserving semantics when you boot into it.   -- I@ Rev. Peter da Silva, ULC.	                                 WWFD?  F "Be conservative in what you generate, and liberal in what you accept" 	-- Matthew 10:16 (l.trans)e   ------------------------------  # Date: Fri, 11 Jan 2002 19:50:57 GMTt* From: "Bill Todd" <billtodd@metrocast.net>: Subject: Re: historical evidence of what went wrong at DECB Message-ID: <BmH%7.438341$C8.31858029@bin4.nnrp.aus1.giganews.com>  5 "Peter da Silva" <peter@taronga.com> wrote in messagel, news:a1nef4$249t$1@citadel.in.taronga.com...D > In article <b4G%7.614867$8q.49627269@bin2.nnrp.aus1.giganews.com>,+ > Bill Todd <billtodd@metrocast.net> wrote:eH > >I may indeed have gotten a tad carried away on this particular point,J > >especially as mixed-language file names should be fairly unusual.  I'llE > >stand by the assertion that multiple representations of *the same*aI > >presentation character (e.g., single-character and composed) should bet, > >treated as identical on look-up, however. >h > In the user interface? Sure. >-J > In the file system? That's putting a heck of a lot of code into a commonJ > path for operations that are supposed to be fast. Have you looked at the# > Unicode conversion tables lately?>  H Within the past couple of years, anyway.  The amount of code required toH cover all cases is non-trivial, but the amount of code *executed* on anyJ single look-up operation is insignificant compared with the other costs of the look-up.   - bill   ------------------------------   Date: 11 Jan 2002 19:49:58 GMT( From: peter@taronga.com (Peter da Silva): Subject: Re: historical evidence of what went wrong at DEC2 Message-ID: <a1nfl6$252e$1@citadel.in.taronga.com>  , In article <3C3F2C31.DCBDD723@jetnet.ab.ca>,- Ben Franchuk  <bfranchuk@jetnet.ab.ca> wrote:  >Peter da Silva wrote: >> >He also seems to think@ >> >that a filesystem design can be completely divorced from the% >> >hardware the system was built on.t   >> Pretty much, yes.  > >I got a DEC tape here... How do I make this '/' for linux? :)  G I've booted RSX-11 on a DECtape. I don't see how it would be any hardert to boot 2.11BSD on it.  C >Right now all disk hardware assumes 8 bits is a byte and all media @ >is blocks of 2^n. That is a bit of a pain for me as am building7 >a 12 bit computer. Do I go 8 bits or 12 bits for data?   D Try them both. I mean, there's no reason you can't go with 336 wordsJ to a block, packed 2 words to 3 octets. Or 288 words if you want the blockI size to be a multiple of 12, that would only waste 5/27th of the space...h a little over 1/6th.   >What if; >I had a replica of a computer that uses a magnetic drum asv
 >main memory.t? >I say no! file system and computer hardware are tied together.   F I didn't say they should be completely divorced, I said they could be.   -- l@ Rev. Peter da Silva, ULC.	                                 WWFD?  F "Be conservative in what you generate, and liberal in what you accept" 	-- Matthew 10:16 (l.trans)    ------------------------------  # Date: Fri, 11 Jan 2002 20:04:34 GMT * From: "Bill Todd" <billtodd@metrocast.net>: Subject: Re: historical evidence of what went wrong at DECB Message-ID: <mzH%7.615288$8q.49706979@bin2.nnrp.aus1.giganews.com>  5 "Peter da Silva" <peter@taronga.com> wrote in messageh, news:a1ne4j$240q$1@citadel.in.taronga.com... > In article@ <francini1026-ya02408000R1001020302190001@news.ne.mediaone.net>,/ > John J Francini <francini1026@mac.com> wrote:mH > >Suppose someone -- a financial analyst -- works on a spreadsheet file named L > >"April Sales Data" on MacOS X, which is UNIX under the hood, but uses theL > >HFS+ filesystem as the default. (It also has UFS available as an option.) > >iL > >He saves it to disk.  He then later makes a word-processing document thatC > >references that spreadsheet (using Microsoft-like object linking L > >facilities), and, instead of using a GUI-oriented file selector, he typesF > >the filename in directly.  He makes an error typing it in -- enters "april > >sales data" instead.e >rJ > That's a user-interface problem. The user interface should at this pointJ > provide a list of file names that may match the one he entered and allowK > him to select one. That would also deal with the situation where one nameeC > was "may sales forecast" and the other was "may sales forecasts".l  L That might be nice, but in all probability cannot be accomplished without anI exhaustive scan of all files in the current directory (if a full filespecFJ wasn't entered:  otherwise, since some directory name could also have beenL distorted, you must scan all files on the entire system).  So it's often not a feasible solution.  K Does that mean that a smaller (but very significant) class of such problemsi& should not be addressed?  I think not.   >g( > A file system is not a user interface.  K But since users are arguably the most important clients of the file system,rI it needs to accommodate the needs of that user interface.  In the examplegI above, its look-up mechanism should make reasonable attempts to allow theoH interface to resolve common distortions - and in the case of distortionsI caused by case (or multiple binary representations of the same character)BJ only (which definitely qualify as common and as far more counter-intuitive= than simple user-entry mistakes), this is not that difficult.a   - bill   ------------------------------  % Date: Fri, 11 Jan 2002 12:46:23 -0700t+ From: Ben Franchuk <bfranchuk@jetnet.ab.ca>-: Subject: Re: historical evidence of what went wrong at DEC, Message-ID: <3C3F410F.A2E57270@jetnet.ab.ca>   Peter da Silva wrote:   I > >Writing C (or its descendents), on the other hand, subjects you to the E > >tyranny. Perish forbid that you might forget that, say, the OpenGL M > >"glTexImage3D()" routine begins with a lower-case "gl", and you code it asbJ > >"GLTexImage3D()".  The linker will say "Duh?  I can't find that routine > >anywhere!". > K > No routine should ever be called glTexImage3D. Any API that requires moretL > than 30 or 40 basic calls (or methods) is broken. With that few names, youN > can make them readable. That routine, and all variants of it, should have an/ > interface that corresponds to something like:m > / >         mygldisplay at: (x, y, z) put: image.U > I > >The calls in the UNIX libraries, taken together, comprise an amazingly , > >varied set of different naming standards. > O > I don't know any library that you could call one of "the UNIX libraries" thatsN > includes a name like "glTexImage3D". I believe you're assigning blame to the > wrong place.  G But is this not a sign that the language is a fault. C was designed forh amF small systems. 64KB was the max memory at the time. Even big computersF at the time like the PDP-10 with 256KW of memory does not compare with? Bloatware 2003. You can't write Glib.display.put(a,b,c) or evenaH Graphlib->display->put(a,b,c) thus making the interface portable because> the language was designed for a few global macros and defines.E And I don't think objects are the not way for segmentation because in2 mostA cases a complex structure can't be shared with another structure.j ObjectshF work for small stuff not big stuff. A better structure of hierarchy isE needed. Computer science people tend to be big on abstract theory butv littleH with a working software. While I know 64k is too small of a abc compilerA need 64Meg be still to small. BTW I still want a ADC for multipleg	 precisionh work!y   -- o% Ben Franchuk - Dawn * 12/24 bit cpu *e+ www.jetnet.ab.ca/users/bfranchuk/index.html    ------------------------------  % Date: Fri, 11 Jan 2002 12:55:52 -0700r+ From: Ben Franchuk <bfranchuk@jetnet.ab.ca> : Subject: Re: historical evidence of what went wrong at DEC, Message-ID: <3C3F4348.318C775A@jetnet.ab.ca>   Peter da Silva wrote:i > E > >Right now all disk hardware assumes 8 bits is a byte and all media B > >is blocks of 2^n. That is a bit of a pain for me as am building9 > >a 12 bit computer. Do I go 8 bits or 12 bits for data?  > F > Try them both. I mean, there's no reason you can't go with 336 wordsL > to a block, packed 2 words to 3 octets. Or 288 words if you want the blockK > size to be a multiple of 12, that would only waste 5/27th of the space...y > a little over 1/6th.  F Most likly I would pack 1536 bytes into 512 24 bit words. Thus I would useaG 3 sectors on a floppy for 1 logical block. No wasted space just strangee formating. t -- e% Ben Franchuk - Dawn * 12/24 bit cpu *-+ www.jetnet.ab.ca/users/bfranchuk/index.htmle   ------------------------------  % Date: Fri, 11 Jan 2002 15:29:00 -0500w From: "JD" <dyson@jdyson.com>a: Subject: Re: historical evidence of what went wrong at DEC2 Message-ID: <vPH%7.207$vb1.35477@news1.iquest.net>  J <jmfbahciv@aol.com> wrote in message news:a1mf03$qb7$2@bob.news.rcn.net.../ > In article <10JAN200219011881@gerg.tamu.edu>,m- >    carl@gerg.tamu.edu (Carl Perkins) wrote:I$ > >"JD" <dyson@jdyson.com> writes...I > >}Yes it does...   The mistaken tradition of normalizing the display ofHG > >}filenames to a fixed case (e.g. RAD50/SIXBIT) might have given rise C > >}to the mistaken notion that filename comparison should be case f > insensitive. > >}John > >nF > >The problem is that you just made a false statement. This notion isG > >*not* mistaken. All filename comparisons should be case insensitive. I > >This is true, good, and essential in any properly designed filesystem.r > > > Note that JD does not know what a filesystem is.  He appearsB > to think that it is separate from an OS.  He also seems to think> > that a filesystem design can be completely divorced from the# > hardware the system was built on.t > ( Are you starting personal insults again?  J Of course, older systems didn't distinguish very well between filesystems,A kernels and command line semantics.   This is EXACTLY the problem I that you might be having by incorporating DCL limitations into filesystem  or OS/kernel code.  O Again, it seems that it is you who forget that stuff isn't all mangled togetherxF on well designed systems.    FFS (for example) provides both unix fileJ semantics and optimizations for real hardware devices.   The optimizationsB for the real hardware shouldn't impose LIMITATIONS onto the higher level portions of the code.g  J It is interesting that when I decided to back away from the argument, then= you childishly make a personal insult.   Waiting for apology.n   John   ------------------------------   Date: 11 Jan 2002 20:19:40 GMT( From: peter@taronga.com (Peter da Silva): Subject: Re: historical evidence of what went wrong at DEC2 Message-ID: <a1nhcs$260b$1@citadel.in.taronga.com>  A In article <j8G%7.117860$Yf.8045386@bin3.nnrp.aus1.giganews.com>,t) Bill Todd <billtodd@metrocast.net> wrote:t > 6 >"Peter da Silva" <peter@taronga.com> wrote in message- >news:a1n907$2159$1@citadel.in.taronga.com...g. >> In article <3C3EA01D.7B7D19D1@mediasec.de>,7 >> Jan C. Vorbrggen  <jvorbrueggen@mediasec.de> wrote:4M >> >Is there actually an alive human language where case differences (if theya >haveaG >> >the concept of case at all) routinely connote semantic differences?e >>7 >> I don't even *pronounce* polish and Polish the same.a > F >The replies to Jan's question that I've seen so far seem to treat his. >inclusion of 'routinely' above as irrelevant.  
 # wc words   234936  234936 2486813 words # grep '^[A-Z]' words | wc    24256   24256  237674  C So case is significant in about 1/10th of the words in the spellingf dictionary.h  A # grep '^[A-Z]' words | tr '[A-Z]' '[a-z]' | sort > /tmp/target  n # sort words > /tmp/wordsa& # comm -12 /tmp/target /tmp/words | wc     1322    1323    9641  ( And 1300 of them have both cases listed.   --  @ Rev. Peter da Silva, ULC.	                                 WWFD?  F "Be conservative in what you generate, and liberal in what you accept" 	-- Matthew 10:16 (l.trans),   ------------------------------    Date: 11 Jan 2002 14:35:59 -0600 From: briggs@encompasserve.org: Subject: Re: historical evidence of what went wrong at DEC3 Message-ID: <ipEyzocaW+Gf@eisner.encompasserve.org>w  Z In article <3C3F2C31.DCBDD723@jetnet.ab.ca>, Ben Franchuk <bfranchuk@jetnet.ab.ca> writes: > Peter da Silva wrote:s > N >> This is because UNIX uses layering to isolate the filesystem implementation9 >> from the rest of the system. And this is a good thing.t >> e >> >He also seems to think@ >> >that a filesystem design can be completely divorced from the% >> >hardware the system was built on.t >> c >> Pretty much, yes. >> i > ? > I got a DEC tape here... How do I make this '/' for linux? :)   @ newfs and mount.  I assume that a vintage 1980's copy of BSD forA the 730 has TU58 support.  If you're running on an HSC-50 you mayt! need to port Linux to the PDP-11.g  F But that's irrelevant.  We're talking file system design.  And DECtapeC is a rewriteable, block addressible medium.  We _can_ design a filenE system for an arbitrary rewriteable, block structured medium and haven3 it work on DECtape.  And in fact we _have done so_.o  3 You did realize that DECtape is a trademark, right?X   	John Briggs   ------------------------------    Date: 11 Jan 2002 14:26:11 -0600 From: rivie@cougar.no.domain: Subject: Re: historical evidence of what went wrong at DEC3 Message-ID: <slrna3uj6t.4ts.rivie@cougar.no.domain>-  H In article <a1nef4$249t$1@citadel.in.taronga.com>, Peter da Silva wrote:J > In the file system? That's putting a heck of a lot of code into a commonJ > path for operations that are supposed to be fast. Have you looked at the# > Unicode conversion tables lately?l   Aha! The truth will out!  F I've always suspected that the only reason Unix was case-sensitive wasG that someone was too lazy to do a case-insensitive compare in the firstcH versions and, rather than fixing the problem, invented a rationalization. to make it appear to be The Right Thing To Do. -- e
 Roger Ivie ivie@cc.usu.edum    > -----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----A http://www.newsfeeds.com - The #1 Newsgroup Service in the World!t> -----==  Over 80,000 Newsgroups - 16 Different Servers! =-----   ------------------------------  % Date: Fri, 11 Jan 2002 15:46:14 -0500t From: "JD" <dyson@jdyson.com>d: Subject: Re: historical evidence of what went wrong at DEC2 Message-ID: <G3I%7.210$vb1.35662@news1.iquest.net>  J <jmfbahciv@aol.com> wrote in message news:a1mfaq$qb7$3@bob.news.rcn.net...4 > In article <7Up%7.167$vb1.27100@news1.iquest.net>,# >    "JD" <dyson@jdyson.com> wrote:2 > <snip> > = > >Again:  FILENAMES aren't necessarily English proper names.o > 8 > They never were.  Just about everything we shipped was? > an acronym or a description of the _function_ of the program.. > P Or even a date.   On real (current) systems, it is easy to use a short formatted9 text identification, that includes some descriptive info.w   John   ------------------------------  % Date: Fri, 11 Jan 2002 15:48:47 -0500s From: "JD" <dyson@jdyson.com>e: Subject: Re: historical evidence of what went wrong at DEC2 Message-ID: <36I%7.211$vb1.35588@news1.iquest.net>  J <jmfbahciv@aol.com> wrote in message news:a1menc$qb7$1@bob.news.rcn.net...- > In article <3C3DC83C.B4129E83@mediasec.de>, K >    Jan C. =?iso-8859-1?Q?Vorbr=FCggen?= <jvorbrueggen@mediasec.de> wrote:hJ > >> Anyone who believes that capitalization isn't part of standard usage @ > >> probably believes that ee cummings writing style is normal. > >IK > >You don't seem to understand we want a case-preserving, case-insensitive J > >system - just about the only thing NTFS got right, with ODS-5 followingK > >suit. Capitalization is an aid to text comprehension, but it changes theIA > >meaning only in very rare cases. _That_ is the proper analogy.i > A > Note that most of the argument for case sensitivity is smushinghB > the documenation of the application into a filename <shudder>^4. > D Nope, YOU are the individual who reduces the argument to the absurd.  D Case senstitivity has nothing to do with including all documentationG in the filename.   It has to do with being sensitive to the differencessI of the actual characters used in filenames, and the abillity to naturallyhD handle comparisons and NOT being roman/english centric, and avoidingG ad-hoc rules that create equivalence sets that have little reason other  than legacy OS behavior.   John   ------------------------------    Date: 11 Jan 2002 14:38:01 -0600 From: rivie@cougar.no.domain: Subject: Re: historical evidence of what went wrong at DEC3 Message-ID: <slrna3ujt2.4ts.rivie@cougar.no.domain>   H In article <a1nf62$24mm$1@citadel.in.taronga.com>, Peter da Silva wrote:0 > In article <ur8owy9zw.fsf@spacy.Boston.MA.US>,7 > Christopher Stacy  <cstacy@spacy.Boston.MA.US> wrote:o > E >>Probably he is unaware of Unix filesystem which do this.  I've used E >>lots of flavors of Unix systems since 1979, and I am unaware of it,i > J > You mentioned one. NTFS. Unix systems that support NTFS and UNIX systemsM > that support CIFS access to NTFS volumes already experience this behaviour.sK > There's also the various case-preserving extensions to FAT and FAT32. I'msM > pretty sure you can get a Linux version right now that will happily install M > into your C: FAT partition on right next to your Windows or WinNT directoryeN > and operate under FAT/FAT32 case-preserving semantics when you boot into it.  F Another excellent example is HFS+ under MacOS X. IIRC, there have been: some trouble installing stuff under MacOS X because of theH case-insensitivity. I'm thinking it was Perl, but I don't really recall;B I had enough unrelated trouble with MacOS X that I moved my Mac toF NetBSD until my wife bought a camcorder and insisted that I re-install MacOS 9. --  
 Roger Ivie ivie@cc.usu.edu-    > -----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----A http://www.newsfeeds.com - The #1 Newsgroup Service in the World!e> -----==  Over 80,000 Newsgroups - 16 Different Servers! =-----   ------------------------------  % Date: Fri, 11 Jan 2002 16:18:56 -0500k From: "JD" <dyson@jdyson.com>e: Subject: Re: historical evidence of what went wrong at DEC2 Message-ID: <jyI%7.217$vb1.35482@news1.iquest.net>  W <rivie@cougar.no.domain> wrote in message news:slrna3uj6t.4ts.rivie@cougar.no.domain...pJ > In article <a1nef4$249t$1@citadel.in.taronga.com>, Peter da Silva wrote:L > > In the file system? That's putting a heck of a lot of code into a commonL > > path for operations that are supposed to be fast. Have you looked at the% > > Unicode conversion tables lately?s >  > Aha! The truth will out! > H > I've always suspected that the only reason Unix was case-sensitive wasI > that someone was too lazy to do a case-insensitive compare in the firstrJ > versions and, rather than fixing the problem, invented a rationalization0 > to make it appear to be The Right Thing To Do. >.K The amazing thing is that 'good' UNIX OSes can implement case insensitivitymR in the filesystem code.   For reverse compatibility with NTFS-like filesystems, it# is easy to add (and has been done.)d  M Case folding during comparisons just isn't desirable to do, except for legacy N reasons.   Any reason to do case folding is a roman/English centric provincial design.a   John   ------------------------------  % Date: Fri, 11 Jan 2002 16:21:47 -0500,5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>@: Subject: Re: historical evidence of what went wrong at DEC2 Message-ID: <6LI%7.590$5Y4.15362@news.cpqcorp.net>  = JD wrote in message <36I%7.211$vb1.35588@news1.iquest.net>.... >>K ><jmfbahciv@aol.com> wrote in message news:a1menc$qb7$1@bob.news.rcn.net...e. >> In article <3C3DC83C.B4129E83@mediasec.de>,L >>    Jan C. =?iso-8859-1?Q?Vorbr=FCggen?= <jvorbrueggen@mediasec.de> wrote:J >> >> Anyone who believes that capitalization isn't part of standard usageA >> >> probably believes that ee cummings writing style is normal.s >> >L >> >You don't seem to understand we want a case-preserving, case-insensitiveK >> >system - just about the only thing NTFS got right, with ODS-5 followingnL >> >suit. Capitalization is an aid to text comprehension, but it changes theB >> >meaning only in very rare cases. _That_ is the proper analogy. >>B >> Note that most of the argument for case sensitivity is smushingC >> the documenation of the application into a filename <shudder>^4.f >>E >Nope, YOU are the individual who reduces the argument to the absurd.i >nE >Case senstitivity has nothing to do with including all documentationTH >in the filename.   It has to do with being sensitive to the differencesJ >of the actual characters used in filenames, and the abillity to naturallyE >handle comparisons and NOT being roman/english centric, and avoiding H >ad-hoc rules that create equivalence sets that have little reason other >than legacy OS behavior.r >o    7 Just so I understand - by 'legacy' you mean 'not UNIX'.l   ------------------------------   Date: 11 Jan 2002 22:01:02 GMT( From: peter@taronga.com (Peter da Silva): Subject: Re: historical evidence of what went wrong at DEC2 Message-ID: <a1nnau$290n$1@citadel.in.taronga.com>  2 In article <adH%7.586$5Y4.15379@news.cpqcorp.net>,4 Fred Kleinsorge <kleinsorge@star.zko.dec.com> wrote:J >>I don't know any library that you could call one of "the UNIX libraries"L >>that includes a name like "glTexImage3D". I believe you're assigning blame >>to the wrong place.   L >glTextImage3D is from the OpenGL libraries, originated by SGI for UNIX. ButM >that isn't much different than the X11 and Motif libraries.  Heck, they evena: >insist on uppercasing the first character of the .h file.  C X was developed for both UNIX and VMS concurrently at MIT. It's now-K available on just about every modern operating system that I know of. MotifnH is based on the OS/2 Presentation Manager and developed by IBM and theirJ partners.  OpenGL was intended from the start to run on multiple operatingL systems, and is managed by another industry consortium... and while it seems= to be shiped with Windows it's an extra-cost product on UNIX.   I I would suggest you direct your ire towards Project Athena at MIT... thataK seems to be where that kind of RansomNote naming convention comes from, andeF it does seem to be mostly isolated to the X universe, thank the stars.   -- t@ Rev. Peter da Silva, ULC.	                                 WWFD?  F "Be conservative in what you generate, and liberal in what you accept" 	-- Matthew 10:16 (l.trans)h   ------------------------------   Date: 11 Jan 2002 22:10:34 GMT( From: peter@taronga.com (Peter da Silva): Subject: Re: historical evidence of what went wrong at DEC2 Message-ID: <a1nnsq$29al$1@citadel.in.taronga.com>  B In article <mzH%7.615288$8q.49706979@bin2.nnrp.aus1.giganews.com>,) Bill Todd <billtodd@metrocast.net> wrote:aM >That might be nice, but in all probability cannot be accomplished without ancJ >exhaustive scan of all files in the current directory (if a full filespecK >wasn't entered:  otherwise, since some directory name could also have beensM >distorted, you must scan all files on the entire system).  So it's often notw >a feasible solution.t  G And this is somehow more processor-intensive than handling full UnicodenD case conversion rules and locale management at the filesystem level?  L >But since users are arguably the most important clients of the file system,: >it needs to accommodate the needs of that user interface.  H It can best do that by reliably and consistently responding to requests.   -- e@ Rev. Peter da Silva, ULC.	                                 WWFD?  F "Be conservative in what you generate, and liberal in what you accept" 	-- Matthew 10:16 (l.trans)    ------------------------------  % Date: Fri, 11 Jan 2002 16:11:49 -0800l+ From: Mark Crispin <mrc@CAC.Washington.EDU>h: Subject: Re: historical evidence of what went wrong at DECU Message-ID: <Pine.NXT.4.50.0201111608560.4273-100000@Tomobiki-Cho.CAC.Washington.EDU>0  % On 11 Jan 2002, Peter da Silva wrote:aK > I would suggest you direct your ire towards Project Athena at MIT... thatDM > seems to be where that kind of RansomNote naming convention comes from, andaH > it does seem to be mostly isolated to the X universe, thank the stars.  < Unfortunately, Microsoft has also succumbed to that disease.  J The disease seems to have its origins in Pascal.  A lot of Pascal programs1 written in the 1980s used that naming convention.a  
 -- Mark --   http://staff.washington.edu/mrciF Science does not emerge from voting, party politics, or public debate.   ------------------------------   Date: 12 Jan 2002 01:38:23 GMT( From: peter@taronga.com (Peter da Silva): Subject: Re: historical evidence of what went wrong at DEC2 Message-ID: <a1o42f$2fnr$1@citadel.in.taronga.com>  M In article <slrna3uj6t.4ts.rivie@cougar.no.domain>,  <ivie@cc.usu.edu> wrote:xG >I've always suspected that the only reason Unix was case-sensitive was0H >that someone was too lazy to do a case-insensitive compare in the firstI >versions and, rather than fixing the problem, invented a rationalizatione/ >to make it appear to be The Right Thing To Do.1  H If you're really concerned, Dennis Ritchie is quite approachable and hasK been quite ready in the past to discuss the history and background of UNIX.o  H Alternatively, quite a lot of old UNIX source code is available from theK PUPS archive. You can examine it yourself. The code starts around line 7530v, in the 6th edition kernel in the Lions book.   -- m@ Rev. Peter da Silva, ULC.	                                 WWFD?  F "Be conservative in what you generate, and liberal in what you accept" 	-- Matthew 10:16 (l.trans)s   ------------------------------  # Date: Sat, 12 Jan 2002 04:05:31 GMT ' From: "David G. Conroy" <dgc@spies.com> : Subject: Re: historical evidence of what went wrong at DECC Message-ID: <fCO%7.28121$q12.2909616296@newssvr21.news.prodigy.com>    In article sJ <Pine.NXT.4.50.0201111608560.4273-100000@Tomobiki-Cho.CAC.Washington.EDU>,, Mark Crispin <mrc@CAC.Washington.EDU> wrote:L > The disease seems to have its origins in Pascal.  A lot of Pascal programs3 > written in the 1980s used that naming convention.i  B All of the examples in the MESA 5.0 Manual (PARC, 1979) have names
 in this form.o dgcs   ------------------------------  % Date: Sat, 12 Jan 2002 18:49:32 +1300t From: "AG" <ang@xtra.co.nz>t: Subject: Re: historical evidence of what went wrong at DEC3 Message-ID: <P4Q%7.1619$Qz1.203891@news.xtra.co.nz>l  5 "Peter da Silva" <peter@taronga.com> wrote in messagew, news:a1nbil$22kh$1@citadel.in.taronga.com...H > In article <V5b%7.883$Qz1.109089@news.xtra.co.nz>, AG <ang@xtra.co.nz> wrote:L > >> How do you in general determine what the uppercase form of an arbitraryI > >> character is? What is the uppercase form of "/home/rne"? What about04 > >> about ""? Should that match a file named "SS"? > I > >But... *can* you have that single character as a valid German word? Orn startb > >a word with it? >oD > I don't speak german, so I have no idea, and I'm not sure why it's
 important.  < Because in your argument above you asked: 'Should that matchD a file named "SS"?' I sort of doubt it could be an issue if there is< no case where that char could/would be used as a stand-aloneH file-name single-char thing. If it's just a totally abstract tag to callB a file something then, as pointed out in some other posts, you canB just use some random generated binary indexes and be done with it.A If, on the other hand, you wish to load it with some conventional @ meaning, then it's surely better to stick with that conventional@ meaning - including the fact that that single character *cannot*9 be used in the way you seemed to imply. So your objection  seems to be a no-issue.3   ------------------------------  % Date: Sat, 12 Jan 2002 19:23:55 +1300f From: "AG" <ang@xtra.co.nz>-: Subject: Re: historical evidence of what went wrong at DEC3 Message-ID: <2BQ%7.1629$Qz1.204836@news.xtra.co.nz>o  5 "Peter da Silva" <peter@taronga.com> wrote in messagev, news:a1nhcs$260b$1@citadel.in.taronga.com...C > In article <j8G%7.117860$Yf.8045386@bin3.nnrp.aus1.giganews.com>,i+ > Bill Todd <billtodd@metrocast.net> wrote:a > >"8 > >"Peter da Silva" <peter@taronga.com> wrote in message/ > >news:a1n907$2159$1@citadel.in.taronga.com... 0 > >> In article <3C3EA01D.7B7D19D1@mediasec.de>,9 > >> Jan C. Vorbrggen  <jvorbrueggen@mediasec.de> wrote:GJ > >> >Is there actually an alive human language where case differences (if they > >havebI > >> >the concept of case at all) routinely connote semantic differences?c > >>9 > >> I don't even *pronounce* polish and Polish the same.  > >oH > >The replies to Jan's question that I've seen so far seem to treat his0 > >inclusion of 'routinely' above as irrelevant.  F I second that one, please note that it is (relatively) easy to come upA with an exceptional cases for (almost?) any language rule. I sure @ know some but the point above was specifically about "routinely"@ and "semantic differences" - NOT spelling errors as such if they# don't trigger semantic differences.eE So, what about, as asked in the original question: "routinely connoten semantic differences?"   >C > # wc words  >   234936  234936 2486813 words > # grep '^[A-Z]' words | wc >    24256   24256  237674 >lE > So case is significant in about 1/10th of the words in the spellingw
 > dictionary._ > A > # grep '^[A-Z]' words | tr '[A-Z]' '[a-z]' | sort > /tmp/targetg > # sort words > /tmp/words-( > # comm -12 /tmp/target /tmp/words | wc >     1322    1323    9641 > * > And 1300 of them have both cases listed. >r   Statistics Statistics ...a  :   A few questions; how many of those words were the proper< nouns, names of the places, personal names etc? Which don't, normally, have the other form.>   The words listed in *both* cases are *obvious* cases arguingD for the case-insensitivity. After all, the dictionary authors neededE to include a duplicate just for the sake of a case-sensitive engines.e= So, count about 5.45% out of your own count against you (1322l out of 24256).B   And finally, what is the relative statistics weight of the words= counted? Did you (from you sample it seems you did) count allb@ and every word in the dictionary as exactly the same probability
 to encounter?i   ------------------------------  # Date: Sat, 12 Jan 2002 06:21:41 GMT-- From: "John E. Malmberg" <wb8tyw@qsl.network>=: Subject: Re: historical evidence of what went wrong at DEC* Message-ID: <3C3FDA92.5020508@qsl.network>   Christopher Stacy wrote:  C > Speaking of which, does anybody know how the hell I'm supposed tohA > delete this random file I created via "tar" on Windows NT, that.? > neither File Explorer nor Emacs are willing to let me delete?b  > It's apparently named "sail.".    H The feedback that my former employer got from Microsoft is that Windows G NT expected that if an application knew how to create a file, it would o know how to delete it.    F Try using dir -x on the file to see if it has an 8.3 name that can be  used to access the file.  H For some of the interesting file names that people have managed to save F things with, there is an article in the Microsoft TechNet database on G how to use the native fully qualified NT path specifications to delete 8 files.  ) I do not know if it can handle your file.R  G You may have to get a delete file program that has been built with the 1( POSIX api instead of the Windows NT API.  E Of course you could always just reformat the drive, and just uncheck e  that file from the restore list.     -JohnC wb8tyw@qsl.network Personal Opinion Onlye   ------------------------------  % Date: Fri, 11 Jan 2002 21:43:01 -0000p= From: "David McKenzie" <david.mckenzie@spitfire0.demon.co.uk> > Subject: Re: How many Fibre Channel - Multiple Hosts/4 FabricsB Message-ID: <1010785387.29143.0.nnrp-07.c1edba74@news.demon.co.uk>  	 Hi Keith,n  J I think some of these four fabric configurations are more possible now dueI to WDM. This means that you would still want two links between the sites, K but if they are dark fibre, you can multiplex many channels down each pipe.   K My experience to dat however in the UK is that the cost of WDM equipment is C close the the cost of a link. In the US that may be very different.n  > "Keith Parris" <keithparris_NOSPAM@yahoo.com> wrote in message7 news:cf15391e.0201102041.5fccf96b@posting.google.com...i4 > "C.W.Holeman II" <cwhii@mail.com> wrote in message$ news:<3C3DD4C5.4D81B376@mail.com>...2 > > Any guestimates as to about how many of these: > >rH > > Configuring Fibre Channel as an OpenVMS Cluster Storage Interconnect$ > > Multiple Hosts With Four Fabrics > >gL http://www.openvms.compaq.com/openvms/fibre/doc/6318pro_003.html#mult_hosts_ four_fabrics_f > >m > > are in existence?n >uF > All the high-availability FC-based clusters I've had any involvementF > with use just two independent FC fabrics.  Perhaps that's due to theF > present situation where FC net gear still seems relatively expensiveB > compared with, for example, GbE.  (And since I tend to work withE > disaster-tolerant clusters, the cost of additional inter-site linksiH > can be an even bigger consideration, especially when you consider thatD > you need redundant SCS-capable inter-site links in addition to the > redundant FC links.) >hF > Triple redundancy would allow retaining the protection of redundancyH > while a failure is being repaired.  (I've seen a case where one of twoG > inter-site cluster links was down for 6 days straight.  That's a longp- > time to depend on the sole surviving link.)i > > > You might think quadruple redundancy is a bit much, but manyC > disaster-tolerant clusters have 4 physical disks for every disk'sxE > worth of data (a 2-member shadowset of 2-member mirrorsets, so thatbH > you still have protection against a disk failure even after a disaster' > that destroys one entire datacenter).r >lE > Another very logical reason for four FC fabrics might be bandwidth. C > Although FC provides a theoretical bandwidth of 100 megabytes perlF > second, I can easily imagine some applications which might need four% > times that bandwidth, or even more.-A > ---------------------------------------------------------------dA > Keith Parris | parris at encompasserve dot org | Consulting on: @ > Clusters, Disaster Tolerance, Performance, I/O, Storage & SANs   ------------------------------  # Date: Fri, 11 Jan 2002 20:28:54 GMTe# From: "John Smith" <a@nonymous.com>D9 Subject: Re: HP admits it will kill VMS if merger suceedso> Message-ID: <aWH%7.167656$pa1.49810902@news3.rdc1.on.home.com>  > "John McLean" <mcleanj@swissonline.delete.ch> wrote in message/ news:3C370E10.3652965E@swissonline.delete.ch...s >t >gF > OTOH the business side of Compaq's PCs usually makes a profit, smallG > perhaps but still a profit.  The Dell purchasing model doesn't reallyeG > work that well if a company wants a whole bunch of new PCs on desks -hI > 100, 200 ... a few thousand.  This is an area that Compaq can sell into = > and one where they can get maintenance contracts and so on.  >h  F No so John.  For a large enough order, even if it is spread out over aH period of time, eg. 1 year, Dell will maintain a consistent build, rightI down to the last screw, and a consistent software image pre-loaded on thesD machines. They have to do this for companies that order thousands ofF machines when their leases expire and they want to have a single buildH across the organization. What Dell will do in this case is adjust prices- every N months that the contract is in force.n   ------------------------------  % Date: Fri, 11 Jan 2002 15:41:57 -0500,+ From: "Main, Kerry" <Kerry.Main@compaq.com>o9 Subject: RE: HP admits it will kill VMS if merger suceedsrT Message-ID: <BE56C50EA024184DAF48F0B9A47F5CF401AB1B76@kaoexc01.americas.cpqcorp.net>   JF -  D >> The mere though of having to depend on a web server to get to the: critical part of the machine doesn't sound right to me.<<<   Ahhh, but can a VT terminal:  D - keep a log file of all console activity (normal and console typing? activity) for later review and/or backups for security reasons?o  G - keep a log file for later review  of non-system console activity such E as HSC/J/G's disk controllers, network controllers (Cisco, etc) - anyi! device that uses an asci console?j  D - monitor console messages and send email/page someone if a selected event message occurs?a  B - allow a system manager to remotely gain access (via a variety ofF secure measures) to the ">>>" console prompt from a hotel room or from- home? This is great when installing software.t  D Anyway, check out the following FAQ for ConsoleWorks as I believe it; will address most of the usual system manager questions:=20b; http://www.tditx.com/services_support_faqs.html#console FAQnF http://www.tditx.com/products_consoleworks.html Console Works Overview  E Bottom line is that ConsoleWorks is only monitorring the console ascitE o/p, so if ConsoleWorks server totally died, one would just hook up asD terminal and continue until it was fixed. ConsoleWorks also offers aG disaster tolerant console solution that would also address the issue of * one ConsoleWorks server being unavailable.   Regards,  
 Kerry Main Senior Consultante Compaq Canada Corp.2 Professional Servicesg Voice: 613-592-4660) Fax  :  819-772-7036 Email: Kerry.Main@Compaq.com     -----Original Message-----4 From: JF Mezei [mailto:jfmezei.spamnot@videotron.ca] Sent: January 10, 2002 10:23 PMd To: Info-VAX@Mvb.Saic.Comr9 Subject: Re: HP admits it will kill VMS if merger suceedso     "Main, Kerry" wrote:G > Good one that is web enabled, has scan files for known error messagesaG > (including SAN HSG80 errors) and also supports OpenVMS as a server orG > client as well  E  Am I the only one to cringe at the though of a web enabled console ?o The mereF though of having to depend on a web server to get to the critical part of the" machine doesn't sound right to me.   ------------------------------  % Date: Fri, 11 Jan 2002 22:12:34 -0500n% From: JF Mezei <jfmezei@videotron.ca>n9 Subject: Re: HP admits it will kill VMS if merger suceedsy, Message-ID: <3C3FA9A2.804693A0@videotron.ca>   "Main, Kerry" wrote: > Ahhh, but can a VT terminal: > F > - keep a log file of all console activity (normal and console typingA > activity) for later review and/or backups for security reasons?u  # My PDA can do that, so can DECterm.c  I > - keep a log file for later review  of non-system console activity suchnG > as HSC/J/G's disk controllers, network controllers (Cisco, etc) - anyv# > device that uses an asci console?P  ? if you have accessto VMS, don't you have access to that stuff ?o  F > - monitor console messages and send email/page someone if a selected > event message occurs?-  B this can be done as a detached process on VMS. Much more reliable.  D > - allow a system manager to remotely gain access (via a variety ofH > secure measures) to the ">>>" console prompt from a hotel room or from/ > home? This is great when installing software.p  J The disaster tolerant cluster I had built with decsercver access to OPA0:sK allowed exactly that with passwrod protection (and gaining dialup access tov' the decservers required authenticator).o  N Furthermore, you are talking to someone who has rebooted a machine in MontrealL from a payphone in Townsville australia with a PDA and accoustic coupler and no fancy software.   ------------------------------  % Date: Fri, 11 Jan 2002 13:57:15 -0500a* From: WILLIAM WEBB <WWEBB1@email.usps.gov>7 Subject: RE: Huge EV7 Alpha Galaxy Cluster being built!t- Message-ID: <0033000047569866000002L062*@MHS>i  7 =0AWonder if there are armed guards doing a cube searchs( for Charlie Matco mugs even as we speak?   :^)y   WWWebb   -----Original Message-----/ From: Info-VAX-Request@Mvb.Saic.Com at INTERNET & Sent: Friday, January 11, 2002 1:50 PMB To: Webb, William W Raleigh, NC; Info-VAX@Mvb.Saic.Com at INTERNET7 Subject: RE: Huge EV7 Alpha Galaxy Cluster being built!-    & Interesting, but read on to the part :  ?   Terry Shannon, who runs Shannon knows Compaq, tells me that asB   512CPU Open VMS Galaxy instance is highly unlikely though.Galaxy7   APMP lets VMS and its manager define instances, usingEB   predetermined dedicated resources as well as shared memory. In aC   32-way GS320 you can carve out four eight CPU instances, he said.b  E   VMS doesn't scale at all well beyond 10 processors, he added - so ad!   512CPU instance seems unlikely.h  F I'm most interested in the last statement. I hadn't heard that, Terry.H In fact as I recall at the Diamond Forum last August, the new EV68(or 7= )o would scale well to ~64 CPUs.n   Bob Ceculski wrote:m > D > try to do this all you windoze and unix cronies w/your junk os ... >o) > http://www.theinquirer.net/10010206.htmo >u* > Compaq builds biggest Alpha cluster ever >  > Ayr on a G-String1' > By Eva Glass, 10/01/2002 11:16:49 BSTr > 8 > A LASSIE FROM Compaq in Ayr gave me a bell on my cell.F > Ayr Annie - as I'll call her - tells me that the factory up there inD > bonnie Scotland has just built and tested the largest ever cluster! > using the Alpha microprocessor.n >rE > The machine is being built for NoSuchAgency, and she reckons it's ac > real coppa whoppa. >nH > She slips me word that it could be a 1.2GHz EV7 VMS system maybe with=  D > 512 CPUs per VMS Galaxy instance and say 64 instances per cluster. > H > Hey, she said the Big Q has even been flashing photos of these babies=  F > round to its select customers who it trusts - obviously we ain't one
 > of them! >eF > Well - 32 processors per GS system was the limit with EV6 but things > are looking up now...T >cF > And in other news, Annie told me that there are some "restructuring"; > plans set to happen in Ayr - maybe as early as next week.  > B > Wouldn't it be better if Compaq talked to the INQ to prevent theB > relentless drip, drip, drip from all these lovely ladies wearing > moleskin hats?=e   ------------------------------  % Date: Fri, 11 Jan 2002 14:31:59 -0500 5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>D7 Subject: Re: Huge EV7 Alpha Galaxy Cluster being built!s2 Message-ID: <a8H%7.584$5Y4.15310@news.cpqcorp.net>  = Don Sykes wrote in message <3C3F341B.F136933E@pacbell.net>...p ><' >Interesting, but read on to the part :e > @ >  Terry Shannon, who runs Shannon knows Compaq, tells me that aC >  512CPU Open VMS Galaxy instance is highly unlikely though.Galaxyp8 >  APMP lets VMS and its manager define instances, usingC >  predetermined dedicated resources as well as shared memory. In a D >  32-way GS320 you can carve out four eight CPU instances, he said. >aF >  VMS doesn't scale at all well beyond 10 processors, he added - so a" >  512CPU instance seems unlikely. ><G >I'm most interested in the last statement. I hadn't heard that, Terry.gI >In fact as I recall at the Diamond Forum last August, the new EV68(or 7). >would scale well to ~64 CPUs. >g    H Let's say that we won't be able to test that question until V8.0 of VMS.  F It is also one of those "it depends" questions.  Some applications andK workloads scale, and some don't.  I believe that VMS has a little way to gocL to scale "well" beyond 16 CPUs in general workloads (however, the results ofG the biggest EV7 system we are running here in the lab - a 16p - show it  scales *well* to 16).e  E How well things scale depends on the HW, the OS, and the application.b   ------------------------------  % Date: Fri, 11 Jan 2002 16:27:36 -0500u% From: JF Mezei <jfmezei@videotron.ca>c7 Subject: Re: Huge EV7 Alpha Galaxy Cluster being built! , Message-ID: <3C3F58C3.FE2969D3@videotron.ca>  G >   VMS doesn't scale at all well beyond 10 processors, he added - so an# >   512CPU instance seems unlikely.u  M But if the big expensive box is to run Tru64 or Linux, how would they scale ?c  N But a 512 processor box could sill run 51 "nodes" in a cluster, each having 10J CPUs with the last one having 12 CPUs. And that is still well under the 96' node limit. Or 64 nodes of 8 CPUs each.-   ------------------------------    Date: 11 Jan 2002 13:07:30 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen)o/ Subject: Re: Inquirer take on Ayr Alpha monster$3 Message-ID: <yZkmZZWrCtle@eisner.encompasserve.org>e  u In article <VlF%7.66363$Sj1.27350481@typhoon.ne.mediaone.net>, "Terry C. Shannon" <terryshannon@mediaone.net> writes:3 > B > "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> wrote in message. > news:5AE%7.568$5Y4.15227@news.cpqcorp.net...K >> One modification to what I said.  I believe that a Marvel with more thanr > 32J >> processors would need to use hard partitions, not soft partitions.  VMSK >> needs the CPU namespace to be 0-31.  That will not change until we go todM >> V8.0.  So I believe you would hard partition at the 32p boundry, and couldn@ >> run multiple Galaxy partitions within the 32p hard partition. >>J >> Of course, my contact in the gov space says the report is out to lunch. > M > I'd go with the Government contact in this regard. But gee, it sure is nicee6 > to know that there will be an OpenVMS Version 8! ;-}  F Some years ago, the chief VMS Product Manager told a DECUS crowd thereF had been so much fuss from customers about going to 7.0 that he wanted' there never to be a release called 8.0.   5 Luckily he moved to another job, with less influence.s  J Unluckily he subsequently moved to yet _another_ job, with more influence.  ) Luckily, I think Fred can beat him up :-)    ------------------------------  % Date: Fri, 11 Jan 2002 14:46:03 -050025 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>h/ Subject: Re: Inquirer take on Ayr Alpha monster 2 Message-ID: <mlH%7.587$5Y4.15381@news.cpqcorp.net>  $ Larry Kilgallen wrote in message ...I >In article <VlF%7.66363$Sj1.27350481@typhoon.ne.mediaone.net>, "Terry C.1, Shannon" <terryshannon@mediaone.net> writes: >>C >> "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> wrote in messageo/ >> news:5AE%7.568$5Y4.15227@news.cpqcorp.net...cL >>> One modification to what I said.  I believe that a Marvel with more than >> 32tK >>> processors would need to use hard partitions, not soft partitions.  VMSfL >>> needs the CPU namespace to be 0-31.  That will not change until we go toH >>> V8.0.  So I believe you would hard partition at the 32p boundry, and couldiA >>> run multiple Galaxy partitions within the 32p hard partition.  >>>sK >>> Of course, my contact in the gov space says the report is out to lunch.  >>I >> I'd go with the Government contact in this regard. But gee, it sure iso nice7 >> to know that there will be an OpenVMS Version 8! ;-}  > G >Some years ago, the chief VMS Product Manager told a DECUS crowd therecG >had been so much fuss from customers about going to 7.0 that he wanted9( >there never to be a release called 8.0. >r6 >Luckily he moved to another job, with less influence. > K >Unluckily he subsequently moved to yet _another_ job, with more influence., >a* >Luckily, I think Fred can beat him up :-)  J Dunno who this is, but I haven't actually "beat someone up" since I was 17 (and he picked the fight).  F In any case, as you get more mature, and have things you don't want toH break - you need to take care when rolling a major release.  I too thinkD this is something that you want to minimize or never do if possible.  G I'll give you some recent thinking here...  The release of Itanium willaI probably *not* be V8.0.  It will probably be a V7.x version number.  ThataC is, a minor release.  We are not planning on breaking any privlegedsI interfaces or datastructures on VAX or Alpha.  Yeah, you'll have to buildaK from sources for Itanium (or use the image translation stuff) but there are + no binary compatability issues for Itanium.o  J We are now working on exactly when a V8.0 release should be done, and whatL should be in it.  Things that break binary compatability for instance - like9 > 32 CPUs, or LBN's > 32.  And major new functionality...0  = I keep telling you guys, we aren't *just* working on Itanium.c   ------------------------------  % Date: Fri, 11 Jan 2002 16:04:39 -0500s2 From: "Sue Skonetski" <susan.skonetski@compaq.com>% Subject: Invitation to Compaq webcaste2 Message-ID: <rqI%7.589$5Y4.15441@news.cpqcorp.net>  ( Invitation to Compaq / Attunity Webcast: Opening BASEstar to the WebnD Earlier this past year, the Compaq OpenVMS organization held severalJ customer forums regarding e-business and the use of web-based technologiesF such as XML, JDBC, Java, etc. with existing / legacy applications. TheI strong feedback received was that while many of the existing applicationstL are the core foundation that the function or business are build on, the need@ to "open up" these applications and web-enable them is critical.  L One of the follow-up actions from the OpenVMS Manufacturing Industries forumF was to investigate the use of Attunity, a Compaq partner in e-businessH software, and a Compaq manufacturing middleware product called BASEstar.E BASEstar is a real-time factory floor middleware product used by many K world-wide manufactures in both discrete and process industries. One of thehD needs discussed in the customer forum was making it easy to create aL web-based user interface for BASEstar so that factory floor and IT personnel@ could monitor manufacturing operations via browsers and wirelessF technologies such as Personal Data Assistants (PDA), cell phones, etc.J Attunity, working with Compaq, created a proof of concept of just how easyE it is to take existing applications and web enable them. Attunity hasdK created a technical white paper (attached) describing the approach and willaE present their company, product capabilities, and describe, as well as.G demonstrate, the opening up of existing applications to standards based  interfaces.r   Intended Audience:  D   a.. IT professionals looking to see how to bring existing / legacy, applications into the 21st century, - easilyI   b.. IT professionals or Plant managers in the manufacturing industry ort@ other industries where data is needed by users wherever they areH   c.. Customers who have or maintain OpenVMS, Tru64 or Microsoft Windows applicationsL   d.. System Integration partners looking for tools to increase productivity2 through web-based and standards based technologies	 Schedule:b   Wednesday January 30th       a.. 9:00 am - 10:00 am EST     b.. 9:00 pm - 10:00 pm EST8 The web casts are being opened to a world-wide audience.  I The Web, Java, and the new standards have changed the way people work and|' the way new applications are developed.t  J With the products available from Attunity, you can easily get the opennessJ and productivity gains from your existing applications without introducingK any changes to them. I would strongly suggest registering for this web castm7 to see and hear how to take advantage of this approach.w   How to Register:  L   email your name, daytime phone number, and which session you would like toL attend to BASEstar@attunity.com.  There are a limited number of openings andC they will be filled on a first come, first severed basis.  An email 9 confirmation will be sent to you.  Please register early.l   Regards,  
 John C. Egolfu  ( Manufacturing Industries Segment Manager
 OpenVMS Group  Compaq, Nashua, NH   ------------------------------    Date: 11 Jan 2002 15:13:16 -0800$ From: JKoska@bender.com (John Koska)1 Subject: Re: OpenVMS Disk Services for Windows NTv= Message-ID: <aa5ec19e.0201111513.512e92ed@posting.google.com>n  D I believe OpenVMS Disk Services for Windows NT (NTDS) is part of theE Enterprise Integration Packages also.  Specifically, prior to releaser 3.2 from September 2001.  = A possible important note regarding this software... It is mycC understanding that NTDS will NOT be qualified with OpenVMS 7.3, andsF NTDS has been removed from the Enterprise Integration License Packages! with release 3.2, September 2001.2   :) jck 0  E My views are my own, and do not necessarily represent my employer(s).s  v Fabio Cardoso <fabiopenvms@yahoo.com.br> wrote in message news:<20020109123051.98641.qmail@web20201.mail.yahoo.com>...0 > I never worked with this product but I believe* > it should be under SANworks portfolio of > products.i > 	 > Regardsd >  > FC e > # > --- Mark Iline - Info-VAX account  > <ivax@meng.ucl.ac.uk> wrote:7 > > I've just been looking at the SPD for 'OpenVMS Diske > > Services for Windows NT' y7 > > ( http://www.compaq.com/info/SP6080/SP6080PF.PDF ),s > > and it's for version  5 > > 1.1 that only refers to NT 4 & OpenVMS 7.1 & 6.2.i > > % > > Has this product got any future ?n > > 7 > > Has anyone tried it with versions of VMS beyond 7.1e > > ?e > >  > >  > > Mark > >  > > % > > Mark Iline	system@meng.ucl. ac.uki1 > > Dept Mech Eng, University College, London. UK  >  >  > =====m > =========================  > =  > F?io dos Santos Cardoson > OpenVMS System Manager > Rio de Janeiro - Brazilo > fabiopenvms@yahoo.com.br > =========================u > =s > 4 >                                                    > Do You Yahoo!?( > Send FREE video emails in Yahoo! Mail!# > http://promo.yahoo.com/videomail/v   ------------------------------    Date: 11 Jan 2002 16:34:13 -0800. From: SPAMSINK2001@YAHOO.COM (Alan E. Feldman)" Subject: Re: Re[2]: DCL whish list< Message-ID: <343f30ae.0201111634.497be19@posting.google.com>  o Valentin Likoum <valentin.likoum@ncc.volga.ru> wrote in message news:<298320601.20020111102912@ncc.volga.ru>...v? > On 10.01.2002 Alan E. Feldman <SPAMSINK2001@YAHOO.COM> wrote:n > t > > Valentin Likoum <valentin.likoum@ncc.volga.ru> wrote in message news:<1829801283.20020110105652@ncc.volga.ru>... [snip]G > >>   While we are on the topic, why not to show PID of the process inaB > >> SHOW QUEUE and SHOW ENTRY if job is executing. Something like > >> s9 > >>   Entry  Jobname         Username             Statust9 > >>   -----  -------         --------             ------nK > >>     812  BCK             B_USER               Executing (PID:000046D1) K > >>                                                         ^^^^^^^^^^^^^^oF > >>   I'm a bit lazy to search through SHOW SYSTEM output to find outK > >> in which process certain entry is executing every time I need to checkm > >> how this entry is doing.- > >>   Thank you.1 >   H > > This command procedure will get you the pid and other info about the > > job. > [procedure snipped]r > D >   Thank you, Alan. It's very useful. But when user comes to me and@ > asks "What the hell is my job doing?", it's still a three-stepG > procedure: 1. SHOW QUEUE to figure out which entry his job is; 2. SHO H > SYS or your procedure to figure out process PID; 3. SHO PROC or SDA toG > dig into. My point was to make this procedure two-step (skipping step B > 2). IMHO it doesn't cost much (but may be backward compatibility
 > issues?) >   Thank you.  F Your welcome. Actually, you *can* use SHO PROC with my procedure. Just run something like   $ BAT 123 /ACC/QUOTA !  B You can even use /CONTINUOUS. Be sure to leave a space between theB entry number and the qualifiers, and no spaces between qualifiers.  F For SDA, you can probably modify my procedure by adding something that# writes and runs a command file like>  $ $ define /user sys$input sys$command $    analyze/system  set proc /index=<pid>o  D You have to make my .com file write the above .com file with the pidB already substituted in because you can't do symbol substitution inA data lies. With a little work, it can do pretty close to what you  want, I think.   Disclaimer: JMHO           a Alan E. Feldmani afeldman;gfigroup.comp   ------------------------------  % Date: Fri, 11 Jan 2002 16:20:24 -0500n% From: JF Mezei <jfmezei@videotron.ca>-D Subject: Re: Steve Jobs not affraid to sell his "different" products, Message-ID: <3C3F5714.A39A0821@videotron.ca>  B > >UDF, JFS, ext2fs, etc.  Sticking to ODS-2/ODS-5 may be good forF > >clustering, but let's say I want a better (faster) file system thatK > >doesn't necessarily have to be clustered.  Currently, its more difficult 8 > >under VMS because I have to do a driver AND an ACP.    L But can't you mount "foreign"  drives through NFS and have each drive servedB by its native OS much better than any OS can serve foreign disks ?   ------------------------------  % Date: Sat, 12 Jan 2002 11:37:11 +1300r# From: Bruce Hoult <bruce@hoult.org>pW Subject: Re: Sv: Sv: Sv: Younger recruits versus experienced veterans ( was Re: The dem > Message-ID: <bruce-2955B1.11371112012002@news.paradise.net.nz>   In article aH <bwF%7.245916$WW.13248185@bgtnsc05-news.ops.worldnet.att.net>, "Stephen * Fuld" <s.fuld.pleaseremove@att.net> wrote:  ; > "Steve O'Hara-Smith" <steveo@eircom.net> wrote in message-3 > news:20020111072003.728d6ab2.steveo@eircom.net...t& > > On Fri, 11 Jan 2002 14:01:18 +1300( > > Bruce Hoult <bruce@hoult.org> wrote: > >t > >aK > > BH> Why is it that two (or three?) mouse buttons is the *right* number,n > but J > > BH> a somewhat larger number of possible modifier key combinations is  > > toos$ > > BH> many?  Where is the optimum? > >0C > > Simple, I can *see* the mouse buttons and *know* that they havenJ > > something to do with pointer related actions. Three buttons is optimalI > > because that's how many fingers I have available for pressing buttonst6 > > while the thumb and little finger guide the mouse. > K > One thing that hasn't been mentioned in this thread is the prevalence (onsB > Windows mice at least) of some kind of scroll device (a wheel orJ > occasionally rocker switch).  I find these to be usefull in that I don't@ > have to move the mouse to the correct spot on the window, etc.  4 Those mice work fine on Macs, too.  If you want one.   -- Bruce   ------------------------------  % Date: Fri, 11 Jan 2002 14:01:41 -0800c' From: David Mathog <mathog@caltech.edu>a$ Subject: Re: Unix filesystems on VMS+ Message-ID: <3C3F60C5.40441691@caltech.edu>V   Fred Kleinsorge wrote: >   L > We of course will have this very issue with Itanium.  There will be a fileL > on the ODS volume that will actually contain a FAT partition.  The "files"N > inside this container file will not be visible to standard VMS utilities andM > applications.  A seperate utility will allow files to be copied into/out of-; > the container, and to be directoried, deleted, renamed...0  L This sort of thing has been done on VMS through these sorts of kludgy access tools forever.H COPY in, COPY out is a pain even when done well (to varying extents with exchange, ftp,P PCX, smbclient etc.). At best you could sometimes get DECNET access to a foreign fileL system on a remote machine (as with Pathworks/Mac) and relatively normal DCL level interactionGA with a remote file system that way.  I really think it would be aaJ wonderful feature if there was a supported way to mount an abritrary local deviceP (like a CD) or file (a container) or even a remote device (smb,nfs, etc) in such a < way that it appeared to be, to the greatest extent possible,N a normal VMS file system (accessed through a logical).  Sure, it might only be able tohP represent files as CRLF (and I do hope that the long standing problems with that	 file typetJ get resolved first, ie SORT) but that would be more than adequate for many uses.  w  , For the FAT case, maybe something like this:  4   $ mount/foreign/format=FAT/parameters=(???) MYFAT:   $ set def MYFAT:[000000]   $ dir 
   autoxec.bat 
   vms.boot   $ type vms.boota   $ edit vms.boot    $ dismount MYFAT  / and exactly the same symantics for the SMB caset     $ mount/foreign/format=FAT -K      /parameters=(remote:"\\machine\$C",user:"domain\fred",password:"blah")e MYFAT:   $ set def MYFAT:[000000]   $ dird
   windows.dirn
   temp.dir   (etc.)   $ set def [.windows]   $ delete worthless_os.*;*      $ dismount MYFAT  N Not that the VMS folks need to do all the work.  If Compaq could supply an API so that userF written code need only provide a fixed set of low level routines (list directory, open file andH move blocks back and forth) delete, rename, get file info, etc.) then it wouldn't be such a longuJ stretch getting from the code for an ftp client or smbclient to direct DCL access to these other file  system types (local or foreign.)   Regards,   David Mathog mathog@caltech.edu   ------------------------------  # Date: Fri, 11 Jan 2002 23:04:29 GMT-2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)$ Subject: Re: Unix filesystems on VMS2 Message-ID: <1cK%7.594$5Y4.15477@news.cpqcorp.net>  U In article <3C3F60C5.40441691@caltech.edu>, David Mathog <mathog@caltech.edu> writes:p  I :COPY in, COPY out is a pain even when done well (to varying extents with3M :exchange, ftp, PCX, smbclient etc.). At best you could sometimes get DECNET sM :access to a foreign file system on a remote machine (as with Pathworks/Mac) IL :and relatively normal DCL level interaction with a remote file system that N :way.  I really think it would be a wonderful feature if there was a supportedM :way to mount an abritrary local device (like a CD) or file (a container) or  N :even a remote device (smb,nfs, etc) in such a way that it appeared to be, to L :the greatest extent possible, a normal VMS file system (accessed through a M :logical).  Sure, it might only be able to represent files as CRLF (and I do rM :hope that the long standing problems with that file type get resolved first,o? :ie SORT) but that would be more than adequate for many uses.  D  E   The closest existing stuff is the foreign file FDL-based converters F   and the TCP/IP Services ADF stuff -- the Attribute Description File +   (ADF) stuff lurks within the NFS client.    J   This stuff is exceedingly ugly, layering a big kludge (ADF- or FDL-basedI   conversions) onto a bigger kludge (determining the record format based -=   on the file extension).  It gets really ugly really fast...r  D   The FAT-related work -- probably an update to the OpenVMS EXCHANGEE   utility -- that is presently underway is not a general-purpose diskrF   interface for FAT disks.  It very likely will not provide a generic F   record structure support interface for FAT files.  It will probably -   not be documented for general use, as well.e   :...L :Not that the VMS folks need to do all the work.  If Compaq could supply an J :API so that user written code need only provide a fixed set of low level M :routines (list directory, open file and move blocks back and forth) delete, nN :rename, get file info, etc.) then it wouldn't be such a long stretch getting L :from the code for an ftp client or smbclient to direct DCL access to these , :other file system types (local or foreign.)  E   This was a "would be nice to have" bullet on the slides I used for gH   an (internal) meeting last Wednesday, and there are definite uses for H   this API and the equivilent API within MOUNT, but there are presently K   NO plans and NO schedule time available to add the necessary connections o   and support for this API.     D   Plugging code into XQP and MOUNT and RMS is an "interesting" task.  G   Could I see uses for this API?  Sure.  I discussed just a few of the oG   potential uses at the (internal) meeting mentioned earlier.  It wouldtH   certainly make one of my current projects easier, and I could see some1   very elegant possibilities for some other work.t  I   The addition of the API is not presently part of nor is it planned for a   the IA-64 port of OpenVMS.      N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Fri, 11 Jan 2002 21:55:34 -0600e, From: "Rich Jordan" <rjordan@mindspring.com>$ Subject: Re: Unix filesystems on VMS3 Message-ID: <a1oc2r$5jg$1@nntp9.atl.mindspring.net>u  I Make it mountable as a 'pseudo-console' device.  Its been so long since IwL worked with one on the old 11/785s that I don't remember the details; SYSGENL connect console, then methods of access... make it feel like home even if it$ is a foreign dos-format partition...   Rich Jordane  $ Fred Kleinsorge wrote in message ... >n >nK >We of course will have this very issue with Itanium.  There will be a file3K >on the ODS volume that will actually contain a FAT partition.  The "files"tI >inside this container file will not be visible to standard VMS utilities  andiL >applications.  A seperate utility will allow files to be copied into/out of: >the container, and to be directoried, deleted, renamed... >t >r >r >    ------------------------------  # Date: Fri, 11 Jan 2002 19:38:08 GMTJ2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) Subject: Re: VMS ECOse2 Message-ID: <AaH%7.585$5Y4.15292@news.cpqcorp.net>  b In article <VBC%7.11856$_p.5269554@typhoon.tampabay.rr.com>, "John N." <JNixon@cfl.rr.com> writes:E :What is the best way to get a CURRENT list of ECOs for (eg) VMS 7.3.a :oI :I found the ECO directory tree, but it only lists part of the names, andu :there are hundreds of them.    H   See the OpenVMS FAQ for related information, and an ECO search engine.  J :                          Several files for each ECO and several versionsI :of each ECO.  I would like something like the old  "need to apply" listseJ :that were kept updated on DSN.  Even they suck, but they were better than. :just a list of hundreds of partial filenames.  I   Those documents are available via http://askq.compaq.com/, and via the b0   folks at the Compaq Customer Support Center.    G   In the case of the askq website -- which serves up most of the CompaqcE   support database -- visit the URL, select the OpenVMS database, andi+   then enter a query such as the following:r  I "Which ECO kits are incorporated in V7.3 and which do I need to install?"r  G   I don't specifically know that the V7.3 version of this "traditional"s#   article is available quite yet...l  L :What I really want is a tool that I can run on a system and it will tell meJ :every current ECO that I have not yet applied and in what order I need to( :apply them (including severity levels).  H   The Compaq Services folks offer an ECO inventory and related services.?   "ProPatch" is one of the names used to identify this service.r  K   The OpenVMS FAQ has some details on this topic (and an update queued for  H   the next edition, to reflect some newly-available details), of course.  N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------    Date: 11 Jan 2002 12:06:11 -0800, From: jeremybarker@email.com (Jeremy Barker)H Subject: Re: VMS on more common hardware (was Re: VMS as a career move )= Message-ID: <5b86b9ee.0201111206.6c71f856@posting.google.com>d  a JF Mezei <jfmezei.spamnot@videotron.ca> wrote in message news:<3C3A146D.9BF412F9@videotron.ca>...e > Bob Koehler wrote:G > >    Back to reality.  A great bit of VMS has since been rewritten iniF > >    BLISS or other HLL.  I know of VAX Macro code fomr VMS 2.x that > >    became BLISS by VMS 4.0.a >  > O > When writing kernel code or deep down stuff for the OS, does using a compileraP > with lots of optimization for the harware ever cause serious problems ?  (I amN > thinking of instruction ordering, and variable "elimination" by keeping them1 > in registers, and possible many other factors).s  F In BLISS you can declare a variable to be VOLATILE so it isn't kept inC a register and, IIRC, is always fetched from its memory address forr	 each use.n  F I wrote power-on boot code for a controller board based on an embeddedD MIPS processor mostly in BLISS and as far as I can remember (this isD going back about 10 years) it needed almost no assembly code.  I canC also remember having a hell of an argument about using BLISS ratheroC than C.  I won when I showed the BLISS code used far less ROM space  than C.a  & jb (ex-DEC, who's a lawyer these days)   ------------------------------    Date: 11 Jan 2002 16:43:44 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen)0H Subject: Re: VMS on more common hardware (was Re: VMS as a career move )3 Message-ID: <J0IxaskFxAGq@eisner.encompasserve.org>s  l In article <5b86b9ee.0201111206.6c71f856@posting.google.com>, jeremybarker@email.com (Jeremy Barker) writes:c > JF Mezei <jfmezei.spamnot@videotron.ca> wrote in message news:<3C3A146D.9BF412F9@videotron.ca>...c >> Bob Koehler wrote:PH >> >    Back to reality.  A great bit of VMS has since been rewritten inG >> >    BLISS or other HLL.  I know of VAX Macro code fomr VMS 2.x that   >> >    became BLISS by VMS 4.0. >> n >> VP >> When writing kernel code or deep down stuff for the OS, does using a compilerQ >> with lots of optimization for the harware ever cause serious problems ?  (I ammO >> thinking of instruction ordering, and variable "elimination" by keeping themm2 >> in registers, and possible many other factors). > H > In BLISS you can declare a variable to be VOLATILE so it isn't kept inE > a register and, IIRC, is always fetched from its memory address fors > each use.i  F Presumably for the same reasons, pragma Volatile was a vendor-specific+ add-on when DEC wrote the VAX Ada compiler.P  = Of course the problem it addresses is not specific to VMS, sohB pragma Volatile was added to the Ada95 standard for all compilers.   ------------------------------    Date: 11 Jan 2002 18:03:48 -08003 From: Eric Smith <eric-no-spam-for-me@brouhaha.com>nH Subject: Re: VMS on more common hardware (was Re: VMS as a career move )0 Message-ID: <qhsn9c9vgb.fsf@ruckus.brouhaha.com>  / Kilgallen@SpamCop.net (Larry Kilgallen) writes:.H > Presumably for the same reasons, pragma Volatile was a vendor-specific- > add-on when DEC wrote the VAX Ada compiler.i > ? > Of course the problem it addresses is not specific to VMS, so D > pragma Volatile was added to the Ada95 standard for all compilers.  A They added a pragma with behavior that is defined by the standard C and required of all implementations?  That's terrible!  Pragmas are E supposed to be implementation specific, and anything that it standard  should NOT be a pragma.  Sigh.   ------------------------------    Date: 11 Jan 2002 22:00:06 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen) H Subject: Re: VMS on more common hardware (was Re: VMS as a career move )3 Message-ID: <Ke3pUamZhJF8@eisner.encompasserve.org>p  f In article <qhsn9c9vgb.fsf@ruckus.brouhaha.com>, Eric Smith <eric-no-spam-for-me@brouhaha.com> writes:1 > Kilgallen@SpamCop.net (Larry Kilgallen) writes::I >> Presumably for the same reasons, pragma Volatile was a vendor-specific . >> add-on when DEC wrote the VAX Ada compiler. >> D@ >> Of course the problem it addresses is not specific to VMS, soE >> pragma Volatile was added to the Ada95 standard for all compilers.a > C > They added a pragma with behavior that is defined by the standard E > and required of all implementations?  That's terrible!  Pragmas areAG > supposed to be implementation specific, and anything that it standarde  > should NOT be a pragma.  Sigh.  = What makes you say that ?  The 1981 Oxford English Dictionarya9 has no such word as "pragma", but looking at the standardd< ( http://www.adapower.com/rm95/arm95_31.html#SEC31 ) we see:  E "A pragma is a compiler directive. There are language-defined pragmas B  that give instructions for optimization, listing control, etc. AnI  implementation may support additional (implementation-defined) pragmas."w   ------------------------------    Date: 11 Jan 2002 15:26:51 -0600G From: simon_clubley@remove_me.altavista.co.uk-Earth.UFP (Simon Clubley)- Subject: Re: VMSBackup to unix3 Message-ID: <yWRAVNR5VwAP@eisner.encompasserve.org>e  U In article <3C3F057F.7000206@mmaz.com>, "Barry Treahy, Jr." <Treahy@mmaz.com> writes:e > K > and lastly, if you still want to do it 'on-the-cheap' you can attempt to  J > use VMSBACKUP from: http://www.panix.com/~kingdon/vms/backup.html which ' > is an incomplete clone but is free...c >   F Assuming that he can get Unix to read an ANSI tape and get the savesetH onto disk, then I have had success with READVMSBACKUP (on the sig tapes) in the past.  I BTW, just how does Unix handle ANSI tapes ? This is one area of Unix thatc" I haven't had any exposure to yet.   Simon.   -- -G Simon Clubley, simon_clubley@remove_me.altavista.co.uk-Earth.UFP       h+ Microsoft: The Lada of the computing world.r   ------------------------------  # Date: Fri, 11 Jan 2002 22:49:01 GMTe' From: nickerson@pundit.ds.boeing.com ()n Subject: Re: VMSBackup to unix( Message-ID: <Gpsq1p.CrG@news.boeing.com>  / In article <a1lgvr$258$1@woodrow.ucdavis.edu>, n* "Ben Ransom" <bransom@ucdavis.edu> writes:H |>Anyone know of a utility or unix "dd" type of method to extract from aK |>VMSBackup set onto a unix system?  I've tried dd with some skip n blocks,u3 |>but that is wild guessing and produces only glop.o |> |>Here's what I get:0 |># dd if=/dev/rmt/tapenr  of=decstuff  ibs=8192 |>0+3 records in |>0+1 records out  |>#o |># cat decstuffE |>VOL1DK1_F                 D%C          1          3HDR105JUL94F.DK1sL |>DK1_F 00010001000100 94187 94187 000000DECVMSBACKUP        HDR2F0819208192 |>M             00 |>#e  C well folks have pointed you to better tools and better use of your bD valuable time; I'll chime in by showing the reformatted headers (???H must be missing some blanks lost in transmission) and some URLs; notice A all the interesting data in the columns below - like the obvious  D 4 character VOL1,HDR1,HDR2; looks like the tape file id is from the F date, way back on 05 Jul 1994 (94187 looks equivalent but I calculate ? it to be day 186 and it should be in columns 41-46 and 47-52); w7 F in header2 showed fixed 01892 08192 records & blocks;  & etc ...; fun stuff;1    K 0        1         2         3         4         5         6         7     yK 123456789012345678901234567890123456789012345678901234567890123456789012345   3 VOL1DK1_F                 D%C          1          3wK HDR105JUL94F.DK1DK1_F 00010001000100 94187 94187 000000DECVMSBACKUP        n HDR2F0819208192M             00"   references: 9 http://www.loc.gov/marc/specifications/specexchtape1.htmlsD http://www.loc.gov/marc/specifications/specexchtape2.html - pre 1977  > |>TIA, and sorry for mentioning the u word on a vms newsgroup.
 |>-Ben Ransomh |> UC Davisi   --bn (Bart Nickerson)h nickerson@pundit.ds.boeing.com (206) 662-0183   ------------------------------  % Date: Fri, 11 Jan 2002 15:19:34 -0500h- From: "Richard D. Piccard" <piccard@ohio.edu> , Subject: Re: what is future of VMS and alpha( Message-ID: <3C3F48C4.A7C1019D@ohio.edu>  E Actually, in educational institutions the DR timescales are sometimesaD quite tight:  even one day missed out of a ten-week quarter can be aH significant handicap, and when running on-line registration for the nextH quarter, the students are informed of the time when they will be allowedE to register, in small batches every fifteen minutes, so that the peaklH load on the system  (and the phone company) is below saturation.  Even aH ten-minutes outage can have major consequences (we did once schedule tooH many in one hour, with the result that local residents in town could notE get a dial tone!).  This has provided much of the justification for a F diesel generator capable of handling the machines and air conditionersB in our computer center, combined with a battery-backup UPS to ride- through the start-up cycle of the generator.    	 						RDPt     Rob Young wrote: >  [snip]M >         Disaster Tolerance) is fairly straightforward and/or easy.  Can youeO >         say the same Ginny?  Guess:  .edu ... DR?  "We don't need no stinkingA$ >         DR, we aren't real-world!" -- pB ==================================================================B Dick Piccard                           Academic Technology ManagerB piccard@ohio.edu                                 Computer ServicesB http://oak.cats.ohiou.edu/~piccard/                Ohio University   ------------------------------    Date: 11 Jan 2002 15:29:49 -0600+ From: young_r@encompasserve.org (Rob Young)t, Subject: Re: what is future of VMS and alpha3 Message-ID: <DEJjeAl4U8B3@eisner.encompasserve.org>e  X In article <3C3F48C4.A7C1019D@ohio.edu>, "Richard D. Piccard" <piccard@ohio.edu> writes: > G > Actually, in educational institutions the DR timescales are sometimesaF > quite tight:  even one day missed out of a ten-week quarter can be aJ > significant handicap, and when running on-line registration for the nextJ > quarter, the students are informed of the time when they will be allowedG > to register, in small batches every fifteen minutes, so that the peak J > load on the system  (and the phone company) is below saturation.  Even aJ > ten-minutes outage can have major consequences (we did once schedule tooJ > many in one hour, with the result that local residents in town could notG > get a dial tone!).  This has provided much of the justification for awH > diesel generator capable of handling the machines and air conditionersD > in our computer center, combined with a battery-backup UPS to ride/ > through the start-up cycle of the generator. e >    	Oh?   	DR != High Availability. . .   : 	Would not wish this on anybody, but being in Ohio, do you% 	recall anything about April 3, 1974?h  + http://www.interaxs.net/pub/hgr/tornado.htmp   				Robv   >  >  > Rob Young wrote: >> d > [snip]N >>         Disaster Tolerance) is fairly straightforward and/or easy.  Can youP >>         say the same Ginny?  Guess:  .edu ... DR?  "We don't need no stinking% >>         DR, we aren't real-world!"h > -- ND > ==================================================================D > Dick Piccard                           Academic Technology ManagerD > piccard@ohio.edu                                 Computer ServicesD > http://oak.cats.ohiou.edu/~piccard/                Ohio University   ------------------------------  + Date: Sat, 12 Jan 2002 00:15:21 +0000 (UTC)p From: david20@alpha2.mdx.ac.uk, Subject: Re: what is future of VMS and alpha+ Message-ID: <a1nv6p$ev3$1@aquila.mdx.ac.uk>f  X In article <3C3F48C4.A7C1019D@ohio.edu>, "Richard D. Piccard" <piccard@ohio.edu> writes: >tF >Actually, in educational institutions the DR timescales are sometimesE >quite tight:  even one day missed out of a ten-week quarter can be aeI >significant handicap, and when running on-line registration for the nextlI >quarter, the students are informed of the time when they will be allowedcF >to register, in small batches every fifteen minutes, so that the peakI >load on the system  (and the phone company) is below saturation.  Even asI >ten-minutes outage can have major consequences (we did once schedule toolI >many in one hour, with the result that local residents in town could nottF >get a dial tone!).  This has provided much of the justification for aG >diesel generator capable of handling the machines and air conditionerspC >in our computer center, combined with a battery-backup UPS to ride . >through the start-up cycle of the generator.  >i
 >						RDP >I  A For a number of years now we have been setting up a backup of oureE corporate student system down at Docklands in London during clearing.pJ (This is the time after the A-Level results when Students, who haven't forF instance quite got the grades they need, are ringing around to find a  University place).I This time is so crucial to us that it is worth having a system setup with<K regular automatic downloads of Oracle data from the live systems so that inOA the event of a disaster we would lose at most half an hours data.-< The site also has a hundred desktop systems with telephones.  J On the administrative side disaster recovery is as crucial to Educational O Institutions as it is for most large organisations. On the Academic side thingsoN may not be quite as time critical although any prolonged outages affecting theK teaching program will affect our reputation and in extremis could result inc+ disaffected students sueing the University.o  
 David Webb VMS and Unix team leader CCSS Middlesex University   >N >Rob Young wrote:t >> e >[snip]aN >>         Disaster Tolerance) is fairly straightforward and/or easy.  Can youP >>         say the same Ginny?  Guess:  .edu ... DR?  "We don't need no stinking% >>         DR, we aren't real-world!"T >-- C >==================================================================aC >Dick Piccard                           Academic Technology ManagerlC >piccard@ohio.edu                                 Computer ServicesyC >http://oak.cats.ohiou.edu/~piccard/                Ohio University    ------------------------------  % Date: Fri, 11 Jan 2002 15:51:28 -0500-0 From: "warren sander" <warren.sander@compaq.com>+ Subject: Re: who else got the VMS brochure?a2 Message-ID: <lfI%7.588$5Y4.15373@news.cpqcorp.net>  : For those snail mail impaired the brochure is available at  K http://www.openvms.compaq.com/openvms/brochures/openvms_brochure/index.htmle and L http://www.openvms.compaq.com/openvms/brochures/openvms_brochure/openvms_bro	 chure.pdf,    it's been up since November 30th   -warrene --B ------------------------------------------------------------------6 Warren Sander                        OpenVMS MarketingK Compaq Computer Corporation          Work:  warren.sander@remove.compaq.comlL 200 Forest Street MR01-3/J1          Personal: sander@remove.ma.ultranet.com3 Marlboro, MA 01752                   (508) 467-4875s5    My opinions are my own and I only speak for myselfo,          Read http://www.openvms.compaq.com/B ------------------------------------------------------------------    ? "Terry C. Shannon" <terryshannon@mediaone.net> wrote in messageC8 news:S7q%7.65288$Sj1.26824660@typhoon.ne.mediaone.net... >M5 > "Jan-Erik Sderholm" <aaa@aaa.com> wrote in messageo# > news:3C3E068C.47AE3314@aaa.com...r? > > I got both the umbrella and the brochure a few weeks beforef. > > Christmas. Both delivered to me in Sweden. >nA > Nothing has materialized in Ashland, the Toxic Waste Capital ofs > Taxachusetts!n >n >    ------------------------------  # Date: Sat, 12 Jan 2002 03:14:36 GMTi1 From: "David J. Dachtera" <djesys.nospam@fsi.net>i+ Subject: Re: who else got the VMS brochure? ' Message-ID: <3C3FAAF5.AAE3D0C4@fsi.net>s  F Sorry - this post is for the sole purpose of attempting to consolidate the wrapped URL.   warren sander wrote: > < > For those snail mail impaired the brochure is available at > M > http://www.openvms.compaq.com/openvms/brochures/openvms_brochure/index.htmlc > andtW > http://www.openvms.compaq.com/openvms/brochures/openvms_brochure/openvms_brochure.pdf    -- n David J. Dachtera  dba DJE Systems- http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/i   ------------------------------  # Date: Sat, 12 Jan 2002 03:33:08 GMT ' From: "C.W.Holeman II" <cwhii@mail.com>e+ Subject: Re: who else got the VMS brochure?:( Message-ID: <3C3FAE6B.4E42490E@mail.com>   warren sander wrote: > < > For those snail mail impaired the brochure is available at > M > http://www.openvms.compaq.com/openvms/brochures/openvms_brochure/index.htmld   Which states that:  F   Studies show that the average company loses half its customers every   five years.   G   It costs five to 10 times as much to get a new customer as it does to    keep one.   D   Keeping customers is becoming more and more a matter of end-to-end   responsiveness.g   -- i C.W.Holeman II  cwhii@acm.org				http://emle.org- http://also.as/cwhii		http://JulianLocals.comg   ------------------------------  # Date: Fri, 11 Jan 2002 19:24:56 GMTn2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) Subject: Re: xml and OpenVMS2 Message-ID: <c_G%7.582$5Y4.15207@news.cpqcorp.net>  _ In article <200201110726.IAA09376@sinet1.fom.fgan.de>, Rudolf Wingert <win@fom.fgan.de> writes:f   :is XML useable under OpenVMS? m     From the OpenVMS FAQ:l   XML:  8   Source code of an XML Parser is available from Oracle.     Also see:d  '     http://www.python.org/sigs/xml-sig/m    ?   An XML parser is available as part of OpenVMS V7.3 and later.      ----  . :If not, will we see this in the near furture?  E   That depends greatly on exactly what you mean -- a one- or two-line-A   question can only occasionally be unambiguously answered.  WhenpD   posting, please include OpenVMS version, platforms, and backgroundE   to the question -- all of which can help folks provide you with themH   answer you (really) seek, and potentially provide you with the answer    you seek rather more quickly.e  F   Given that XML is a text data taggging scheme, you could use XML on H   any VAX/VMS or OpenVMS version ever shipped -- you might well be able G   to design a hardware upgrade to the Babbage Engine that would permit >F   that widget to (noisily) generate XML.  Of course, there are variousD   current vendors that currently generate substantial XML noise, but?   I digress. :-)  (http://www.ee.ryerson.ca/~elf/ancient-comp/)   N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Fri, 11 Jan 2002 23:46:49 -0600sC From: "Craig A. Berry" <craig.berry@nospam.signaltreesolutions.com>r Subject: Re: xml and OpenVMS= Message-ID: <3C3FCDC9.6040103@nospam.signaltreesolutions.com>    Rudolf Wingert wrote:    > M > is XML useable under OpenVMS? If not, will we see this in the near furture?1    G XML what?  Used how?  Are you looking for a parser?  A content creationnD tool?  A display or analysis utility?   An API for building your own> tools, or end-user tools?  If you need a parser, do you need aF validating parser or only one that ensures well-formedness?  What kindB of content are you creating / analyzing / storing / transmitting / transforming / displaying?  D See <http://www.xml.org/xml/xmlfaq.shtml> among other places for theE myriad applications, sublanguages, and standards associated with XML.y  E If your goal is to display XML documents with associated stylesheets,rE the Mozilla version available for OpenVMS Alpha is as good as Mozilla L on any other platform for doing this (and at this point that's pretty good).  A A good XML document creation tool is probably tougher to come by.2= There are emacs plug-ins that do this, so when the emacs portuB periodically mentioned here becomes available that may be the best alternative.  G But perhaps you are not working with documents at all.  Perhaps you arec9 using XML for EDI between remote databases, or performing.D transformations on datasets, or describing database schemas.  XML isD used for all these things, and without knowing which one or ones areD relevant to your situation it's a bit hard to steer you in the right
 direction.  G As far as XML parsers go, Compaq offers one as part of the e-biz/Apache F suite.  IIRC, it is not a validating parser, which may be a good thingE or a bad thing depending on what you want to do.  And I believe thereiC are multiple parsers available written in Java which should work on 4 VMS, though I have no specific experience with them.  B Some time back Martin Vorlaender made some VMS mods to build James@ Clark's expat parser on VMS.  It is not the latest, greatest, orD sexiest XML parser, but my impression is that expat is the referenceE standard for all the others.  The XML::Parser Perl module is built ongC top of expat.  Martin's changes (to both expat and XML::Parser) are  here:p  H <http://www.xray.mpe.mpg.de/mailing-lists/vmsperl/2001-01/msg00032.html>  ? Or you can get a version of Perl that has expat and XML::ParserrF included with it at <http://www.sidhe.org/vmsperl>.  (If you just want< expat and not Perl, it's in the [.perlbuild.libraries.expat] directory.)r   ------------------------------  % Date: Fri, 11 Jan 2002 16:55:16 -0500e= From: Alexandre Pechtchanski <alex*@*rockvax.rockefeller.edu>dY Subject: Re: Younger recruits versus experienced veterans  ( was Re: The demise of compaqo8 Message-ID: <ennu3uo4qg5q2e96hehb3u8pcphjs99puo@4ax.com>  L On Thu, 10 Jan 2002 23:24:57 GMT, Charles Richmond <richmond@ev1.net> wrote: [ piggy-backing ]t   >jmfbahciv@aol.com wrote:E >> B* >> In article <3C3C1037.4ABD9256@ev1.net>,0 >>    Charles Richmond <richmond@ev1.net> wrote: >> > a0 >> >       [snip...]     [snip...]     [snip...] >> >( >> > ...That's why I see a lot of people: >> >coming home from the office at eight or nine at night. >> >: >> I don't understand this.  There must be an awful lot of >> meetings going on.2  > "Dear, I'm going to the meeting and from her right to you" ;-)   -- S* [ When replying, remove *'s from address ]0 Alexandre Pechtchanski, Systems Manager, RUH, NY   ------------------------------  # Date: Fri, 11 Jan 2002 19:40:57 GMT . From: "aaron spink" <aaronspink@earthlink.net>O Subject: Re: Younger recruits versus experienced veterans ( was Re:     The demlF Message-ID: <ddH%7.22780$zw3.2252429@newsread1.prod.itd.earthlink.net>  E "Christian Bau" <christian.bau@cbau.freeserve.co.uk> wrote in messagee. news:3C3EBD79.DEA279D3@cbau.freeserve.co.uk...E > Ok, you have so much more experience designing user interfaces than. > anyone at Apple does.1 >9H I would say that I have a lot more experience using a lot more differentK interfaces than the majority of the people at Apple have.  If you have been F designing the same crap for the past 15 years, it doesn't mean you are actually good at it.   Aaron Spinkn speaking for Myself Inc.   ------------------------------   Date: 11 Jan 02 11:42:54 -0800* From: "Charlie Gibbs" <cgibbs@sky.bus.com>O Subject: Re: Younger recruits versus experienced veterans ( was Re:     The demv/ Message-ID: <3381.776T2711T7025736@sky.bus.com>h  3 In article <3C3EBD79.DEA279D3@cbau.freeserve.co.uk>o: christian.bau@cbau.freeserve.co.uk (Christian Bau) writes:   >Paul Guertin wrote: > A >> As someone once said about shift-, option-, ctrl- and command-8C >> clicking, "The Mac does have a multibutton mouse, it's just that 6 >> Apple put all the buttons but one on the keyboard."   :-)U  E >> Shift-clicking has got to be one of the least intuitive operations2D >> there is. Why not have a click-while-turning-down-the-brightness- >> knob, while you're at it? >sG >Intuitiveness for beginners: The mouse is there to initiate an action. F >There are several possible actions, mostly clicking and dragging. TheH >modifier keys on the keyboard are there to modify the action. BasicallyF >the same action, but slightly different. The shift key Is Always UsedE >To Change The Behaviour Of Other User Interface Elements, as you can F >see what happened when I pressed the shift key while typing the firstH >letter of each word. So it is obvious that pressing the shift key might >modify other actions as well. >nG >Did you get this: The shift key on the keyboard is a modifier key that E >modifies the behaviour of other user interface elements. That is itst	 >purpose.m  ? Thanks for the sermon.  One of the reasons Apple irritates manyc< people is that they seem to claim total knowledge of what is; "intuitive" and what isn't.  Anyone who disagrees with them-; obviously has his brain wired incorrectly; how dare someoneu< suggest that what's intuitive to one person might be totally opaque to someone else?   @ Drag a file icon to the trash; it's deleted.  Drag a folder iconB to the trash; it's deleted.  I almost lost my lunch the first timeB I saw someone drag a floppy disk icon to the trash.  My (obviouslyC incorrectly wired) brain extrapolated the other examples to suggestoB that the entire disk was going to be trashed (formatted, perhaps).  G >There is no "click-while-turning-down-the-brightness-knob" because thesF >brightness knob on your monitor is intended for a single purpose, and6 >overloading its functionality would be non-intuitive. >=E >> If your interface uses a meta key to put the mouse in some specialiB >> mode, then put that key on the mouse itself, not on a unrelated >> peripheral.  F >The keyboard is not an "unrelated peripheral", it is one of the threeC >essential components (from the users point of view) that make up a D >modern computer: Monitor, keyboard, and mouse. All three components >interact together.   B The brightness knob is on the monitor.  The mouse button is on the? mouse.  These are two of the "three components" which "interact>F together".  Exactly how does "click-while-turning-down-the-brightness- knob" differ from shift-click?  @ To turn it around, I'll paraphrase what you said above: it couldA easily be argued that "the shift key on your keyboard is intendeduD for a single purpose [changing the meaning of other keys on the sameE keyboard], and overloading its functionality would be non-intuitive."aB (Hmmm, is it possible to put two keyboards on the ADB bus?  If so,> should the shift key on one affect keys pressed on the other?)  > Oops, I forgot.  Apple is the sole arbiter of what constitutes "intuitive".  My bad.c   --" cgibbs@sky.bus.com (Charlie Gibbs)5 Remove the first period after the "at" sign to reply.rD I don't read top-posted messages.  If you want me to see your reply,? appropriately trim the quoted text and put your reply below it.o   ------------------------------  % Date: Sat, 12 Jan 2002 14:46:57 +1300a# From: Bruce Hoult <bruce@hoult.org>oO Subject: Re: Younger recruits versus experienced veterans ( was Re:     The demt> Message-ID: <bruce-BCBAE6.14465712012002@news.paradise.net.nz>  @ In article <3381.776T2711T7025736@sky.bus.com>, "Charlie Gibbs"  <cgibbs@sky.bus.com> wrote:d  A > Thanks for the sermon.  One of the reasons Apple irritates manyD> > people is that they seem to claim total knowledge of what is= > "intuitive" and what isn't.  Anyone who disagrees with them== > obviously has his brain wired incorrectly; how dare someoneA> > suggest that what's intuitive to one person might be totally > opaque to someone else?v   But they *don't* claim that.  H What Apple does (or at least did 15 years ago) and which pretty much no D one else does (then or now) is implement the same functionallity in I every reasonable way they can think of and then *TEST* it on real people 6C and carefully record and analyse to see which techniques cause the , fewest user mistakes.p  F Of course no one does this because it's ten times more expensive than B just implementing the first thing you think of and stopping there.    B > Drag a file icon to the trash; it's deleted.  Drag a folder iconD > to the trash; it's deleted.  I almost lost my lunch the first timeD > I saw someone drag a floppy disk icon to the trash.  My (obviouslyE > incorrectly wired) brain extrapolated the other examples to suggesthD > that the entire disk was going to be trashed (formatted, perhaps).  E Yes that's dumb, and even Mac people have been saying its dumb for a w decade or more.a  @ The problem is that the steps leading to that behaviour were in  themselves quite sensible.  G You see, the original Macs had only a single floppy drive (and no hard nI disk) and the RAM was smaller than a floppy.  It was common to work with rG multiple floppy disks in the same drive at the same time.  When a disk  H was in the drive and you wanted to use another one you used the "eject" F menu item and the disk was ejected and you put a new disk in.  But it F wasn't *forgotten*.  The  Icon for the disk was still on the desktop, D the disk was still logically mounted, any folders you had open were E still visible, programs could have files on the disk open.  When you vG eventually did something with the disk that couldn't be satisfied from  H the cache in RAM the system ejected whatever was currently in the drive - and asked you to put the needed disk back in.d  E The "drag the icon to the trash" was at first something that you did eI *only* when the disk was already physically out of the drive.  It closed rF the data structures and logically unmounted the disk.  But it was too H much hassle to eject the disk and *then* drag the icon to the trash, so B it was changed so that you could do the whole thing in one action.  F Which was a mistake on the part of Apple, and everyone knows that and  agrees.   F It also was a fairly short-lived mistake, as they pretty soon added a I *different* menu item, "put away", which not only ejected disks but also  G unmounted them.  Well, actually, it was there nearly from the start in gC the role of returning files from the desktop to wherever they were =H before you dragged them to the desktop, and it got enhanced to make the & system eject and forget disks as well.  G Even later, when people started to use floppy disks very seldom -- and hF almost never multiple disks in the same drive at the same time -- the = "eject" menu item was changed to both eject and forget disks.   I Today, in MacOS X, if you start to drag a disk icon the trash changes to sI an "eject" icon.  So you can use the same physical action as before, but > it's clear what is happening.e    H The summary is that, yes, you've been able to eject disks on the Mac by I dragging them to the trash for a very long time, and that isn't really a mH very good user interface.  But it was the *only* way to make the system G forget a disk for really quite a short period of the last 18 years and -F there's been really no reason to show it to beginners as *the* method  for all the rest of the time.r   -- Bruce   ------------------------------  % Date: Sat, 12 Jan 2002 01:02:26 -0500  From: Paul Guertin <pg@sff.net>lO Subject: Re: Younger recruits versus experienced veterans ( was Re:     The demaA Message-ID: <5ihv3u00jqbv5pdqncdp5qmmhqa9uj6c7s@news.newsguy.com>e  9 Christian Bau <christian.bau@cbau.freeserve.co.uk> wrote:-   > Paul Guertin wrote:  > > F > > Shift-clicking has got to be one of the least intuitive operationsE > > there is. Why not have a click-while-turning-down-the-brightness-l > > knob, while you're at it?g > H > Did you get this: The shift key on the keyboard is a modifier key thatF > modifies the behaviour of other user interface elements. That is its
 > purpose.  > I bet most beginners would say the purpose of the shift key isA to type capital letters and punctuation characters printed at the  top of keycaps.0  H > There is no "click-while-turning-down-the-brightness-knob" because theG > brightness knob on your monitor is intended for a single purpose, andt7 > overloading its functionality would be non-intuitive.t  B Coming soon from Apple: the monitor with just one knob. Turn it toB adjust brightness, hold shift while turning it to adjust contrast,A use it with command to set the vertical hold, and poke at it withh  the mouse to degauss the screen.  A > > Apple got many things right with the MacOS interface, but the ( > > shift-click thing isn't one of them. > E > Ok, you have so much more experience designing user interfaces thanp > anyone at Apple does.o  7 I'm not a cook, but I can tell when my soup needs salt.0   Paul Guertin
 pg@sff.net   ------------------------------  % Date: Fri, 11 Jan 2002 17:27:10 -0500:* From: WILLIAM WEBB <WWEBB1@email.usps.gov>E Subject: RE: Younger recruits versus experienced veterans ( was Re: Te- Message-ID: <0033000047596866000002L062*@MHS>a  6 =0AThe movie was [loosely] based on Arthur C. Clarke's Childhood's End.  = WWWebb  [who saw the premiere in Atlanta at Martin's Cineramas: along with the kid who was generally regarded as the otherC "class intellectual". I wonder what ever became of Chris Mulrooney?t9 We took the bus there becuase we were too young to drive.e7 And I refuse to let my son see it until it comes aroundh to a a theater.t5 Some things just shouldn't be seen on small screens.]-     -----Original Message-----/ From: Info-VAX-Request@Mvb.Saic.Com at INTERNETl& Sent: Friday, January 11, 2002 5:09 PMB To: Webb, William W Raleigh, NC; Info-VAX@Mvb.Saic.Com at INTERNETE Subject: RE: Younger recruits versus experienced veterans ( was Re: T      Terje Mathisen wrote:r >s > Steve O'Hara-Smith wrote:p > >?& > > On Thu, 10 Jan 2002 20:39:49 +01008 > > Terje Mathisen <terje.mathisen@hda.hydro.com> wrote: > > D > > TM> Not _nearly_ as bad as current movies about the same things. > > H > >         Can you cite any examples where movies have been better tha= n evenH > > quite bad books in the same genre ? I am quite sure that there are = *no*H > > science fiction films that improve on novels with similar topics (l= et& > > alone the ones they are based on). >M > 2001?, >a > Otherwise, I agree.   F Wasn't the book "2001" written at the same time as, or even after, the screenplay?e  H I'll also mention HAL 9000 to keep this on topic for at least two of th= ee the newsgroups :-)=    ------------------------------  + Date: Sat, 12 Jan 2002 00:42:35 +0000 (UTC)O From: david20@alpha2.mdx.ac.ukE Subject: RE: Younger recruits versus experienced veterans ( was Re: Tn+ Message-ID: <a1o0pr$fh1$1@aquila.mdx.ac.uk>o  Z In article <0033000047596866000002L062*@MHS>, WILLIAM WEBB <WWEBB1@email.usps.gov> writes:7 >=0AThe movie was [loosely] based on Arthur C. Clarke's. >Childhood's End.a >e  J Wrong book. 2001 was based on Arthur C Clarke's short story "The Sentinel"& which featured a Monolith on the Moon.     See   " http://www.elstree.co.uk/2001.html   For  more information on 2001.    
 David Webb VMS and Unix team leader CCSS Middlesex University  > >WWWebb  [who saw the premiere in Atlanta at Martin's Cinerama; >along with the kid who was generally regarded as the other4D >"class intellectual". I wonder what ever became of Chris Mulrooney?: >We took the bus there becuase we were too young to drive.8 >And I refuse to let my son see it until it comes around >to a a theater.6 >Some things just shouldn't be seen on small screens.] >  >  >-----Original Message----- 0 >From: Info-VAX-Request@Mvb.Saic.Com at INTERNET' >Sent: Friday, January 11, 2002 5:09 PMsC >To: Webb, William W Raleigh, NC; Info-VAX@Mvb.Saic.Com at INTERNETIF >Subject: RE: Younger recruits versus experienced veterans ( was Re: T >r >w >Terje Mathisen wrote: >> >> Steve O'Hara-Smith wrote: >> >' >> > On Thu, 10 Jan 2002 20:39:49 +0100i9 >> > Terje Mathisen <terje.mathisen@hda.hydro.com> wrote:l >> >E >> > TM> Not _nearly_ as bad as current movies about the same things.s >> >I >> >         Can you cite any examples where movies have been better tha=d >n eveneI >> > quite bad books in the same genre ? I am quite sure that there are =  >*no*yI >> > science fiction films that improve on novels with similar topics (l=  >etX' >> > alone the ones they are based on).g >> >> 2001? >> >> Otherwise, I agree. >aG >Wasn't the book "2001" written at the same time as, or even after, thev >screenplay? >bI >I'll also mention HAL 9000 to keep this on topic for at least two of th=a >e >the newsgroups :-)=   ------------------------------  % Date: Fri, 11 Jan 2002 19:05:36 +0100 , From: Steve O'Hara-Smith <steveo@eircom.net>K Subject: Re: Younger recruits versus experienced veterans ( was Re: The dem 7 Message-ID: <20020111190536.2524b5c9.steveo@eircom.net>n  " On Fri, 11 Jan 2002 10:24:57 +00009 Christian Bau <christian.bau@cbau.freeserve.co.uk> wrote:     CB> Intuitiveness for beginners:   	ROTFLMAO, what a concept.  J CB> Did you get this: The shift key on the keyboard is a modifier key thatH CB> modifies the behaviour of other user interface elements. That is its
 CB> purpose. e  B 	The shift key is a left over from mechanical typewriters where itH moved (shifted) the mechanism up and down so as to strike the paper withL either the upper or lower letters. Given this experience why is it intuitiveE that the electronic version affects physically disconnected objects ?m  I CB> The keyboard is not an "unrelated peripheral", it is one of the threeaF CB> essential components (from the users point of view) that make up aG CB> modern computer: Monitor, keyboard, and mouse. All three components   B 	Monitor(s), keyboard(s), mouse(mice), bitpad(s), digitiser(s) ...H OK I have never seen a system with all the plurals, but I have certainlyI seen each of them and several combinations. If I have two mice in use for K different purposes (OK not many systems will do this) then which one shouldrJ be affected by the shift key on the keyboard ? Might the shift key also beI expected to interact with the remote control for a TV tuner card ? Spooky.* action at a distance is *never* intuitive.  G CB> Ok, you have so much more experience designing user interfaces thana CB> anyone at Apple does.   @ 	I suspect a good proportion of the regulars in here do, many ofI them with direct personal feedback from real users concentrating on doing]I real things (as distinct from evaluation users concentrating on what theyo) think of the interface they are testing).I   -- uH C:>WIN                                          |     Directable MirrorsN The computer obeys and wins.                    |A Better Way To Focus The SunL You lose and Bill collects.                     |  licenses available - see:J                                                 |   http://www.sohara.org/   ------------------------------  % Date: Fri, 11 Jan 2002 15:48:02 -0600r) From: "GerardS" <GerardS@PrairieTech.Net>eK Subject: Re: Younger recruits versus experienced veterans ( was Re: The deml/ Message-ID: <u3un8h5q4k2316@corp.supernews.com>r   | Steve O'Hara-Smith wrote:e | CB> Christian Bau9" | CB> Intuitiveness for beginners: |a | ROTFLMAO, what a concept.s |iL | CB> Did you get this: The shift key on the keyboard is a modifier key thatJ | CB> modifies the behaviour of other user interface elements. That is its | CB> purpose. |aC | The shift key is a left over from mechanical typewriters where ittJ | moved (shifted) the mechanism up and down so as to strike the paper withN | either the upper or lower letters. Given this experience why is it intuitiveG | that the electronic version affects physically disconnected objects ?g  Q The  SHIFT  key is used to choose either the lower or upper character on the key,TQ not just the letters.    Also, don't confuse the shift lock on manual typewriters O with the modern caps-lock on most computer keyboards. --------------- Gerard S.:   ------------------------------  % Date: Fri, 11 Jan 2002 15:38:35 -0800>* From: "Jack Peacock" <peacock@simconv.com>Y Subject: Re: Younger recruits versus experienced veterans ( was Re: Thedemise of compaq )>6 Message-ID: <nKK%7.2656$3g7.202542@news.webusenet.com>  9 "Steve O'Hara-Smith" <steveo@eircom.net> wrote in message-1 news:20020111074725.6aa55981.steveo@eircom.net... C > Can you cite any examples where movies have been better than evenc% > quite bad books in the same genre ?i  K Dr. Strangelove...the movie was brilliant and more coherent than any Philipe K. Dick story.    Jack Peacock    ------------------------------  % Date: Fri, 11 Jan 2002 21:31:41 +0100 3 From: Terje Mathisen <terje.mathisen@hda.hydro.com>,Y Subject: Re: Younger recruits versus experienced veterans ( was Re: Thedemise of compaq )e- Message-ID: <3C3F4BAD.F5115E3F@hda.hydro.com>t   Steve O'Hara-Smith wrote:t > $ > On Thu, 10 Jan 2002 20:39:49 +01006 > Terje Mathisen <terje.mathisen@hda.hydro.com> wrote: > B > TM> Not _nearly_ as bad as current movies about the same things. > K >         Can you cite any examples where movies have been better than even I > quite bad books in the same genre ? I am quite sure that there are *no*nG > science fiction films that improve on novels with similar topics (leti$ > alone the ones they are based on).   2001?t   Otherwise, I agree.e   Terjei -- e  - <Terje.Mathisen@hda.hydro.com>@ "almost all programming can be viewed as an exercise in caching"   ------------------------------  % Date: Fri, 11 Jan 2002 17:08:35 -0500a' From: Dave Perks <Dave_Perks@mitel.com>tY Subject: Re: Younger recruits versus experienced veterans ( was Re: Thedemise of compaq ) ) Message-ID: <3C3F6263.F0C37156@mitel.com>d   Terje Mathisen wrote:  >  > Steve O'Hara-Smith wrote:- > >t& > > On Thu, 10 Jan 2002 20:39:49 +01008 > > Terje Mathisen <terje.mathisen@hda.hydro.com> wrote: > >dD > > TM> Not _nearly_ as bad as current movies about the same things. > >tM > >         Can you cite any examples where movies have been better than eventK > > quite bad books in the same genre ? I am quite sure that there are *no* I > > science fiction films that improve on novels with similar topics (lete& > > alone the ones they are based on). >  > 2001?m >  > Otherwise, I agree.y  F Wasn't the book "2001" written at the same time as, or even after, the screenplay?o  H I'll also mention HAL 9000 to keep this on topic for at least two of the the newsgroups :-)   ------------------------------   End of INFO-VAX 2002.021 ************************