1 INFO-VAX	Thu, 03 Mar 2005	Volume 2005 : Issue 123       Contents: Re: Backup /ALIAS question Build problems - ZLIB  Re: Build problems - ZLIB & Re: DECC : toupper/tolower performance& Re: DECC : toupper/tolower performance& Re: DECC : toupper/tolower performance& Re: DECC : toupper/tolower performance& Re: DECC : toupper/tolower performance& Re: DECC : toupper/tolower performance DECC floating point question DECC floating point question  Re: DECC floating point question  Re: DECC floating point question  Re: DECC floating point question  Re: DECC floating point question# Re: Fiorina and The Peter Principle # Re: Fiorina and The Peter Principle # Re: Fiorina and The Peter Principle # Re: Fiorina and The Peter Principle # Re: Fiorina and The Peter Principle # Re: Fiorina and The Peter Principle # Re: Fiorina and The Peter Principle # Re: Fiorina and The Peter Principle # Re: Fiorina and The Peter Principle # Re: Fiorina and The Peter Principle  Re: GBLSECTIONS loop Re: IA64 unraveling... IBM knows itanium is a dud ...J RE: Interesting error message I just got trying to access accuwea	ther.comJ RE: Interesting error message I just got trying to access accuwea	ther.com& Re: new Itanium after Tukwila: Poulson& Re: new Itanium after Tukwila: Poulson& Re: new Itanium after Tukwila: Poulson& Re: new Itanium after Tukwila: Poulson8 Re: OpenVMS Seminar in Toronto (2005-02-24) a few points8 Re: OpenVMS Seminar in Toronto (2005-02-24) a few points8 Re: OpenVMS Seminar in Toronto (2005-02-24) a few points8 Re: OpenVMS Seminar in Toronto (2005-02-24) a few points& Re: Problem with a DEC terminal server Re: vms mail forwarding  Re: vms mail forwarding < Re: What is the Difference between Shadow and Mirrored disk?  F ----------------------------------------------------------------------   Date: 2 Mar 2005 19:00:25 -0800 $ From: "AEF" <spamsink2001@yahoo.com># Subject: Re: Backup /ALIAS question C Message-ID: <1109818825.855808.230590@g14g2000cwa.googlegroups.com>    Bob Koehler wrote:> > In article <422516E3.7B201985@comcast.net>, David J Dachtera# <djesys.nospam@comcast.net> writes:  > > > > > However, UN*X-like "links" ("aliases") are new with ODS-5. > >  > D >    Yes, but I got the impression that the OP was talking about old >    fashioned file aliases.    F I don't think it matters. The difference between VMS aliases and "hard/ links" is what happens with the DELETE command.    See   1     http://h71000.www7.hp.com/portability/51.html    for details.  B I'd expect /IMAGE to always restore the disk correctly. But if youE changed the status of a disk from aliases to hard links or vice versa @ in the midst of a series of incremental backups, I don't know ifC restoring with /INCREMENTAL would get that right. IOW, are modified C volume attributes copied by incremental save operations and applied D during /INCREMENTAL restore operations? What about the volume label?   ------------------------------   Date: 2 Mar 2005 17:03:56 -0800  From: mcbill20@yahoo.com Subject: Build problems - ZLIBC Message-ID: <1109811836.288768.155000@f14g2000cwb.googlegroups.com>   F I am still trying to build all the modules in order to eventually have6 a working GDCHART. I have been working from this page:  ; http://www.quadratrix.be/products/open_source_material.html   C I have downloaded the ZLIB 1.2.2 (the page above says 1.2.1 but all D that seems to be available at the target is 1.2.2). After unzipping,E every .c and .h file needed to be fixed; that is, every line in every 0 file had an embedded carriage return at the end.  D I then downloaded the VMS build proc and patches. It was a small zipD file containing a small backup saveset that also needed to be fixed.D The contents were just two files: DESCRIP.MMS and QPORT.COM, as with" most of the software on this page.  D Here is the output I get when trying to build. For this one, I don't really even know where to look.    $ @qport  , Building ZLib 1.2.1 for OpenVMS Alpha V7.3-1  y CC/PREFIX=ALL/DEFINE=(QPORT=1,"NO_vsnprintf")/INCLUDE=([],BUILD_DEV:[ZLIB])/OBJECT=BUILD_DEV:[ZLIB.CONTRIB.VMS.OBJ_ALPHA]  BUILD_DEV:[ZLIB]EXAMPLE.C y CC/PREFIX=ALL/DEFINE=(QPORT=1,"NO_vsnprintf")/INCLUDE=([],BUILD_DEV:[ZLIB])/OBJECT=BUILD_DEV:[ZLIB.CONTRIB.VMS.OBJ_ALPHA]  BUILD_DEV:[ZLIB]ADLER32.C y CC/PREFIX=ALL/DEFINE=(QPORT=1,"NO_vsnprintf")/INCLUDE=([],BUILD_DEV:[ZLIB])/OBJECT=BUILD_DEV:[ZLIB.CONTRIB.VMS.OBJ_ALPHA]  BUILD_DEV:[ZLIB]COMPRESS.Cy CC/PREFIX=ALL/DEFINE=(QPORT=1,"NO_vsnprintf")/INCLUDE=([],BUILD_DEV:[ZLIB])/OBJECT=BUILD_DEV:[ZLIB.CONTRIB.VMS.OBJ_ALPHA]  BUILD_DEV:[ZLIB]CRC32.C    #  define NO_vsnprintf ......................^ > %CC-W-MACROREDEF, The redefinition of the macro "NO_vsnprintf"A conflicts with a current definition because the replacement lists + differ.  The redefinition is now in effect. 4 at line number 193 in file BUILD_DEV:[ZLIB]zutil.h;2D %MMK-F-ERRUPD, error status %X10B91260 occurred when updating target BUILD_DEV:[ $ ZLIB.CONTRIB.VMS.OBJ_ALPHA]CRC32.OBJ $   G One strange thing is the object file that it is complaining about. When F I run the qport.com again, I would expect it to stop at the same place< but it complains about a different file (GZIO.OBJ instead of CRC32.OBJ):    $ @qport  , Building ZLib 1.2.1 for OpenVMS Alpha V7.3-1  P CC/PREFIX=ALL/DEFINE=(QPORT=1,"NO_vsnprintf")/INCLUDE=([],BUILD_DEV:[ZLIB])/OBJE@ CT=BUILD_DEV:[ZLIB.CONTRIB.VMS.OBJ_ALPHA] BUILD_DEV:[ZLIB]GZIO.C   #  define NO_vsnprintf ......................^ > %CC-W-MACROREDEF, The redefinition of the macro "NO_vsnprintf" conflicts with a= current definition because the replacement lists differ.  The  redefinition is no w in effect.4 at line number 193 in file BUILD_DEV:[ZLIB]zutil.h;2D %MMK-F-ERRUPD, error status %X10B91260 occurred when updating target BUILD_DEV:[ # ZLIB.CONTRIB.VMS.OBJ_ALPHA]GZIO.OBJ    ------------------------------  % Date: Thu, 03 Mar 2005 05:51:48 +0100 2 From: martin@radiogaga.harz.de (Martin Vorlaender)" Subject: Re: Build problems - ZLIB; Message-ID: <422697e4.524144494f47414741@radiogaga.harz.de>    mcbill20@yahoo.com wrote: H > I am still trying to build all the modules in order to eventually have8 > a working GDCHART. I have been working from this page: > = > http://www.quadratrix.be/products/open_source_material.html  > E > I have downloaded the ZLIB 1.2.2 (the page above says 1.2.1 but all F > that seems to be available at the target is 1.2.2). After unzipping,G > every .c and .h file needed to be fixed; that is, every line in every 2 > file had an embedded carriage return at the end.  G This happens when the archive has not been packed on a VMS machine. You I can reverse the effect by telling unzip to treat all files in the archive  as text files (option "-a").   F > I then downloaded the VMS build proc and patches. It was a small zipF > file containing a small backup saveset that also needed to be fixed.F > The contents were just two files: DESCRIP.MMS and QPORT.COM, as with$ > most of the software on this page. > F > Here is the output I get when trying to build. For this one, I don't! > really even know where to look.  > 
 > $ @qport > . > Building ZLib 1.2.1 for OpenVMS Alpha V7.3-1 ... { > CC/PREFIX=ALL/DEFINE=(QPORT=1,"NO_vsnprintf")/INCLUDE=([],BUILD_DEV:[ZLIB])/OBJECT=BUILD_DEV:[ZLIB.CONTRIB.VMS.OBJ_ALPHA]  > BUILD_DEV:[ZLIB]CRC32.C  >  > #  define NO_vsnprintf > ......................^ @ > %CC-W-MACROREDEF, The redefinition of the macro "NO_vsnprintf"C > conflicts with a current definition because the replacement lists - > differ.  The redefinition is now in effect. 6 > at line number 193 in file BUILD_DEV:[ZLIB]zutil.h;2F > %MMK-F-ERRUPD, error status %X10B91260 occurred when updating target1 > BUILD_DEV:[ZLIB.CONTRIB.VMS.OBJ_ALPHA]CRC32.OBJ  > $   C You have defined NO_vsnprintf on the command line. There is another . definition of that macro in ZUTIL.H, line 193.   I > One strange thing is the object file that it is complaining about. When H > I run the qport.com again, I would expect it to stop at the same place> > but it complains about a different file (GZIO.OBJ instead of
 > CRC32.OBJ):   E As the above is only a warning, the compiler does generate CRC32.OBJ. H When you call QPORT.COM again, MMS sees that and continues with the next file.   C As ZUTIL.H is likely #include'd in all sources, the cure is simple:    I Either remove the definition of NO_vsnprintf from the command line (could ? be in QPORT.COM as well as DESCRIP.MMS), or extend ZUTIL.H like    #ifndef NO_vsnprintf #  define NO_vsnprintf #endif   HTH,   Martin  I P.S.: I think J.F.Pieronne has a binary distribution of ZLIB at his site. 6 Dunno, however, whether GDCHART can be built using it. --  B    Emacs would be a great   | Martin Vorlaender  |  OpenVMS rules!5    operating system,        | work: mv@pdv-systeme.de H    if only it came with     |   http://www.pdv-systeme.de/users/martinv/<    a decent editor...       | home: martin@radiogaga.harz.de   ------------------------------  % Date: Wed, 02 Mar 2005 12:54:39 -0800  From: Z <Z@no.spam> / Subject: Re: DECC : toupper/tolower performance ( Message-ID: <eSpVd.2603$hb.314@fe07.lga>    martinkirby12@yahoo.co.uk wrote:) >>have you looked at  _toupper/_tolower ?   G > _toupper is only valid if the character is upper case and _tolower if D > it is lower so would need to test with isupper etc. first which is* > probably much the same as using toupper.  < No, those are the in-line macro versions of toupper/tolower.  F If you're processing a lot of bytes, one at a time, the call overhead I with the toupper/tolower functions is significant and you should look at  @ the _to macros, just be aware of the caveats when using a macro.   ------------------------------  % Date: Wed, 02 Mar 2005 12:58:14 -0800  From: Z <Z@no.spam> / Subject: Re: DECC : toupper/tolower performance ( Message-ID: <AVpVd.2604$Yc.986@fe07.lga>    martinkirby12@yahoo.co.uk wrote:H >>From the "HP C Run-time Library Reference Manual for OpenVMS Systems", > V8.2 version:  > "DescriptionF > The _tolower macro is equivalent to the tolower function except thatE > its argument must be an uppercase letter (not lowercase, not EOF)."    That doc is wrong.   $ cc/ver) Compaq C V6.5-001 on OpenVMS Alpha V7.1-2    $ help cc run-time tolower   C  Run-time_functions   tolower, _tolower   C Convert their argument, a character, to lowercase.  If the argument 8 is not an uppercase character, it is returned unchanged.   Syntax:    #include <ctype.h>   int tolower(int character);    int _tolower(int character);   ------------------------------  % Date: Wed, 02 Mar 2005 22:39:33 -0000 1 From: wspencer@ap.dontspamme.org (Warren Spencer) / Subject: Re: DECC : toupper/tolower performance 1 Message-ID: <960DBE4A2wspenceraporg@216.168.3.30>   1 jfmezei.spamnot@teksavvy.com (JF Mezei) wrote in  7 <1109697967.846299e2b11a48e9742a810baa402624@teranews>:    > D >I need to do a case insensitive comparison. I looked at the toupperI >macro and while it works, i was wondering about peformance ( it seems to  >do a lot of checking).  > F >Would it be faster to just have an array of 256 bytes and just lookup@ >the translated character based on the value of the untranslated >character ? > $ >newchar = uppercase[oldchar] ;  ??? > E >or does DECC end up taking mroe juice out of the VAX CPU for such an  >operation compared to:  >  >newchar = toupper(oldchar) ;  > 6 >(I can setup my own translation table fairly easily).  K I know you're asking about performance, but in case your application could  K one day run on Solaris, I'll pass on what I learned yesterday working on a  L port from OpenVMS to Solaris:  _toupper() on VMS != _toupper() on Solaris.  F As documented in the man pages, Solaris' _toupper() returns undefined I results if the character passed to it is NOT lower case.  OpenVMS on the  H other hand, kindly returns an upper case character when you pass it one.  H I suspect your "newchar = uppercase[oldchar] ;" may perform better than > toupper(), but I can guarantee it is definitely more portable.   ws   --   Warren Spencer Senior Software Engineer The Associated Press   ------------------------------  % Date: Wed, 02 Mar 2005 18:27:28 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> / Subject: Re: DECC : toupper/tolower performance B Message-ID: <1109805310.efec191c5d223a536ad4b542e8a30d3d@teranews>   Dave Froble wrote:I > That's interesting.  How do you handle such?  Both upper and lower case 0 > accent characters have ascii values above 190.  C Yes, but in case you didn't know, a single bytes is 8 bits and thus  handles values from 0 to 255.   H people outside the uSA know to compile C programs with /UNSIGNED so thatG the compiler doesn't purposefully add code to chop off the 8th bit from 3 characters. To me, an  is just another character.    G VMS's SORT has the ability to have smart collating sequences that allow B you to treat an  the same as an e or  so that the text comes outA sorted the right way, instead of using the raw ascii value of the $ chartacter which would put  after z  B It is just that most users of VMS don't think of this when writingE software and since it isn't there by default, not much software makes 
 use of it.   ------------------------------  % Date: Wed, 02 Mar 2005 20:21:14 -0500 ' From: Dave Froble <davef@tsoft-inc.com> / Subject: Re: DECC : toupper/tolower performance 0 Message-ID: <112cp3hmlv9k37e@corp.supernews.com>   JF Mezei wrote:  > Dave Froble wrote: > I >>That's interesting.  How do you handle such?  Both upper and lower case 0 >>accent characters have ascii values above 190. >  > E > Yes, but in case you didn't know, a single bytes is 8 bits and thus   > handles values from 0 to 255.   > I'm aware of that.  It's the sorting that I'm wondering about.  J > people outside the uSA know to compile C programs with /UNSIGNED so thatI > the compiler doesn't purposefully add code to chop off the 8th bit from 5 > characters. To me, an  is just another character.   > I > VMS's SORT has the ability to have smart collating sequences that allow D > you to treat an  the same as an e or  so that the text comes outC > sorted the right way, instead of using the raw ascii value of the & > chartacter which would put  after z  E That's the answer to my question.  I figured that there probably was  I something that would ease such, rather than your having to have a tag or   such for sorting.   D > It is just that most users of VMS don't think of this when writingG > software and since it isn't there by default, not much software makes  > use of it.   Learn something new every day.   ------------------------------  % Date: Wed, 02 Mar 2005 22:28:47 -0600 6 From: "Craig A. Berry" <craigberry@mac.com.spamfooler>/ Subject: Re: DECC : toupper/tolower performance D Message-ID: <craigberry-BD6DDC.22284702032005@news.isp.giganews.com>  1 In article <960DBE4A2wspenceraporg@216.168.3.30>, 3  wspencer@ap.dontspamme.org (Warren Spencer) wrote:   H > As documented in the man pages, Solaris' _toupper() returns undefined ; > results if the character passed to it is NOT lower case.    E If that's true, and I have no reason to doubt your account, then the  G Solaris implementation does not comply with the ANSI C standard, which  ? states that the argument is returned unmodified unless it is a  : character in the current locale for which islower is true.   ------------------------------   Date: 2 Mar 2005 19:18:26 -0800  From: damnihatethat@gmail.com % Subject: DECC floating point question C Message-ID: <1109819906.558841.250000@g14g2000cwa.googlegroups.com>   = I would really like the following code to produce the result:  temp = 63250   but instead it produces: temp = 63249    I compile it on vms 7.3.1 using:   $cc test
 $link test	 $run test   / Am I wrong to expect the result I am expecting?  Thanks for your input,   DH     #include <stdio.h>   main() {      double odds1 = 2.75;   double odds2 = 2.30;   double result;   unsigned int temp;     result = odds1 * odds2; +   temp = (unsigned int) (result * 10000.0);    printf("temp = %d\n", temp);   }    ------------------------------  * Date: Wed, 2 Mar 2005 21:35:00 -0600 (CST)* From: sms@antinode.org (Steven M. Schweda)% Subject: DECC floating point question 2 Message-ID: <05030221350034_27800279@antinode.org>   From: damnihatethat@gmail.com   ? > I would really like the following code to produce the result:  > temp = 63250 >  > but instead it produces: > temp = 63249  1 > Am I wrong to expect the result I am expecting?    >   double odds2 = 2.30;  =    Well, we can all relax now.  This takes care of the annual < floating-point math problem posting in comp.os.vms for 2005.  @    A quick Google search for "floating-point math basics" found:  -       http://www.petebecker.com/js200006.html   8 which seems to cover the usual material reasonably well.  $       [...]  From Knuth, once again:  9           [E]very well-rounded programmer ought to have a 9           knowledge of what goes on during the elementary :           steps of floating point arithmetic. This subject<           is not at all as trivial as most people think, and9           it involves a surprising amount of interesting             information.  A    No doubt others can point to other sources.  Other geezers are 5 welcome to join me in my annual bout of head-shaking.   H ------------------------------------------------------------------------  4    Steven M. Schweda               (+1) 651-699-98183    382 South Warwick Street        sms@antinode-org     Saint Paul  MN  55105-2547    ------------------------------  % Date: Wed, 02 Mar 2005 19:39:47 -0800 # From: "Tom Linden" <tom@kednos.com> ) Subject: Re: DECC floating point question ( Message-ID: <opsm1lslolzgicya@hyrrokkin>  > On 2 Mar 2005 19:18:26 -0800, <damnihatethat@gmail.com> wrote:  ? > I would really like the following code to produce the result:  > temp = 63250 >  > but instead it produces: > temp = 63249 > " > I compile it on vms 7.3.1 using: > 
 > $cc test > $link test > $run test  > 1 > Am I wrong to expect the result I am expecting?  > Thanks for your input, >  > DH >  >  > #include <stdio.h> >  > main() > {  >  >   double odds1 = 2.75; >   double odds2 = 2.30; >   double result; >   unsigned int temp; >  >   result = odds1 * odds2; - >   temp = (unsigned int) (result * 10000.0);   >   printf("temp = %d\n", temp); >  > }  >   C It appears that there is some truncation taking place, as well as   
 convertingG  from decimal to binary float.  I tried the same with PL/I just confirm    FREJA> creat t.pli t: proc options(main); dcl (x,y,r) float bin(53);	 x = 2.75;  y=2.3; r = x*y*10000; put skip list(r);  end t; *EXIT* FREJA> pli t
 FREJA> link t  FREJA> run t     6.324999999999999E+04  FREJA> creat d.pli t: proc options(main); dcl (x,y,r) float dec(31);	 x = 2.75;  y=2.3; r = x*y*10000; put skip list(r);  end t; *EXIT* FREJA> pli dG N.B. I meant to write fixed dec(31) so go the warnings but it gives the  correct result.      dcl (x,y,r) float dec(31);
 .........^- %PLIG-W-FLTDPREC, The precision specified for                    "R" exceeds theI                  implementation's limit of FLOAT DECIMAL(15). The maximum 3                  precision of 15 has been supplied. ; at line number 2 in file SYS$SYSDEVICE:[TOM.SCRATCH]D.PLI;1    dcl (x,y,r) float dec(31); .......^- %PLIG-W-FLTDPREC, The precision specified for                    "Y" exceeds theI                  implementation's limit of FLOAT DECIMAL(15). The maximum 3                  precision of 15 has been supplied. ; at line number 2 in file SYS$SYSDEVICE:[TOM.SCRATCH]D.PLI;1    dcl (x,y,r) float dec(31); .....^- %PLIG-W-FLTDPREC, The precision specified for                    "X" exceeds theI                  implementation's limit of FLOAT DECIMAL(15). The maximum 3                  precision of 15 has been supplied. ; at line number 2 in file SYS$SYSDEVICE:[TOM.SCRATCH]D.PLI;1   9 %PLIG-I-SUMMARY, Completed with 0 errors, 3 warnings, and *                  0 informational messages.
 FREJA> link d $ %LINK-W-WRNERS, compilation warnings<          in module T file SYS$SYSDEVICE:[TOM.SCRATCH]D.OBJ;1 FREJA> ru d      6.32500000000000E+04 FREJA>  I I can only conclude that you are using the wrong language, try fortran or  PL/I.      --  C Using Opera's revolutionary e-mail client: http://www.opera.com/m2/    ------------------------------  % Date: Thu, 03 Mar 2005 00:07:10 -0500 ' From: Dave Froble <davef@tsoft-inc.com> ) Subject: Re: DECC floating point question 0 Message-ID: <112d6b8ogcn0hb1@corp.supernews.com>   Tom Linden wrote: @ > On 2 Mar 2005 19:18:26 -0800, <damnihatethat@gmail.com> wrote: > @ >> I would really like the following code to produce the result: >> temp = 63250  >> >> but instead it produces:  >> temp = 63249  >># >> I compile it on vms 7.3.1 using:  >> >> $cc test 
 >> $link test  >> $run test >>2 >> Am I wrong to expect the result I am expecting? >> Thanks for your input,  >> >> DH  >> >> >> #include <stdio.h>  >>	 >> main()  >> { >> >>   double odds1 = 2.75;  >>   double odds2 = 2.30;  >>   double result;  >>   unsigned int temp;  >> >>   result = odds1 * odds2;. >>   temp = (unsigned int) (result * 10000.0);! >>   printf("temp = %d\n", temp);  >> >> } >> > E > It appears that there is some truncation taking place, as well as    > convertingH > from decimal to binary float.  I tried the same with PL/I just confirm >    <big PL/I snip>    >  >  6.32500000000000E+04  > FREJA> > K > I can only conclude that you are using the wrong language, try fortran or  > PL/I.  >  >   A (This kind of stuff from C always brings out the language bigots)   I Well, maybe the wrong language.  I'd suggest BASIC, as below, first on a   VAX:   10      Declare Long T            a1 = 2.75          a2 = 2.30            a = a1 * a2'          print using "######.######", a             r = a * 10000.0'          print using "######.######", r             t = a * 10000.0          print t   Ready    run  NONAME     2-MAR-2005 23:39          6.325000   63250.000000   63250  Ready   # However, not so fast.  On an Alpha:    $ run t        6.325000   63250.000000   63249  $   G Now that I got in my required language promotion, I'll guess that your  G problem is using DOUBLE (8 byte floating point) on an Alpha.  There is  E no such thing.  Alpha converts such to IEEE floating point, does the  G calculations, and then converts it back.  Guess what, there are 3 less  * bits of prescision in IEEE floating point.   You might try:  5 temp = (unsigned int) ( (result * 10000.0) + .00001);   , That fixes my example in BASIC on the Alpha.   Or, get a VAX.   Dave   ------------------------------  % Date: Thu, 03 Mar 2005 01:00:00 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> ) Subject: Re: DECC floating point question B Message-ID: <1109828831.47567526acb94f1e0a5f85bd7f9f2a36@teranews>   "Steven M. Schweda" wrote:/ >       http://www.petebecker.com/js200006.html   H Thanks for the pointer. I knew that floating point was not accurate. ButE I found a tidbit in there I didn't know: it mentions that calculators / don't work in base 2. Is that really the case ?   7 Do they have CPUs that can handle base 10 calculation ?   @ To answer the original poster's question, one solution is to useG integers. You want 6 digit precision, then multiply your floating point C value by 100000 and then do your calculations as integers, and when , done, divide by 100000 to print your result.   ------------------------------  * Date: Wed, 2 Mar 2005 23:53:32 -0600 (CST)* From: sms@antinode.org (Steven M. Schweda)) Subject: Re: DECC floating point question 2 Message-ID: <05030223533257_27800279@antinode.org>  - From: JF Mezei <jfmezei.spamnot@teksavvy.com>   J > Thanks for the pointer. I knew that floating point was not accurate. ButG > I found a tidbit in there I didn't know: it mentions that calculators 1 > don't work in base 2. Is that really the case ?  > 9 > Do they have CPUs that can handle base 10 calculation ?   C    I don't know about modern ones, but the HP-35 (and friends) did.   7       http://www.hpl.hp.com/hpjournal/72jun/jun72a1.htm 7       http://www.hpl.hp.com/hpjournal/72jun/jun72a1.pdf   E       [...]  Each word consists of 14 binary-coded-decimal digits, so '       each word is 56 bits long.  [...]   B    Presumably it was easier to do decimal calculations than to theH decimal-binary-decimal conversions.  Plus, when "10 Enter 3 /" comes outE as 3.333333333, no one is amazed, but when "3.3 Enter 10 *" comes out E 32.99999999 (or 33.00000001), users may reasonably be expected to get  annoyed.  F    (Of course, the HP-45 fixed the too-many-digits-displayed problem. H Among other things, this teaches us that when one is spending $395 for a? calculator, it pays to wait for the second generation product.)   H ------------------------------------------------------------------------  4    Steven M. Schweda               (+1) 651-699-98183    382 South Warwick Street        sms@antinode-org     Saint Paul  MN  55105-2547    ------------------------------  $ Date: Wed, 2 Mar 2005 15:32:47 -0500# From: "John Smith" <a@nonymous.com> , Subject: Re: Fiorina and The Peter Principle, Message-ID: <48CdnXS01pttv7vfRVn-1g@igs.net>   John Vottero wrote: 0 > "John Smith" <a@nonymous.com> wrote in message( > news:47Sdnb7wLpaYXbjfRVn-iw@igs.net...G >> http://www.nytimes.com/2005/03/02/business/worldbusiness/02bank.html  >> >> March 2, 2005* >> Fiorina Called Candidate for World Bank >> By ELIZABETH BECKER >>F >> ASHINGTON, March 1 - Carleton S. Fiorina, who lost her job as chiefH >> executive of Hewlett-Packard almost three weeks ago, has emerged as a	 >> strong C >> candidate to become president of the World Bank, according to an  >> official in >> the Bush administration.  > E > The World Bank uses OpenVMS.  Wouldn't it be ironic if Carly starts ' > demanding better handling of OpenVMS!     4 Only if it impacts on her flight schedules to Davos.  F If they appoint her, we should write her about the need to protect theL Bank's investment in VMS by her writing a letter to the Chariman & CEO of HPF encouraging them to advertise it so its marketshare will grow, therebyJ encouraging HP to continue investing in it. She could also ask for EV8 and beyond rather than Itanic.   --- OpenVMS - The classics never go out of style.    ------------------------------  % Date: Wed, 02 Mar 2005 18:30:24 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> , Subject: Re: Fiorina and The Peter PrincipleB Message-ID: <1109805482.b48929dfad008cf48321d11904dd52b3@teranews>   David Mathog wrote:   G > Proven record alright, but not in a good sense.  Bush would only name B > her to the World Bank if he wanted to destroy that institutiion.  F Carly has been a staunch republican supporter and had been rumoured toE get a government appointment for some time. World bank appointment is 3 better than what I had expected her to get offered.   E And there is no way that the rest of the world will accept that a war H criminal be nominated as head of world bank. So Carly might get the job.E  She'll coin new phrases such as "Agile banking", "The adaptive World 	 Bank" :-)    ------------------------------  % Date: Wed, 02 Mar 2005 19:37:40 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> , Subject: Re: Fiorina and The Peter PrincipleB Message-ID: <1109809509.e93fe102bb0c02bbf65f230d2d364954@teranews>   John Smith wrote: H > If they appoint her, we should write her about the need to protect theN > Bank's investment in VMS by her writing a letter to the Chariman & CEO of HPH > encouraging them to advertise it so its marketshare will grow, thereby  G There is another side of the coin: if you tell her her new organisation F runs VMS, she may decide to find out whcih part of the World Bank runs@ VMS and then make sure they replace it with Bill gates software.   ------------------------------  % Date: Wed, 02 Mar 2005 20:28:02 -0500 ' From: Dave Froble <davef@tsoft-inc.com> , Subject: Re: Fiorina and The Peter Principle0 Message-ID: <112cpgd2adph72c@corp.supernews.com>   John Smith wrote: F > http://www.nytimes.com/2005/03/02/business/worldbusiness/02bank.html >  > March 2, 2005 ) > Fiorina Called Candidate for World Bank  > By ELIZABETH BECKER  > E > ASHINGTON, March 1 - Carleton S. Fiorina, who lost her job as chief N > executive of Hewlett-Packard almost three weeks ago, has emerged as a strongN > candidate to become president of the World Bank, according to an official in > the Bush administration.  " There is no justice in this world.  " Long time employees are being cut.  D The idiot gets what, somewhere over $40 million, as she's fired for  incomptence.  / The laid off employee cannot find a decent job.   6 The idiot is being considered for something like this.   ------------------------------   Date: 2 Mar 2005 17:38:04 -0800 $ From: "AEF" <spamsink2001@yahoo.com>, Subject: Re: Fiorina and The Peter PrincipleB Message-ID: <1109813884.106281.90850@z14g2000cwz.googlegroups.com>   Dave Froble wrote: > John Smith wrote:  > > D http://www.nytimes.com/2005/03/02/business/worldbusiness/02bank.html > >  > > March 2, 2005 + > > Fiorina Called Candidate for World Bank  > > By ELIZABETH BECKER  > > G > > ASHINGTON, March 1 - Carleton S. Fiorina, who lost her job as chief G > > executive of Hewlett-Packard almost three weeks ago, has emerged as  a strongD > > candidate to become president of the World Bank, according to an official in  > > the Bush administration. > $ > There is no justice in this world. > $ > Long time employees are being cut. > E > The idiot gets what, somewhere over $40 million, as she's fired for  > incomptence. > 1 > The laid off employee cannot find a decent job.  > 8 > The idiot is being considered for something like this.     Yep. Pretty amazing.  B And the same with Cappellas being recommended for new CEO of HP by M.Lynch.  E Let me see if I got it straight: The merger was frowned on in part as F an attempt for two drunks to hold each other up. Compaq employees wereF begging to be saved by the merger as they saw their company going downD the tubes. So Curly didn't do much good for Compaq and Carly for hp.B Now both Curly and Carly are gone from hp. Yet Meryll Lynch favorsE Cappellas for new CEO of hp? Huh? And (I don't know if this is right) F wasn't Carly the one who brought Lucent's stock down to single digits?F And they both get millions for all this? What does ML get out of Curly as CEO?   9 I may well have some or all of this wrong -- corrections, D clarifications, de-obfuscations, etc., and answers to the questions, welcome (except from BT! :-)   ------------------------------  % Date: Wed, 02 Mar 2005 22:30:58 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> , Subject: Re: Fiorina and The Peter PrincipleB Message-ID: <1109819898.92647bffb7da66412e84d8d539086b55@teranews>  
 AEF wrote:D > Now both Curly and Carly are gone from hp. Yet Meryll Lynch favors$ > Cappellas for new CEO of hp? Huh?     @ Who benefits from mergers ? Bankers.  Curly had paid millions toF investment bankers to seek a potential buyer to save Compaq. And theseD bankers then get millions in comissions when they execute the mergerF legal and financial paperwork. Consider just the comissions to replaceE HP to HPQ on the NYSE. Consider Deutche Bank: they probably made more F money off the loans and other financial services than the value of theH HP stock they owned, so they didn't care about the Compaq merger being a1 mistake, they saw more profits in it going ahead.   G So Merryl Lynch sees dollars signs when they see Curly. They know he is I incompetant and without visions and will seek a buyer within a few years.   F Capellas succeeded where Palmer failed: He got another job and managedC to save his reputation. palmer is forever gone, with just a "guest"  appearance on AMD's board.  H Now,  to cause a few missed hertbeats: imagine if HP were to hire Robert Palmer to replace Carly.	 <ducking>    ------------------------------  $ Date: Wed, 2 Mar 2005 22:28:47 -0500# From: "John Smith" <a@nonymous.com> , Subject: Re: Fiorina and The Peter Principle, Message-ID: <wrydnSbHNv99G7vfRVn-qQ@igs.net>   JF Mezei wrote:  > AEF wrote:E >> Now both Curly and Carly are gone from hp. Yet Meryll Lynch favors $ >> Cappellas for new CEO of hp? Huh? >  > B > Who benefits from mergers ? Bankers.  Curly had paid millions toH > investment bankers to seek a potential buyer to save Compaq. And theseF > bankers then get millions in comissions when they execute the mergerH > legal and financial paperwork. Consider just the comissions to replaceG > HP to HPQ on the NYSE. Consider Deutche Bank: they probably made moretH > money off the loans and other financial services than the value of theB > HP stock they owned, so they didn't care about the Compaq merger; > being a mistake, they saw more profits in it going ahead.i > F > So Merryl Lynch sees dollars signs when they see Curly. They know heG > is incompetant and without visions and will seek a buyer within a fewn > years. >eH > Capellas succeeded where Palmer failed: He got another job and managedE > to save his reputation. palmer is forever gone, with just a "guest"h > appearance on AMD's board. >nC > Now,  to cause a few missed hertbeats: imagine if HP were to hirer! > Robert Palmer to replace Carly.i > <ducking>     H Might as well toss Ken Lay and Bernie Ebbers into the list of candidatesJ too.....after all they've run companies just as big into the ground too so0 they certainly have the experience to tackle HP.   --C OpenVMS - The classics never go out of style.....they're just neverr* appreciated by the company that owns them.   ------------------------------  % Date: Wed, 02 Mar 2005 23:03:28 -0500u( From: Bill Todd <billtodd@metrocast.net>, Subject: Re: Fiorina and The Peter Principle= Message-ID: <mKSdnUqFoaYPEbvfRVn-hw@metrocastcablevision.com>e   JF Mezei wrote:M   ...a  0   Consider Deutche Bank: they probably made moreH > money off the loans and other financial services than the value of the > HP stock they ownedl  A Let's consider them a bit more carefully than that (the SEC IIRC tC certainly did).  They were not by and large voting stock that they -G owned, but rather stock which they held for others and had an explicit rC fiduciary obligation to vote in the best interests of those others oG (rather than in their own).  The fact that they failed to isolate that  D decision from their own business decisions eventually resulted in a A fine, I believe (though probably not much more than a hand-slap).e   - bill   ------------------------------  % Date: Wed, 02 Mar 2005 23:36:33 -0500 ' From: Dave Froble <davef@tsoft-inc.com>o, Subject: Re: Fiorina and The Peter Principle0 Message-ID: <112d4horcui2rf4@corp.supernews.com>   John Smith wrote:k > JF Mezei wrote:w >  >>AEF wrote: >>E >>>Now both Curly and Carly are gone from hp. Yet Meryll Lynch favors $ >>>Cappellas for new CEO of hp? Huh? >> >>B >>Who benefits from mergers ? Bankers.  Curly had paid millions toH >>investment bankers to seek a potential buyer to save Compaq. And theseF >>bankers then get millions in comissions when they execute the mergerH >>legal and financial paperwork. Consider just the comissions to replaceG >>HP to HPQ on the NYSE. Consider Deutche Bank: they probably made moretH >>money off the loans and other financial services than the value of theB >>HP stock they owned, so they didn't care about the Compaq merger; >>being a mistake, they saw more profits in it going ahead.e >>F >>So Merryl Lynch sees dollars signs when they see Curly. They know heG >>is incompetant and without visions and will seek a buyer within a fewp >>years. >>H >>Capellas succeeded where Palmer failed: He got another job and managedE >>to save his reputation. palmer is forever gone, with just a "guest"  >>appearance on AMD's board. >>C >>Now,  to cause a few missed hertbeats: imagine if HP were to hire ! >>Robert Palmer to replace Carly.g >><ducking>h >  >  > J > Might as well toss Ken Lay and Bernie Ebbers into the list of candidatesL > too.....after all they've run companies just as big into the ground too so2 > they certainly have the experience to tackle HP.  A Might as well.  Shouldn't take them long.  HP's financial status  & shouldn't be a big challenge for them.   ------------------------------  % Date: Thu, 03 Mar 2005 00:45:27 -0500n- From: JF Mezei <jfmezei.spamnot@teksavvy.com>=, Subject: Re: Fiorina and The Peter PrincipleB Message-ID: <1109827961.77f305ee84a0daf37568d2502ba78e48@teranews>   John Smith wrote:LJ > Might as well toss Ken Lay and Bernie Ebbers into the list of candidatesL > too.....after all they've run companies just as big into the ground too so2 > they certainly have the experience to tackle HP.  - Well, Ken lay was a board member at Compaq...A  F But in fairness, there is a big difference between being way over your head and being dishonest.E  B Curly and Palmer were incompetant and didn't know how to fix theirF company. The others were dishonest and played with money as a hobby to" feed their own aperites for money.  D Ebbers was the head of a tiny asset-less long distance firm and usedH .com money to buy a real company like MCI and use all sorts of financialH tricks to hide teh fact that Worldcom should have never been able to buy anything as big as MCI.d  - BTW, does anyone know if MCI still runs VMS ?A   ------------------------------  # Date: Thu, 03 Mar 2005 01:28:51 GMTp5 From: rdeininger@mindspringdot.com (Robert Deininger)d Subject: Re: GBLSECTIONS looppL Message-ID: <rdeininger-0203052028460001@user-uinj5jj.dialup.mindspring.com>  I In article <rXrqF4r4puRp@eisner.encompasserve.org>, Kilgallen@SpamCop.netL (Larry Kilgallen) wrote:  4 >In article <+mHGgWFBvKsD@eisner.encompasserve.org>,= koehler@eisner.nospam.encompasserve.org (Bob Koehler) writes:nB >> In article <IYWdnSUs0I_eSbnfRVn-rg@megapath.net>, "Paul Dembry"  <pade@nospam.trifox.com> writes:O >>> When I boot, decwinwdows does not come up because there are not enough freehO >>> GBLSECTIONS. The message says that I need 426 and I only have 210. AutobootzN >>> offers to regen to fix this and I answer yes. It goes through the generateG >>> and update phase, the system reboots and I still only have 210 free-M >>> GBLSECTIONS. How do I increase this? I tried modifying the parameter filer >>> but to no avail. >>> Thanks,R >>> Paul >> mH >>    The offer by AUTOGEN to fix this never works.  Do a conversationalF >>    boot, put the parameter into sys$system:modparams.dat, and do an >>    autogen by hand. >eH >The offer by AUTOGEN sometimes works (it has for me - never say never). >qG >I think the flaw is in the underlying design of the ADD_xxx mechanism. # >Not that I have a better design....  I I saw no mention of the VMS version.  Substantial work was done in V7.3-2iJ to fix problems around AUTOGEN and some of the layered products and greedy IO adapters.  @ DECW was typically a victim since it is the last resource-hungryG application to start up.  DECW tries to tell Autogen about its resourcetG needs, but there were some other products (that start sooner) that takeoG significant resources without telling Autogen.  So, for example, DECW'saD additional GBLSECTIONS were getting "stolen" before DECW could claimJ them.  Without feedback, additional passes didn't help because Autogen got the same input every time.  J V7.3-2 fixed a number of problems, but there are enough complex situationsB that some problems could have slipped through testing.  Autogen isJ supposed to be able to tune for VMS, the tightly-coupled layered products,G and supported IO adapters.  It ought to only take one extra reboot, andDI everything should have enough resources to start up.  (Assuming you don'tf& have a totally memory-starved system.)  F As your workloads get more demanding and complicated, you will need toJ help Autogen, either by putting MIN_* and ADD_* hints in MODPARAMS.DAT, or by using feedback, or both.   H If Autogen isn't doing its job, make sure you aren't making things worseI by supplying bad input in MODPARAMS.DAT.  If you find cases where AutogenZF isn't performing as it should, report via your usual support channels.   ------------------------------   Date: 2 Mar 2005 18:43:17 -0600i+ From: young_r@encompasserve.org (Rob Young)  Subject: Re: IA64 unraveling...p3 Message-ID: <5BvCyiNDsIuV@eisner.encompasserve.org>w  s In article <1109432228.4461a445ebd73a65b4c09319fb329da3@teranews>, JF Mezei <jfmezei.spamnot@vaxination.ca> writes:  > Rob Young wrote:H >>         Anyhow.... it is a good business decision on IBM's part.  TheH >>         common chipset shows up in 2007, until then they use existing >>         chipset.    > J > Ok, so Intel is working on a new chipset that will apply to IA64 in 2007, > ? What is the code name for this project ?  " 	Common Serial Interconnect - CSI:  m http://www.eet.com/semi/news/showArticle.jhtml;jsessionid=2OIKSKWWGRTXQQSNDBCCKHSCJUMEKJVN?articleID=60404677.   				Robs   ------------------------------   Date: 2 Mar 2005 13:22:16 -0800  From: bob@instantwhip.comi' Subject: IBM knows itanium is a dud ...aB Message-ID: <1109798536.191539.91370@f14g2000cwb.googlegroups.com>  G think about it ... if IBM ever thought that down the road itanium wouldeD become superior to power, they would have never shut down their lineG especially w/montecito coming out ... no, I think they realize that DEC1E engineers had it right with EV8 and the reasons they gave why itaniumeE would not succeed, and they smartly picked up where Compaq/HP dropped1D alpha EV8, and now they will dominate the high end server market forE years to come ... NIH (not invented here) syndrome will spell the end A for HP ... they continue to now insist the alpha teams version ofsD tukwila is bad and that they can do better ... they will never admit0 failure, and that will lead to their failure ...   ------------------------------   Date: 2 Mar 2005 14:06:42 -0600t4 From: kuhrt@nospammy.encompasserve.org (Marty Kuhrt)S Subject: RE: Interesting error message I just got trying to access accuwea	ther.com-3 Message-ID: <e4$oQhI63H5W@eisner.encompasserve.org>5  J In article <d025g8$dnu$1@news.mdx.ac.uk>, david20@alpha2.mdx.ac.uk writes: > In article <DA4AD590CAF06845B671C398333A89C6082E2662@ohms.electric.ci.austin.tx.us>, "Stuart, Ed" <Ed.Stuart@austinenergy.com> writes:J >> Interestingly enough, the accuweather servers are not reporting what OS) >>they are running according to Netcraft.e >>4 >>OS  Server Last changed IP address Netblock Owner C >>unknown  OSU/3.9c;UCX  9-Nov-2003  207.242.93.24   Accuweather   V> >>unknown  unknown  8-Nov-2003  207.242.93.24   Accuweather   D >>unknown  OSU/3.9c;UCX  28-Mar-2002  207.242.93.24   Accuweather   F >>unknown  OSU/3.8aw2;UCX  26-Feb-2002  207.242.93.24   Accuweather   D >>unknown  OSU/3.9c;UCX  25-Feb-2002  207.242.93.24   Accuweather   F >>unknown  OSU/3.8aw2;UCX  24-Feb-2002  207.242.93.24   Accuweather   D >>unknown  OSU/3.9c;UCX  22-Feb-2002  207.242.93.24   Accuweather   J >>Compaq Tru64  OSU/3.8aw2;UCX  6-Mar-2001  207.242.93.24   Accuweather   M >>Compaq Tru64  Microsoft-IIS/5.0  5-Mar-2001  207.242.93.24   Accuweather   fH >>Compaq Tru64  OSU/3.8aw2;UCX  3-Mar-2001  207.242.93.24   Accuweather  >> > 3 > Well the earlier entries are interesting as well.i > - > OSU running on Tru64 with UCX tcpip stack ?cK > I think I recall you could build OSU for Tru64 but I didn't think anybody  > had really used it. 5 > And what about running Microsoft-IIS/5.0  on Tru64.a >   E Netcraft freely admits that it can't figure out uptimes, some server t; combinations, and what not, for various OSs.  Start with...   2 http://uptime.netcraft.com/up/accuracy.html#uptime   and read their take on it.   ------------------------------  # Date: Thu, 03 Mar 2005 01:29:01 GMToL From: winston@SSRL.SLAC.STANFORD.EDU (Alan Winston - SSRL Central Computing)S Subject: RE: Interesting error message I just got trying to access accuwea	ther.come6 Message-ID: <00A402D5.C9689C32@SSRL.SLAC.STANFORD.EDU>  J In article <d04di9$6q0$1@news.mdx.ac.uk>, david20@alpha2.mdx.ac.uk writes: >In article <00A40269.579088D1@SSRL.SLAC.STANFORD.EDU>, winston@SSRL.SLAC.STANFORD.EDU (Alan Winston - SSRL Central Computing) writes:L >>In article <d048nd$4mq$1@news.mdx.ac.uk>, david20@alpha2.mdx.ac.uk writes: >>>In article <00A40217.187E5C08@SSRL.SLAC.STANFORD.EDU>, winston@SSRL.SLAC.STANFORD.EDU (Alan Winston - SSRL Central Computing) writes: >>>>In article <DA4AD590CAF06845B671C398333A89C6082E2662@ohms.electric.ci.austin.tx.us>, "Stuart, Ed" <Ed.Stuart@austinenergy.com> writes:M >>>>> Interestingly enough, the accuweather servers are not reporting what OSo, >>>>>they are running according to Netcraft. >>>>>27 >>>>>OS  Server Last changed IP address Netblock Owner hF >>>>>unknown  OSU/3.9c;UCX  9-Nov-2003  207.242.93.24   Accuweather   A >>>>>unknown  unknown  8-Nov-2003  207.242.93.24   Accuweather   eG >>>>>unknown  OSU/3.9c;UCX  28-Mar-2002  207.242.93.24   Accuweather    I >>>>>unknown  OSU/3.8aw2;UCX  26-Feb-2002  207.242.93.24   Accuweather   NG >>>>>unknown  OSU/3.9c;UCX  25-Feb-2002  207.242.93.24   Accuweather   hI >>>>>unknown  OSU/3.8aw2;UCX  24-Feb-2002  207.242.93.24   Accuweather   OG >>>>>unknown  OSU/3.9c;UCX  22-Feb-2002  207.242.93.24   Accuweather   tM >>>>>Compaq Tru64  OSU/3.8aw2;UCX  6-Mar-2001  207.242.93.24   Accuweather   yP >>>>>Compaq Tru64  Microsoft-IIS/5.0  5-Mar-2001  207.242.93.24   Accuweather   K >>>>>Compaq Tru64  OSU/3.8aw2;UCX  3-Mar-2001  207.242.93.24   Accuweather t >>>> >>>>M >>>>That's kind of the fault of the OSU webserver, which doesn't provide that-M >>>>information and doesn't give you a way to configure that information intolJ >>>>the reply stream.  You'd have to change the source code and recompile. >>>>7 >>>As I reported in a previous posting that isn't true.oC >>>Netcraft reports that I'm running OpenVMS with OSU 3.10a and UCXd >>>see >>>h= >>>http://uptime.netcraft.com/up/graph/?host=alpha2.mdx.ac.ukb >>>9 >>> N >>>It also shows historical records which also show that I was running OSU 3.8 >>>with UCX on OpenVMS in 2002.  >>O >>NetCraft may have been beaten up enough to recognize OSU as being on VMS, bute' >>look at what OSU 3.10alpha gives you:I >>+ >>HTTP/1.0 404 200 Sending data and caching  >>MIME-version: 1.0t >>Server: OSU/3.10alpha;UCXs >>Content-type: text/htmle! >>Content-transfer-encoding: 8bits. >>Last-Modified: Tue, 05 Feb 2002 13:34:45 GMT >>Content-length: 256e% >>Date: Wed, 02 Mar 2005 12:08:08 GMTh >> >>6 >>Do you see "VMS" or "OpenVMS" in the Server: header? >>N >Sorry all I meant was that Netcraft can either determine or has been setup to2 >assume that anyone running OSU is running on VMS.N >I fully accept that OSU doesn't return that information in the server header.  ( Then I think we have vehement agreement.   -- Alan    ------------------------------   Date: 2 Mar 2005 12:48:37 -0800  From: icerq4a@spray.se/ Subject: Re: new Itanium after Tukwila: Poulson C Message-ID: <1109796517.553118.256250@l41g2000cwc.googlegroups.com>5   Roger Ivie wrote:m; > On 2005-03-02, icerq4a@spray.se <icerq4a@spray.se> wrote:R > >C > > Dave Froble wrote:E > >> In 2001 when Compaq killed Alpha, the main reason was that IA-64w WAS to= > >> be an inexpensive product usable from the desktop to the@ enterprise.. > >W > > Not desktop, > D > http://groups-beta.google.com/group/comp.arch/msg/6fbf60d0494a5264  ; I was specifically referring to the year 2001 announcement.s   ------------------------------   Date: 2 Mar 2005 19:20:04 -0600h+ From: young_r@encompasserve.org (Rob Young)p/ Subject: Re: new Itanium after Tukwila: Poulson 3 Message-ID: <w8c4AsktSTGt@eisner.encompasserve.org>   \ In article <1109659668.323757.258880@l41g2000cwc.googlegroups.com>, icerq4a@spray.se writes: > JF Mezei wrote:: >> icerq4a@spray.se wrote: >> >4 >> > http://news.zdnet.com/2100-9584_22-5593047.html >>H >> Intel would have been much better off announcing this last year whichG >> would have help counter all the bad news that gave much steam to theu* >> rumours that IA64 wasn't going to last. > + > I think Tukwila was sufficient last year.e > D >> It will be interesting to see how many details Intel will release > today.H >> IA64 is still one generation behind (Power has had dual core for some	 >> time).e > 1 > Yes, being late on dual-core has not been good.  > C >> But that still doesn't give much trust in IA64's future. EV8 was $ >> announced but never materialised. > I > You have in the past said that there was nothing beyond 2007, now there)H > is, but as expected there is always a negative side on everything from > your side. >   B 	I would say "always" is a bit strong.  Now if you would have said' 	for the last 10 years, I'd agree.  HA!r  5 From: mezei...@eisner.decus.org (Jean-Francois Mezei)s" Subject: Meaning of OpenVMS vs VMS Date: 1996/02/04  I If you hear someone say VMS, it means that they have not succombed to theaL meaningless marketing fluff that harmed VMS more than it helped by confusing5 customers who were already fearing the death of VMS. e   ---s   				Robg   ------------------------------  % Date: Wed, 02 Mar 2005 20:35:59 -0500y' From: Dave Froble <davef@tsoft-inc.com>s/ Subject: Re: new Itanium after Tukwila: Poulson.0 Message-ID: <112cpv9ge041r0d@corp.supernews.com>   icerq4a@spray.se wrote:d > Dave Froble wrote: >  >>icerq4a@spray.se wrote:i >> >>E >>>What I try to say is that the main reasons for large L3 caches ares >  > to > 5 >>>alleviate for a less powerful system architecture.e >> >>I must confess to curosity.s >>C >>In the above you're admitting that you think the itanic isn't the. >  > best > 
 >>of designs.  >  > > > Yes, it's bus/system architecture is not the best out there.F > That is pretty obvious and the main reason why for example POWER5 is > ahead. >  > F >>Another post claimed that when HP's technical people saw the success >  > of > D >>DEC and IBM with out-of-order designs they advised management that >  > EPIC > F >>should be dropped, and that sticking with it was a management, not a >>technical decision.m >  >  > I have not seen that. > > I guess lots of people have had opinions of EPIC many times. >  > 1 >>How many applications run pretty much in cache?  >  >  > Very difficult to answer.d >  > < >>How many users tweek and optimize the compilation of their >  > applications?n > # > It varies very much, hard to say.b >  > 1 >>How many users buy computers to run benchmarks?c >  >  > Not many.a >  > A >>Itanic is a rather poor design that due to state of the art CPU E >>manufacturing will allow most people to run their applications with 6 >>acceptable performance, in spite of the poor design. >  > ) > I don't think Itanium is a poor design.d >  > D >>So my question is, just what is it about this CPU design that gets# >>people to advocate and defend it?d >  > E > I don't know, perhaps because it is different, and some people like F > diversity. It is pretty fun to read the IA64 manuals and think about' > how things can be done and optimized.  > F > I advocate a bit because many people portray it as particularly bad,E > sometimes for reasons other than technical. That is not what I haveu > seen when coding for IA64. > H > I have never had any illusions that Itanium would outperform anything.C > I think it is doing a very good job for being an in-order design.a >   E No, IA-32 is an example of that, and with a few exceptions for which P/ IA-32 is a poor choice, it out-performs itanic.l  B If VAX would have been continued, with performance enhanced using I methods similar to what was done with the N-VAX, that would have been an a" example of a good in-order design.   ------------------------------  % Date: Wed, 02 Mar 2005 17:32:06 -0800s# From: "Tom Linden" <tom@kednos.com>>/ Subject: Re: new Itanium after Tukwila: Poulsona( Message-ID: <opsm1fvsapzgicya@hyrrokkin>  G On Wed, 02 Mar 2005 20:35:59 -0500, Dave Froble <davef@tsoft-inc.com>  v wrote:   > icerq4a@spray.se wrote:o >> Dave Froble wrote:  >> >>> icerq4a@spray.se wrote:> >>>o >>>oG >>>> What I try to say is that the main reasons for large L3 caches aree >>  to >>7 >>>> alleviate for a less powerful system architecture.K >>>y >>> I must confess to curosity.o >>>dE >>> In the above you're admitting that you think the itanic isn't theg >>  best >> >>> of designs.hA >>   Yes, it's bus/system architecture is not the best out there.tG >> That is pretty obvious and the main reason why for example POWER5 is.	 >> ahead.- >>H >>> Another post claimed that when HP's technical people saw the success >>  of >>F >>> DEC and IBM with out-of-order designs they advised management that >>  EPIC >>H >>> should be dropped, and that sticking with it was a management, not a >>> technical decision.L >>   I have not seen that.? >> I guess lots of people have had opinions of EPIC many times.u >>3 >>> How many applications run pretty much in cache?h >>   Very difficult to answer. >>> >>> How many users tweek and optimize the compilation of their >>  applications?n% >>  It varies very much, hard to say.d >>3 >>> How many users buy computers to run benchmarks?n >>   Not many. >>C >>> Itanic is a rather poor design that due to state of the art CPUaG >>> manufacturing will allow most people to run their applications withf8 >>> acceptable performance, in spite of the poor design., >>   I don't think Itanium is a poor design. >>F >>> So my question is, just what is it about this CPU design that gets% >>> people to advocate and defend it?dH >>   I don't know, perhaps because it is different, and some people likeG >> diversity. It is pretty fun to read the IA64 manuals and think abouts( >> how things can be done and optimized.H >>  I advocate a bit because many people portray it as particularly bad,F >> sometimes for reasons other than technical. That is not what I have >> seen when coding for IA64.LJ >>  I have never had any illusions that Itanium would outperform anything.D >> I think it is doing a very good job for being an in-order design. >> > H > No, IA-32 is an example of that, and with a few exceptions for which  1 > IA-32 is a poor choice, it out-performs itanic.d > E > If VAX would have been continued, with performance enhanced using  sL > methods similar to what was done with the N-VAX, that would have been an  $ > example of a good in-order design.  H Not only that, but as has been recently discussed here, the bloat factorF between VAX and Itanic is probably around 6, and that coupled with theI relatively slower main memory access, causes some concern for performance>K of large applications.  I have no doubt that the usual marketing benchmarks  will be impressive.a           -- sC Using Opera's revolutionary e-mail client: http://www.opera.com/m2/f   ------------------------------  $ Date: Wed, 2 Mar 2005 15:38:06 -0500# From: "John Smith" <a@nonymous.com>sA Subject: Re: OpenVMS Seminar in Toronto (2005-02-24) a few points., Message-ID: <E7CdnbzNxcmyubvfRVn-qA@igs.net>   Keith Cayemberg wrote: > Peter Weaver wrote:  >u >> Neil Rieck wrote: >>* >>> <bob@instantwhip.com> wrote in message@ >>> news:1109550872.067891.88350@o13g2000cwo.googlegroups.com... >>>lG >>>> Oracle would create a band of very unhappy customers if they triedd >>>> to eol rdb  >>>> >>>pE >>> The following interesting Oracle RDB points were mentioned at thee >>> seminar: >>> ...  >> >>C >> One other point from Norman's session that I recall is that on auF >> slide he quoted Larry Ellison as saying that the way Oracle treatedC >> the RDB customers is a good example of how to treat customers ofpF >> products a company buys. So I do not think there would be anyway heF >> would EOL RDB after making a comment like that. Unfortunately I didG >> not write down the exact quote since I figured I would get it on ther >> slides later. >> >  > 8 > Oracle Praised By Thousands of Satisfied Rdb Customers4 > http://www.oracle.com/corporate/press/2337269.html >r > Nine Years Later - PDF5 > http://www.oracle.com/peoplesoft/Rdb_CaseStudyE.pdfi >n& > An Oracle Rdb-centric VMS TUD TidbitC > http://www.shannonknowshpc.com/stories.php?story=04/10/10/1936825h > # > Oracle Rdb Statement of Directione >rL http://www.oracle.com/technology/products/rdb/htdocs/rdb7/rdb_statement_of_d
 irection.html  >r >u@ > The above shows there is plenty of collateral stating Oracle'sD > commitment to Rdb. I would say it is also partly due the continualC > industry-leading innovation of its development team and quick newRH > feature release cycle despite having 3 development trees (7.0, 7.1 andE > 7.2). The Oracle Classic team has over 90 platforms to contend with,G > providing (in many aspects) a lowest-common denominator functionality>G > (Oracle Classic integration of the DLM on OpenVMS is an example where>B > this has not been the case). But Oracle Classic has a completelyF > different DB engine model, and has not yet been able incorporate (orD > make as good use of) many of the features and advantages developedE > originally and often exclusively for Rdb. Some features were mostlysE > unnoticed by the public for many years before they became buzzwordsa > for competing products.o > = >      - Snapshot coordination of serialized transaction data0  >      - Cost-Based Optimization& >      - precompiled optimized queries >      - optimizer hintsC >      - BLOBs (segmented strings) since 1984 (became buzzword c.a.B > 1990)o >      - bit-mapped indicesu* >      - latches (ultra-efficient locking)> >      - BLASTs used for efficient SW-interrupt-based resource > coordination >      - pseudo-ranked indicesB >      - transaction- and quiet point-aware on line backup utility9 >      - true fully-shared clustering and DLM integrationq* >      - cluster-aware rollback resolutionG >      - true default "serializable read transaction isolation" withouteB >        needing an additional TP Monitor front-end, despite and a= >        necessary consequence of Rdb's ultra-scalable directhE >        per-process DB access model (compare Oracle Classic's served % >        data resource bubble model).-C >      - log-based (AIJ) ultra-low transaction overhead Hot-Standbym >        database capability; >      - log-mining (AIJ) transaction replication interfaceg >  >       and more...o > ? > Rdb has also always been at the forefront of implementing SQLeC > standards, going far beyond the basic "Entry-Level" SQL Standardsn. > Compliance of almost all other SQL products.    G So all Larry really needs to go along with his DEC-developed profitablewF world-class rdbms is his very own DEC-developed profitable world-class/ operating system. Wonder if he'd be interested.a   --I OpenVMS - The classics never go out of style....they just aren't marketed9 anymore.   ------------------------------  # Date: Thu, 03 Mar 2005 01:26:59 GMT L From: winston@SSRL.SLAC.STANFORD.EDU (Alan Winston - SSRL Central Computing)A Subject: Re: OpenVMS Seminar in Toronto (2005-02-24) a few points.6 Message-ID: <00A402D5.80EA0095@SSRL.SLAC.STANFORD.EDU>  R In article <7dqdnQEBgc-iWbjfRVn-iQ@igs.net>, "John Smith" <a@nonymous.com> writes: >Neil Rieck wrote:) >> <bob@instantwhip.com> wrote in message2? >> news:1109550872.067891.88350@o13g2000cwo.googlegroups.com... F >>> Oracle would create a band of very unhappy customers if they tried >>> to eol rdb >>>7D >> The following interesting Oracle RDB points were mentioned at the >> seminar:3 >>F >> 1) 50% of the code currently in RDB was put there since the sale of* >> the RDB division from Digital to Oracle >  >cI >No surprise there.... Code that big always has bugs that need fixing andnK >adaptations to changes in the os, and it's it's been 10 years of tinkeringn2 >since Oracle was 'given' Rdb at a firesale price.  K That seems unnecessarily dismissive.  Rdb development has suffered somewhat.M from diversions like Tru-64 development (nobody wanted to buy it) and WindowseK development, but there a lot of new features, performance enhancements, new0K syntax, etc, etc, since the Oracle near-giveaway.  Multifile databases, roweI cache, two flavors of JDBC server, SQL*Net support, etc, etc - lot of newt features since 6.2..  M (And it's worth remembering that Oracle got just about all of the Digital Rdb9K development team and built them a workplace near their old one, rather than63 requiring all development move across the country.).  I I wouldn't argue with a charge  that Oracle's posture toward Rdb has been0L benign neglect - leaving the Rdb team to get on with it so long as they keepB making their revenue goals - when it could have been promoted moreG aggressively.  But I reject what I read as your imputation that Rdb hasA3 essentially been in maintenance mode for ten years..   -- Alans   ------------------------------  % Date: Wed, 02 Mar 2005 20:45:42 -0500n' From: Dave Froble <davef@tsoft-inc.com>oA Subject: Re: OpenVMS Seminar in Toronto (2005-02-24) a few points20 Message-ID: <112cqhdmkb0p42f@corp.supernews.com>   John Smith wrote:n  I > So all Larry really needs to go along with his DEC-developed profitablenH > world-class rdbms is his very own DEC-developed profitable world-class1 > operating system. Wonder if he'd be interested.>  G Well, that would definitely be the end of any type of 'affordable VMS'.>   ------------------------------  $ Date: Wed, 2 Mar 2005 22:39:21 -0500# From: "John Smith" <a@nonymous.com>0A Subject: Re: OpenVMS Seminar in Toronto (2005-02-24) a few pointsF, Message-ID: <Aa2dnfVyab13G7vfRVn-tw@igs.net>  , Alan Winston - SSRL Central Computing wrote:; > In article <7dqdnQEBgc-iWbjfRVn-iQ@igs.net>, "John Smith"5 > <a@nonymous.com> writes: >> Neil Rieck wrote:* >>> <bob@instantwhip.com> wrote in message@ >>> news:1109550872.067891.88350@o13g2000cwo.googlegroups.com...G >>>> Oracle would create a band of very unhappy customers if they tried  >>>> to eol rdby >>>>E >>> The following interesting Oracle RDB points were mentioned at thev >>> seminar: >>>rG >>> 1) 50% of the code currently in RDB was put there since the sale of + >>> the RDB division from Digital to Oraclen >> >>G >> No surprise there.... Code that big always has bugs that need fixing/G >> and adaptations to changes in the os, and it's it's been 10 years of > >> tinkering since Oracle was 'given' Rdb at a firesale price. >OD > That seems unnecessarily dismissive.  Rdb development has suffered
 > somewhatG > from diversions like Tru-64 development (nobody wanted to buy it) andPC > Windows development, but there a lot of new features, performancerE > enhancements, new syntax, etc, etc, since the Oracle near-giveaway.cE > Multifile databases, row cache, two flavors of JDBC server, SQL*Net-4 > support, etc, etc - lot of new features since 6.2. > C > (And it's worth remembering that Oracle got just about all of theyD > Digital Rdb development team and built them a workplace near their@ > old one, rather than requiring all development move across the > country.)U >eF > I wouldn't argue with a charge  that Oracle's posture toward Rdb has > beenD > benign neglect - leaving the Rdb team to get on with it so long as@ > they keep making their revenue goals - when it could have been > promoted moreTE > aggressively.  But I reject what I read as your imputation that Rdbg9 > has essentially been in maintenance mode for ten years.q    K I know that there has been a lot of work on it, and VMS too. My glibness in1I my post was directed more towards statements of performance that were not 0 presented with factual evidence to back them up.  7 Fastest commecial database can mean a lot of things....o* - fastest Rdb database on a single machineL - fastest rdbms on a cluster of 1280's maxed out with memory, cpu's, and the fastest storage going-L - fastest rdbms on a single processor tied to a Patriot 2 missile 15 seconds after launch  > No presentation of experimental method, no claim to greatness.   --- OpenVMS - The classics never go out of style.1   ------------------------------  # Date: Thu, 03 Mar 2005 06:28:08 GMTi1 From: drwho8__NOTME__@att.net (The Eighth Doctor):/ Subject: Re: Problem with a DEC terminal server5H Message-ID: <Y7yVd.315414$w62.313205@bgtnsc05-news.ops.worldnet.att.net>  > In article <b8cd107b.0503020913.19b08bd0@posting.google.com>,  pfleging.lka@ekkw.de says... >sF >The DECserver 90L+ will work as a standalone unit with an external 5VB >power supply, DEC number H7082-AA (110 VAC) or H7082-BA (220 VAC) > @ >You can also install the DECserver in a DEChub 90 or DEChub 900 > 1 >Access is then possible via ThinWire (DEChub90),bA >via Twisted pair: DECrepeater 90T (unshielded), DECrepeater 90T+o >(shielded and unshielded)% >or via fiber-optic: DECrepeater 90FSn >rC >by the way, we have four DEChub90 and one DEChub900 that we do notg, >need anymore, but we are located in germany >h > : >pfleging.lka@ekkw.de (pfleging.michael) wrote in message 9 news:<b8cd107b.0503010035.4458307c@posting.google.com>...nN >> If you have VMS, you can reprogram your Terminalserver via Console Connect:& >> example, from documentation of NCP: >> t >> $ MC NCP  >> =; >> NCP>CONNECT VIA UNA-0 PHYSICAL ADDRESS 08-00-2B-22-45-CD  >>  C >> replace UNA-0 with your Network device (NCP>SHOW KNOWN CIRCUITS)rD >> and 08-00-2B-22-45-CD with the MAC Address of your DECserver 90L+ >> 0H >> you will get a session (port 9) on your Terminal Server, then you can? >> see and reprogram the port settings according to your manualc >> t@ >> drwho8__NOTME__@att.net (The Eighth Doctor) wrote in message B news:<BZaUd.81854$Th1.81032@bgtnsc04-news.ops.worldnet.att.net>... >> > Hello from Gregg C LevineP >> > I have here a DECserver 90L+. According to the manual which I have here, I N >> > connected a PC to one of its ports, using a temporary plug. Then set the 	 terminal eM >> > software to a default of 9600,8,n,1. (Which I suspect isn't the correct t settings.) And  S >> > then when I turned the T/S on, and pressed enter, twice, instead of getting a i prompt -P >> > sequence, I got garbage. Does anyone know the correct way to do a "factory  reset" e >> > for the thing? 
 >> > -----$ >> > Gregg drwho8 atsign att dot net$ >> > "This signature isn't at home."   Hello from the Eighth Doctor Thank you for the advice.0  P Your being in Germany isn't a problem, I think. What are the shipping rates for / something like that from where you are, to me? z  ----- Gregg drwho8 atsign att dot net6 "This signature isn't at home."    ------------------------------  % Date: Wed, 02 Mar 2005 19:25:36 +0000 # From: issinoho <issinoho@gmail.com>p  Subject: Re: vms mail forwarding/ Message-ID: <112c4pio92qfb1@corp.supernews.com>    Binh Nguyen wrote:L > Sorry, all this is new to me.  I'll look into Majordomo and DELIVER today,% > but if you could, please elaborate.t > F > Alternatively, is there a way for me to send a message from my PC to9 > chef@chocolate.com and have the mail get to vms::chef ? D > My work setup is: mail from my PC to my VMS machine to my client's > VMS machine. >  > Thanks.  Binh  > ] > "Hoff Hoffman" <hoff@hp.nospam> wrote in message news:QY3Vd.795$Q14.731@news.cpqcorp.net...f > R >>In article <1%2Vd.48410$uc.23144@trnddc08>, "Binh Nguyen" <binh@egh.com> writes:; >>:Is there a way I can send mail to a vms mail account and A >>:have that account forward the message to a distribution list ?  >> >>  Not directly.n >>K >>  You can use a mail server or other such technology, of course, and thisnJ >>  is definitely the approach I would use -- Majordomo or such allows for7 >>  easy management of mail and subscriptions and such.a >>M >>  As for the OpenVMS MAIL client, there has been an attempt made to prevent L >>  forwarding mail to a distribution list because there is no way to handleK >>  the expected partial failures of such an operation.   There can be only'G >>  one target for a MAIL forward, and it can't be a distribution list.  >>K >>  This has been discussed before, of course -- more than a few folks have;K >>  wanted to try this, and the classic Mail-11 protocol just doesn't allowpJ >>  for it to work correctly.  There is no "correct" implementation choice4 >>  here, obviously, given the Mail-11 restrictions. >>= >>  The Freeware DELIVER package is another potential option.  >>P >> ---------------------------- #include <rtfaq.h> -----------------------------M >>    For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faq P >> --------------------------- pure personal opinion ---------------------------G >>        Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.comS >> >  >  > D My domain mail gets collected on my PC by MDaemon which has account B forwarding enabled for my chosen VMS mail address. This sends all C messages to my VMS box which has TCPWare SMTP server running which YC automagically puts the mail into the correct VMSMail account inbox.   E This also works in reverse: the TCPWare SMTP server is configured to fI forward external mail to my PC and MDaemon does the rest. Note I have to tE use the TCPWARE_SMTP_REPLY_TO logical to ensure the email address is 0 preserved in this direction.  G One other nicety is that enabling the TCPWare IMAP server allows me to eF connect to my VMSMail account from my PC using Thunderbird (or indeed 3 use Mozilla Mail to read it from my Alpha desktop).t   I.   ------------------------------  % Date: Wed, 02 Mar 2005 18:14:19 -0500r- From: JF Mezei <jfmezei.spamnot@teksavvy.com>   Subject: Re: vms mail forwardingB Message-ID: <1109804517.06097f4d7ed93b24f3f6f70a841583ec@teranews>   Binh Nguyen wrote: > A > I'm sorry TCPIP$SMTP_COMMON is not defined on my system.  Wherei > should it be ?  F If that logical doesn't excist on your system, it is likely you have aF different TCPIP software (Multinet, Process) instead of the Digital/HP- one so my previous suggestions may not apply.s   ------------------------------  * Date: Wed, 2 Mar 2005 22:09:10 +0000 (UTC)P From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply)E Subject: Re: What is the Difference between Shadow and Mirrored disk?t$ Message-ID: <d05di6$s9b$1@online.de>  6 In article <38lkqfF5os0vvU1@individual.net>, Roy Omond& <Roy.Omond@BlueBubble.UK.Com> writes:   H > Sorry to harp on about this, but since it's coming from an HP address,F > I'd just like to add the comment that using a *local* quorum disk asA > suggested is not only unnecessary, but actually *increases* the $ > probability of losing the cluster. > . > Node A (1 vote) + local quorum disk (1 vote) > Node B (1 vote)i >  > Node B goes down ... > C > Now you have introduced an unnecessary probability of the cluster I > being lost, equal to the probability of the quorum disk "disappearing".0  F In the situation above, if the quorum disk disappears, then there are G still 2 out of 3 votes there, so the cluster will survive.  The reason  E this is pointless, though, is since if node A goes down, so does the eI disk, so the above is equivalent to giving node A two votes and node B 1 . vote.   H With a two-node cluster without a "proper" quorum disk (which is really H a three-node cluster with the disk taking the place of the third node), G there is NO WAY to have a real cluster, whether or not a "local quorum (H disk" is used and however the votes are distributed.  Either you have a I 50-50 situation: cluster freezes if either node goes down, or you have a -I master-slave situation, where one node must ALWAYS be up for the cluster m7 to be up and it doesn't matter what the other one does.t  F I really fail to see why people these days even bother with less than F optimal solutions.  Just get a cheap VAX and use it as a quorum node;  problem solved.f   ------------------------------   End of INFO-VAX 2005.123 ************************