1 INFO-VAX	Thu, 03 Mar 2005	Volume 2005 : Issue 124       Contents: Carly's Way  Re: Carly's Way ; DCPS printing of ANSI files (where does it get the fonts ?) ? Re: DCPS printing of ANSI files (where does it get the fonts ?) & 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& 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& Re: DECC : toupper/tolower performance& Re: DECC : toupper/tolower performance& Re: DECC : toupper/tolower performance  Re: 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: 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: GBLSECTIONS loop- Re: Identify the process locking a RMS record  Re: Logfile name& 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 points
 Oracle RDB Re: Oracle RDB Re: Oracle RDB Running OpenVMS on DEC PWS 500A # Re: Running OpenVMS on DEC PWS 500A / setting default protection for CDD+ directories  Re: sysgen and decw problem  Re: sysgen and decw problem 1 Re: This is the End... My Only Friend, The End... 1 Re: This is the End... My Only Friend, The End... < Re: What is the Difference between Shadow and Mirrored disk?< Re: What is the Difference between Shadow and Mirrored disk?  F ----------------------------------------------------------------------  $ Date: Thu, 3 Mar 2005 07:12:08 -0500) From: "Neil Rieck" <n.rieck@sympatico.ca>  Subject: Carly's Way; Message-ID: <yaDVd.43201$Vf6.1205090@news20.bellglobal.com>   J I just read this cool article at Technology Review about Carly's reign of F terror (regarding the Labs at HP). Scary stuff that can happen in any = company when the board of directors puts in the wrong person.   E http://www2.technologyreview.com/articles/05/03/wo/wo_delio030405.asp   
 Neil Rieck Kitchener/Waterloo/Cambridge,  Ontario, Canada.8 http://www3.sympatico.ca/n.rieck/links/cool_openvms.html   ------------------------------  % Date: Thu, 03 Mar 2005 08:09:21 -0500 2 From: Bob Willard <BobwBSGS@TrashThis.comcast.net> Subject: Re: Carly's Way0 Message-ID: <a9idnUsk_ZL2kbrfRVn-tQ@comcast.com>   Neil Rieck wrote: L > I just read this cool article at Technology Review about Carly's reign of H > terror (regarding the Labs at HP). Scary stuff that can happen in any ? > company when the board of directors puts in the wrong person.  > G > http://www2.technologyreview.com/articles/05/03/wo/wo_delio030405.asp  >  > Neil Rieck > Kitchener/Waterloo/Cambridge,  > Ontario, Canada.: > http://www3.sympatico.ca/n.rieck/links/cool_openvms.html >  >   D Sounds like Carly will be a great fit with the anti-science folks in this administration. --   Cheers, Bob    ------------------------------  % Date: Thu, 03 Mar 2005 11:41:19 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> D Subject: DCPS printing of ANSI files (where does it get the fonts ?)B Message-ID: <1109867330.f107404b7694afc9ff63a88e44e66ae3@teranews>  F I have been trying to find a PS font to print line drawing and visible& control characters. I know they exist.  G I found TERMINAL_DECTECH.PS in the [SYSFONT] hiearchy, but it generates H postscript errors with a "known" command (testing if a character alreadyE exists in the font directory). (Has anyone been able to use this font  succesfully ?)  E So, I set out to see how DCPS manages to convert ANSI files with line G drawing etc to postscript. I traced the ethernet while DCPS was sending  a file over.  G a VT line segment would be translated to postscript as (qqqqqqqqq) 0 D4   G Unfortunatly, I can't find in the postscript data stream any definition - of D4 :-) (Maybe ethermon dropped some data).   E And looking at this DCPS$DEVCTL.TLB there is no D4 definition either.   E Does anyone have some information as to how/where D4 gets defined ? I K can't find any reference to it in any of the executables or text libraries.   E Does DCPS leave some its its stuff in the printer and when starting a C job, relaises some dictionary is already present and doesn'T bother < downloading it ? Or does it reset the printer between jobs ?   ------------------------------  # Date: Thu, 03 Mar 2005 19:11:36 GMT * From: Paul Anderson <paul.anderson@hp.com>H Subject: Re: DCPS printing of ANSI files (where does it get the fonts ?)5 Message-ID: <030320051340240000%paul.anderson@hp.com>   E In article <1109867330.f107404b7694afc9ff63a88e44e66ae3@teranews>, JF + Mezei <jfmezei.spamnot@teksavvy.com> wrote:   @ > I have been trying to find a PS font to print line drawing and > visible control characters.   F DCPS creates ANSI fonts from the printer's Courier font and loads them on the printer.   G > So, I set out to see how DCPS manages to convert ANSI files with line A > drawing etc to postscript. I traced the ethernet while DCPS was  > sending a file over.  C An easier way to see what DCPS sends to the printer is to collect a  DCPS "diag" file.   /    $ DEFINE /EXECUTIVE_MODE /SYSTEM DCPS$TEST 1 *    $ PRINT /PARAMETERS=DIAG=FILE=BOTH file    Then stop and restart the queue.  D You'll get a DCPS$JOB_entry_DIAG_TO.LOG which contains all DCPS dataD sent to the printer.  (You'll get a "from" log, too, containing data sent back from the printer.)  8 > a VT line segment would be translated to postscript as > (qqqqqqqqq) 0 D4 > > > Unfortunatly, I can't find in the postscript data stream any: > definition of D4 :-) (Maybe ethermon dropped some data).  G Power-cycle the printer before sending your job.  I think what you want @ is included in the prologues that DCPS persistently loads on the= printer.  This data would be sent only in the first job after  power-cycling the printer.   Paul   --    Paul Anderson   OpenVMS Engineering    Hewlett-Packard Company    ------------------------------   Date: 3 Mar 2005 07:01:21 GMT 2 From: "Dave Weatherall" <djw-nothere@nospam.nohow>/ Subject: Re: DECC : toupper/tolower performance ? Message-ID: <DTiotGxQ0bj6-pn2-Z3zpdKPpwKLi@dave2_os2.home.ours>   * On Tue, 1 Mar 2005 17:38:38 UTC, JF Mezei % <jfmezei.spamnot@teksavvy.com> wrote:    > E > I need to do a case insensitive comparison. I looked at the toupper J > macro and while it works, i was wondering about peformance ( it seems to > do a lot of checking). > G > Would it be faster to just have an array of 256 bytes and just lookup A > the translated character based on the value of the untranslated 
 > character ?  > % > newchar = uppercase[oldchar] ;  ???  > F > or does DECC end up taking mroe juice out of the VAX CPU for such an > operation compared to: >  > newchar = toupper(oldchar) ; > 7 > (I can setup my own translation table fairly easily).   F If its a VMS-only stuff I'd be looking at STR$UPCASE. If its portable F then, considering Warren warning about Solaris, I'd still be lookin at9 st$upcase for VMS and whatever Solaris needs for Solaris.   0 I'm sure you're using descriptor'ed strings :-)    --   Cheers - Dave W.   ------------------------------   Date: 3 Mar 2005 01:13:22 -0800  From: martinkirby12@yahoo.co.uk / Subject: Re: DECC : toupper/tolower performance A Message-ID: <1109841202.924563.5870@g14g2000cwa.googlegroups.com>    Craig A. Berry wrote: 3 > In article <960DBE4A2wspenceraporg@216.168.3.30>, 5 >  wspencer@ap.dontspamme.org (Warren Spencer) wrote:  > ? > > As documented in the man pages, Solaris' _toupper() returns 	 undefined < > > results if the character passed to it is NOT lower case. > F > If that's true, and I have no reason to doubt your account, then theB > Solaris implementation does not comply with the ANSI C standard, which @ > states that the argument is returned unmodified unless it is a< > character in the current locale for which islower is true.  B Is this a conflict between UNIX and C standards? The UNIX standard> "IEEE Std 1003.1, 2004 Edition" has the words about "should be+ lowercase" in the definition of "_toupper".    Martin Kirby   ------------------------------   Date: 3 Mar 2005 01:16:17 -0800  From: martinkirby12@yahoo.co.uk / Subject: Re: DECC : toupper/tolower performance C Message-ID: <1109841377.136532.226790@o13g2000cwo.googlegroups.com>   9 I have no idea whether the document or the help is wrong.   F Note that there is no performance gain (in fact a performance hit) for8 using the macro format if the current locale is not "C".   Martin Kirby   ------------------------------   Date: 3 Mar 2005 01:26:50 -0800  From: martinkirby12@yahoo.co.uk / Subject: Re: DECC : toupper/tolower performance C Message-ID: <1109842010.496550.258500@g14g2000cwa.googlegroups.com>   G As Fred pointed out, a problem with STR$UPCASE is it doesn't know about  locales.  F In practice, between ISO-LATIN-1 and DEC Multinational character set IG think the only problem is code values 215 and 247. In ISO-LATIN-1 these E are multiplication and division signs. In DEC MNCS they are upper and D lower case OE ligatures (I think). This seems not to matter for thisB case because the strings are names. It is a pain when dealing with, ODS-5 file specs since they are ISO-LATIN-1.  A ISO-LATIN-1 defines various characters (mostly Icelandic) in code C values that DEC MNCS does not define but the STR$ functions seem to * treat them in the way ISO-LATIN-1 expects.   Martin Kirby   ------------------------------  % Date: Thu, 03 Mar 2005 07:58:43 -0600 6 From: "Craig A. Berry" <craigberry@mac.com.spamfooler>/ Subject: Re: DECC : toupper/tolower performance D Message-ID: <craigberry-0BBDD3.07584303032005@news.isp.giganews.com>  A In article <1109841202.924563.5870@g14g2000cwa.googlegroups.com>, !  martinkirby12@yahoo.co.uk wrote:    > Craig A. Berry wrote: 5 > > In article <960DBE4A2wspenceraporg@216.168.3.30>, 7 > >  wspencer@ap.dontspamme.org (Warren Spencer) wrote:  > > A > > > As documented in the man pages, Solaris' _toupper() returns  > undefined > > > > results if the character passed to it is NOT lower case. > > H > > If that's true, and I have no reason to doubt your account, then theD > > Solaris implementation does not comply with the ANSI C standard, > which B > > states that the argument is returned unmodified unless it is a> > > character in the current locale for which islower is true. > D > Is this a conflict between UNIX and C standards? The UNIX standard@ > "IEEE Std 1003.1, 2004 Edition" has the words about "should be- > lowercase" in the definition of "_toupper".   E My mistake.  I assumed the macro and the function would have exactly  H the same behavior, but they don't.  C defines the function, but not the H macro.  POSIX defines both and specifies the difference between the two  like so:  A "The _toupper() macro is equivalent to toupper() except that the  ( argument c must be a lower-case letter."   See   @ http://www.opengroup.org/onlinepubs/7990989775/xsh/_toupper.html  ? http://www.opengroup.org/onlinepubs/7990989775/xsh/toupper.html    ------------------------------  % Date: Thu, 03 Mar 2005 06:09:02 -0800 # From: "Tom Linden" <tom@kednos.com> / Subject: Re: DECC : toupper/tolower performance ( Message-ID: <opsm2excpgzgicya@hyrrokkin>  @ On 3 Mar 2005 01:26:50 -0800, <martinkirby12@yahoo.co.uk> wrote:  I > As Fred pointed out, a problem with STR$UPCASE is it doesn't know about 
 > locales. > H > In practice, between ISO-LATIN-1 and DEC Multinational character set II > think the only problem is code values 215 and 247. In ISO-LATIN-1 these G > are multiplication and division signs. In DEC MNCS they are upper and F > lower case OE ligatures (I think). This seems not to matter for thisD > case because the strings are names. It is a pain when dealing with. > ODS-5 file specs since they are ISO-LATIN-1. > C > ISO-LATIN-1 defines various characters (mostly Icelandic) in code E > values that DEC MNCS does not define but the STR$ functions seem to , > treat them in the way ISO-LATIN-1 expects. >  > Martin Kirby  H I'll ask again, How do you dynamically change character sets on the fly?C Suppose, for example that I ty[ically use the standard US English    character set,= but I want to run a program  which outputs, say, german text.    >        --  C Using Opera's revolutionary e-mail client: http://www.opera.com/m2/    ------------------------------  % Date: Thu, 03 Mar 2005 08:37:51 -0600 6 From: "Craig A. Berry" <craigberry@mac.com.spamfooler>/ Subject: Re: DECC : toupper/tolower performance D Message-ID: <craigberry-F840DD.08375103032005@news.isp.giganews.com>  G In article <opsm2excpgzgicya@hyrrokkin>, "Tom Linden" <tom@kednos.com>   wrote:  J > I'll ask again, How do you dynamically change character sets on the fly?  G See the documentation for setlocale() for starters, then read chap. 10   of the CRTL manual.    ------------------------------  % Date: Thu, 03 Mar 2005 06:54:42 -0800 # From: "Tom Linden" <tom@kednos.com> / Subject: Re: DECC : toupper/tolower performance ( Message-ID: <opsm2g1gjezgicya@hyrrokkin>  4 On Thu, 03 Mar 2005 08:37:51 -0600, Craig A. Berry  & <craigberry@mac.com.spamfooler> wrote:  H > In article <opsm2excpgzgicya@hyrrokkin>, "Tom Linden" <tom@kednos.com> > wrote: > K >> I'll ask again, How do you dynamically change character sets on the fly?  > H > See the documentation for setlocale() for starters, then read chap. 10 > of the CRTL manual.   F I was hoping to do it from dcl. Ideally a logical which could be set   either form 2 dcl or program.  Don't use CRTL, unless forced to.   --  C Using Opera's revolutionary e-mail client: http://www.opera.com/m2/    ------------------------------  % Date: Thu, 03 Mar 2005 11:09:35 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> / Subject: Re: DECC : toupper/tolower performance B Message-ID: <1109865433.6a434fd06766c12c33b8221619108385@teranews>   Tom Linden wrote:    J > I'll ask again, How do you dynamically change character sets on the fly?C > Suppose, for example that I ty[ically use the standard US English  > character set,? > but I want to run a program  which outputs, say, german text.   E If your computer language treats characters as 8 bit entities without D purposefully chipping off their high order bit, then you need not doG anything. Germans is included in ISO-LATIN-1 character set which si the 7 default on VMS (ok, DEC MCS which is nearly identical).   A If you use C, you need to use /UNSIGNED when compiling so the the F compiler will not generate the code that chops off the 8th bit when itG processes char entities. (or defined all your char entities as unsigned H char, but then the compiler complains about you feeding unsigned char to routines that expect char).    ------------------------------  % Date: Thu, 03 Mar 2005 08:38:23 -0800 # From: "Tom Linden" <tom@kednos.com> / Subject: Re: DECC : toupper/tolower performance ( Message-ID: <opsm2lt9kpzgicya@hyrrokkin>  . On Thu, 03 Mar 2005 11:09:35 -0500, JF Mezei  % <jfmezei.spamnot@teksavvy.com> wrote:    > Tom Linden wrote:  > K >> I'll ask again, How do you dynamically change character sets on the fly? D >> Suppose, for example that I ty[ically use the standard US English >> character set, @ >> but I want to run a program  which outputs, say, german text. > G > If your computer language treats characters as 8 bit entities without F > purposefully chipping off their high order bit, then you need not doI > anything. Germans is included in ISO-LATIN-1 character set which si the 9 > default on VMS (ok, DEC MCS which is nearly identical).  > C > If you use C, you need to use /UNSIGNED when compiling so the the H > compiler will not generate the code that chops off the 8th bit when itI > processes char entities. (or defined all your char entities as unsigned J > char, but then the compiler complains about you feeding unsigned char to > routines that expect char).   J I guess I didn't make myself clear, or perhaps I missed something.  if I   have4 a simple program which prints the collating sequence   FREJA> type asc.pli  asc: proc options(main); dcl c char(256); c = collate(); put skip list(c);  end;F Now if I run the Program from a PuTTY window from W2K , where I have   defined B my character set to be what Microsoft calls US Internationl  I get     FREJA> run asc                                                         123456789:;<=>?@ABCDEFGHIJKLMNO . PQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}  )  ¡¢£¤¥¦§¨©ª«¬­®¯°±²³  x ´µ¶·¸¹º»¼½¾¿ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞßàáâãäåæçèéêëìíîï  ðñòóôõö÷øùúûüýþÿ FREJA> PuTTY FREJA> PuTTY    H However If I login directly and run the same program in a DECWindow it   prints the following	 and hangs    FREJA> run asc                                 123456789:;<=>?@ABCDEFGHIJKLMNO / PQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~   	 INTERRUPT    FREJA>  K So what would be desirable is the ability to set a logical such that when    runniong the< program it could produce whatever character set you desired.     --  C Using Opera's revolutionary e-mail client: http://www.opera.com/m2/    ------------------------------  % Date: Thu, 03 Mar 2005 12:06:10 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> / Subject: Re: DECC : toupper/tolower performance B Message-ID: <1109868822.e8e7e6c0950eb0e498ea3b217fdd691d@teranews>   Tom Linden wrote: H > However If I login directly and run the same program in a DECWindow it > prints the following > and hangs  >  > FREJA> run asc > ! > 123456789:;<=>?@ABCDEFGHIJKLMNO 1 > PQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~  >  > INTERRUPT   A It is normal for it to hang since DECterm understands the control F coharacters between 127 and 160. (eccept for 4 or 5 in there which are" not mapped to a control sequence).  G Consider the <CSI> character at decimal value 156, there si also DCS at F 144 and I think ST which are all part of escape sequences that DECterm# would expect to be in valid format.   C postscript fonts use 144-159 for special characters (mostlty varous  accents without any letters).   F For instance 155 maps to the cedilla that would otherwise go under the c. ( is at 231)  J Your program should only try to output from 32 to 126 and from 160 to 255.   ------------------------------  % Date: Thu, 03 Mar 2005 09:02:03 -0800 # From: "Tom Linden" <tom@kednos.com> / Subject: Re: DECC : toupper/tolower performance ( Message-ID: <opsm2mxpg8zgicya@hyrrokkin>  . On Thu, 03 Mar 2005 12:06:10 -0500, JF Mezei  % <jfmezei.spamnot@teksavvy.com> wrote:    > Tom Linden wrote: I >> However If I login directly and run the same program in a DECWindow it  >> prints the following  >> and hangs >> >> FREJA> run asc  >>" >> 123456789:;<=>?@ABCDEFGHIJKLMNO2 >> PQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~ >> >> INTERRUPT > C > It is normal for it to hang since DECterm understands the control H > coharacters between 127 and 160. (eccept for 4 or 5 in there which are$ > not mapped to a control sequence). > I > Consider the <CSI> character at decimal value 156, there si also DCS at H > 144 and I think ST which are all part of escape sequences that DECterm% > would expect to be in valid format.  > E > postscript fonts use 144-159 for special characters (mostlty varous  > accents without any letters).  > H > For instance 155 maps to the cedilla that would otherwise go under the > c. ( is at 231) > I > Your program should only try to output from 32 to 126 and from 160 to    > 255.  H I know all that.  You still don't understand.  All the program does is   display J the character set, but note they are different in the two different ways IK demonstrated.  So my question is there a logical or other means to define    the H character set, such that when this program is run it will display that  	 character  set.  I hope that is clear.      --  C Using Opera's revolutionary e-mail client: http://www.opera.com/m2/    ------------------------------  # Date: Thu, 03 Mar 2005 17:57:39 GMT * From: "FredK" <fred.nospam@nospam.dec.com>/ Subject: Re: DECC : toupper/tolower performance 0 Message-ID: <neIVd.917$Gb5.535@news.cpqcorp.net>  F You are mixing a lot of things here.  The "traditional" multi-national issuesH for *terminal* like devices are well known.  Your program BTW should notL attempt to output in the C0 or C1 ranges - the "hang" you see is a result ofB one of the controls that requires a sequence terminator.  For ANSIG terminals, you would need to send the proper control sequence to change D the character set being displayed (for instance changing between ISOL Latin-1 and MCS).  DECterm is a fully ANSI compliant terminal implementationI and many UNIX/PC "command" windows are not - so they do not handle things   like - say OSC ... ST correctly.  H This has nothing however to do with say, a Motif application - which hasK it's own mechanisms.  A font file doesn't have to leave entries in the ANSI F control ranges blank - because it isn't defining a set of input/output values.   H Historically, and practically, shifting between different encodings with someG exceptions (like translating between 16 an 8 bit latin encodings) makes  little8 sense - you need to translate the words not the letters.      . "Tom Linden" <tom@kednos.com> wrote in message" news:opsm2mxpg8zgicya@hyrrokkin.... > On Thu, 03 Mar 2005 12:06:10 -0500, JF Mezei' > <jfmezei.spamnot@teksavvy.com> wrote:  >  > > Tom Linden wrote: K > >> However If I login directly and run the same program in a DECWindow it  > >> prints the following  > >> and hangs > >> > >> FREJA> run asc  > >>$ > >> 123456789:;<=>?@ABCDEFGHIJKLMNO4 > >> PQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~ > >> > >> INTERRUPT > > E > > It is normal for it to hang since DECterm understands the control J > > coharacters between 127 and 160. (eccept for 4 or 5 in there which are& > > not mapped to a control sequence). > > K > > Consider the <CSI> character at decimal value 156, there si also DCS at J > > 144 and I think ST which are all part of escape sequences that DECterm' > > would expect to be in valid format.  > > G > > postscript fonts use 144-159 for special characters (mostlty varous ! > > accents without any letters).  > > J > > For instance 155 maps to the cedilla that would otherwise go under the > > c. ( is at 231) > > I > > Your program should only try to output from 32 to 126 and from 160 to  > > 255. > H > I know all that.  You still don't understand.  All the program does is	 > display L > the character set, but note they are different in the two different ways IK > demonstrated.  So my question is there a logical or other means to define  > the H > character set, such that when this program is run it will display that > character  > set.  I hope that is clear.  >  >  > --  E > Using Opera's revolutionary e-mail client: http://www.opera.com/m2/    ------------------------------  % Date: Thu, 03 Mar 2005 12:43:34 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> / Subject: Re: DECC : toupper/tolower performance B Message-ID: <1109871070.465355f7b3805b4f10391316a04b1f22@teranews>   Tom Linden wrote: L > the character set, but note they are different in the two different ways I > demonstrated.     A i got garbage from the windows output you showed, and no valuable > outpout on teh VMS box since the decterm froze just before the intersting output.  < > So my question is there a logical or other means to define > the character set,  E for character cell outpout (such as a write sys$outpout equivalejt on G your programming language), there is no logical on the host running the P program. It is defined in terminal emulator (eg: how to display arriving bytes).  & Your program just spits out raw bytes.  H However, because VMS is inherently an ISO-LATIN1 machine by default, youG should set your terminal emulators to translate remote system data from M latin1 to whatever that host uses. And make sure you use an 8 bit clean path.   C You should be able to output a 255 (0xFF) to any terminal and see a  lowercase y with an umlaut ()  G To test 8 bit cleanliness, output a  (233 or 0xE9). It if comes out as H an i (105 or 0x69), then there is something which snips all 8th bits off your bytes along the way.    ------------------------------  % Date: Thu, 03 Mar 2005 10:45:34 -0800 # From: "Tom Linden" <tom@kednos.com> / Subject: Re: DECC : toupper/tolower performance ( Message-ID: <opsm2rp8zvzgicya@hyrrokkin>  K On Thu, 03 Mar 2005 17:57:39 GMT, FredK <fred.nospam@nospam.dec.com> wrote:    > H > You are mixing a lot of things here.  The "traditional" multi-national > issuesJ > for *terminal* like devices are well known.  Your program BTW should notF > attempt to output in the C0 or C1 ranges - the "hang" you see is a   > result of D > one of the controls that requires a sequence terminator.  For ANSII > terminals, you would need to send the proper control sequence to change F > the character set being displayed (for instance changing between ISOA > Latin-1 and MCS).  DECterm is a fully ANSI compliant terminal    > implementationK > and many UNIX/PC "command" windows are not - so they do not handle things " > like - say OSC ... ST correctly. > J > This has nothing however to do with say, a Motif application - which hasJ > it's own mechanisms.  A font file doesn't have to leave entries in the   > ANSIH > control ranges blank - because it isn't defining a set of input/output	 > values.  > J > Historically, and practically, shifting between different encodings with > someI > exceptions (like translating between 16 an 8 bit latin encodings) makes  > little: > sense - you need to translate the words not the letters. >   L I am having a tough time explaining myself this morning. I want to be able   toK select character set without have to redefine the keyboard or reconfigure    the J system.  My example was bad because it was the issue was distracted by the@ control characters so lets skip those and consider the following asc: proc options(main); dcl c char(256); c = collate();% put skip list (substr(collate(),32));  end asc;  I So I am just printing character 32 through 255 of the collating sequence. I Now when I run this from a DECterm being directly attached (well thru a    KVM) it  displays following FREJA> run asc  P   !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmn opqrstuvwxyz{|}~% and here I hit ^c get the prompt back   J now on W2K I use a different char set US int'l  and logging to same system via PuTTY gives    FREJA> run asc  P   !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmn opqrstuvwxyz{|} >  ¡¢£¤¥¦§¨©ª«¬­®¯°±²³´µ¶·¸¹º»¼½¾ ¿ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ FREJA>  K So what  was trying to ask was if there was a way to change the displayed    char set on the fly?    >  > 0 > "Tom Linden" <tom@kednos.com> wrote in message$ > news:opsm2mxpg8zgicya@hyrrokkin.../ >> On Thu, 03 Mar 2005 12:06:10 -0500, JF Mezei ( >> <jfmezei.spamnot@teksavvy.com> wrote: >> >> > Tom Linden wrote:K >> >> However If I login directly and run the same program in a DECWindow    >> it  >> >> prints the following >> >> and hangs  >> >>  >> >> FREJA> run asc >> >> % >> >> 123456789:;<=>?@ABCDEFGHIJKLMNO 5 >> >> PQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~  >> >>  >> >> INTERRUPT  >> >F >> > It is normal for it to hang since DECterm understands the controlK >> > coharacters between 127 and 160. (eccept for 4 or 5 in there which are ' >> > not mapped to a control sequence).  >> >K >> > Consider the <CSI> character at decimal value 156, there si also DCS    >> at K >> > 144 and I think ST which are all part of escape sequences that DECterm ( >> > would expect to be in valid format. >> >H >> > postscript fonts use 144-159 for special characters (mostlty varous" >> > accents without any letters). >> >K >> > For instance 155 maps to the cedilla that would otherwise go under the  >> > c. (ç is at 231) >> >J >> > Your program should only try to output from 32 to 126 and from 160 to	 >> > 255.  >>I >> I know all that.  You still don't understand.  All the program does is 
 >> displayH >> the character set, but note they are different in the two different  	 >> ways I G >> demonstrated.  So my question is there a logical or other means to   	 >> define  >> theI >> character set, such that when this program is run it will display that  >> character >> set.  I hope that is clear. >> >> >> -- F >> Using Opera's revolutionary e-mail client: http://www.opera.com/m2/ >  >        --  C Using Opera's revolutionary e-mail client: http://www.opera.com/m2/    ------------------------------  % Date: Thu, 03 Mar 2005 05:59:52 -0800 # From: "Tom Linden" <tom@kednos.com> ) Subject: Re: DECC floating point question ( Message-ID: <opsm2eh21mzgicya@hyrrokkin>  G On Thu, 03 Mar 2005 00:07:10 -0500, Dave Froble <davef@tsoft-inc.com>    wrote:   > Tom Linden wrote: A >> On 2 Mar 2005 19:18:26 -0800, <damnihatethat@gmail.com> wrote:  >>A >>> 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  >>> 3 >>> 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); >>>  >>> }  >>> H >>  It appears that there is some truncation taking place, as well as   
 >> converting I >> from decimal to binary float.  I tried the same with PL/I just confirm  >> >  > <big PL/I snip>  >  >>   6.32500000000000E+04 	 >> FREJA> L >>  I can only conclude that you are using the wrong language, try fortran   >> or  >> PL/I. >> > C > (This kind of stuff from C always brings out the language bigots)   H Yes.  The real problem is that you cannot map the finite set of decimals? supported on the computer to the finite set of flaoting binary.   A PL/I supports both fixed and floating decimal, permitting exact    representationF of such numbers.  You certainly would not use C for making financial   calculations- (unless you use a decimal arithmetic package)    > L > 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  > $  > J > Now that I got in my required language promotion, I'll guess that your  J > problem is using DOUBLE (8 byte floating point) on an Alpha.  There is  H > no such thing.  Alpha converts such to IEEE floating point, does the  J > calculations, and then converts it back.  Guess what, there are 3 less  , > bits of prescision in IEEE floating point. >  > You might try: > 7 > temp = (unsigned int) ( (result * 10000.0) + .00001);  > . > That fixes my example in BASIC on the Alpha. >  > Or, get a VAX. >  > Dave       --  C Using Opera's revolutionary e-mail client: http://www.opera.com/m2/    ------------------------------  % Date: Thu, 03 Mar 2005 06:03:17 -0800 # From: "Tom Linden" <tom@kednos.com>	) Subject: Re: DECC floating point questiona( Message-ID: <opsm2enrq7zgicya@hyrrokkin>  . On Thu, 03 Mar 2005 01:00:00 -0500, JF Mezei  % <jfmezei.spamnot@teksavvy.com> wrote:d   >tB > To answer the original poster's question, one solution is to useI > integers. You want 6 digit precision, then multiply your floating point E > value by 100000 and then do your calculations as integers, and whenn. > done, divide by 100000 to print your result.  7 Wrong.  Use a language that supports decimal data type.w   -- oC Using Opera's revolutionary e-mail client: http://www.opera.com/m2/l   ------------------------------   Date: 3 Mar 2005 15:59:21 +0100t. From: huber@NOBODY-mppmu.mpg.de (Joseph Huber)) Subject: Re: DECC floating point questionw+ Message-ID: <MWZFpOtRfzHX@vms.mppmu.mpg.de>r  c In article <1109819906.558841.250000@g14g2000cwa.googlegroups.com>, damnihatethat@gmail.com writes:u? > I would really like the following code to produce the result:  > temp = 63250 > but instead it produces: > temp = 63249  4 In addition to what others said about FP arithmetic,D Your problem is not really FP precision (at least not for 4 decimal M digits), but the C language casting, which is simply truncating the mantissa.iK Other languages have ENTIER or NINT (Nearest INTeger) functions to give Youl the expected result.  ! Replace in Your example the line o,    temp = (unsigned int) (result * 10000.0); by2    temp = (unsigned int) (result * 10000.0 + 0.5);! and You get the expected result. l   -- s6    Joseph Huber, Muenchen  http://www.huber-joseph.de/   ------------------------------  % Date: Thu, 03 Mar 2005 11:44:18 -0500o' From: Dave Froble <davef@tsoft-inc.com>e) Subject: Re: DECC floating point questionM0 Message-ID: <112ef6bta4ids16@corp.supernews.com>   Tom Linden wrote:lI > On Thu, 03 Mar 2005 00:07:10 -0500, Dave Froble <davef@tsoft-inc.com>  n > wrote: >  >> Tom Linden wrote: >>B >>> On 2 Mar 2005 19:18:26 -0800, <damnihatethat@gmail.com> wrote: >>>eB >>>> I would really like the following code to produce the result: >>>> temp = 63250  >>>> >>>> but instead it produces:d >>>> temp = 63249- >>>>% >>>> I compile it on vms 7.3.1 using:- >>>>
 >>>> $cc teste >>>> $link test  >>>> $run test >>>>4 >>>> Am I wrong to expect the result I am expecting? >>>> Thanks for your input,1 >>>> >>>> DHw >>>> >>>> >>>> #include <stdio.h>h >>>> >>>> main()T >>>> { >>>> >>>>   double odds1 = 2.75;e >>>>   double odds2 = 2.30;P >>>>   double result;  >>>>   unsigned int temp;n >>>> >>>>   result = odds1 * odds2;0 >>>>   temp = (unsigned int) (result * 10000.0);# >>>>   printf("temp = %d\n", temp);  >>>> >>>> } >>>>I >>>  It appears that there is some truncation taking place, as well as   k >>> convertingJ >>> from decimal to binary float.  I tried the same with PL/I just confirm >>>0 >> >> <big PL/I snip> >> >>>   6.32500000000000E+04
 >>> FREJA>D >>>  I can only conclude that you are using the wrong language, try  >>> fortran  or 	 >>> PL/I.  >>>c >>D >> (This kind of stuff from C always brings out the language bigots) >  > J > Yes.  The real problem is that you cannot map the finite set of decimalsA > supported on the computer to the finite set of flaoting binary.y > C > PL/I supports both fixed and floating decimal, permitting exact  c > representationH > of such numbers.  You certainly would not use C for making financial   > calculations/ > (unless you use a decimal arithmetic package)i  I As does BASIC.  In the 30 years I've used the DOUBLE data type I've only -H seen a few instances where it has such problems.  In theory the problem I does exist, no question about that.  However, the extra precision it has fG vs IEEE floating point seems to be able to handle most practical usage.4  F Specifically, used for US currency, with two digits after the decimal I point, and some rather reasonable numbers normally used (ie; not lyre or  A yen) it is adequate.  Usage of a rounding function completes the e
 functionally.e  + For absolute purposes, you're 100% correct.a   >>J >> Well, maybe the wrong language.  I'd suggest BASIC, as below, first on 
 >> a  VAX: >> >> 10      Declare Long Th >> >>          a1 = 2.75A >>          a2 = 2.30r >> >>          a = a1 * a2a* >>          print using "######.######", a >> >>          r = a * 10000.0o* >>          print using "######.######", r >> >>          t = a * 10000.0  >>          print tt >> >> Ready >> >> run >> NONAME     2-MAR-2005 23:39 >> >>       6.325000n >>   63250.000000C
 >>   63250 >> Ready >>& >> However, not so fast.  On an Alpha: >>
 >> $ run t >>       6.325000  >>   63250.000000 
 >>   63249 >> $ >>E >> Now that I got in my required language promotion, I'll guess that eG >> your  problem is using DOUBLE (8 byte floating point) on an Alpha.   I >> There is  no such thing.  Alpha converts such to IEEE floating point,  I >> does the  calculations, and then converts it back.  Guess what, there  9 >> are 3 less  bits of prescision in IEEE floating point.  >> >> You might try:  >>8 >> temp = (unsigned int) ( (result * 10000.0) + .00001); >>/ >> That fixes my example in BASIC on the Alpha.r >> >> Or, get a VAX.c >> >> Dave0 >  >  >  >    ------------------------------  $ Date: Thu, 3 Mar 2005 17:40:17 -0000* From: "Richard Brodie" <R.Brodie@rl.ac.uk>) Subject: Re: DECC floating point questiont2 Message-ID: <d07i61$c92$1@blackmamba.itd.rl.ac.uk>  a "Joseph Huber" <huber@NOBODY-mppmu.mpg.de> wrote in message news:MWZFpOtRfzHX@vms.mppmu.mpg.de...o  E > Your problem is not really FP precision (at least not for 4 decimal O > digits), but the C language casting, which is simply truncating the mantissa.SM > Other languages have ENTIER or NINT (Nearest INTeger) functions to give YouU > the expected result.  
 So does C.  Q To summarise: when working with floating point numbers one should be careful thattV one doesn't require ridiculously large precision by doing something like rounding downQ 0.9999999999999999 to 0. Realise that using something harmless looking like (int)nU can do this. This applies even when the calculations look straightforward in base 10.    ------------------------------  % Date: Thu, 03 Mar 2005 13:32:04 -0500b' From: Dave Froble <davef@tsoft-inc.com>y) Subject: Re: DECC floating point question 0 Message-ID: <112elgcf064sf5b@corp.supernews.com>   Tom Linden wrote:lI > On Thu, 03 Mar 2005 11:44:18 -0500, Dave Froble <davef@tsoft-inc.com>  e > wrote: > J >> Specifically, used for US currency, with two digits after the decimal  I >> point, and some rather reasonable numbers normally used (ie; not lyre 0I >> or  yen) it is adequate.  Usage of a rounding function completes the  c >> functionally. >  > I > We had this same discussion in 2004, for currency you need 5 accurate  .	 > decimalnK > places, and float binary is not suitable for that task.  I think the OP  e > sample > demonstartes that. >  Yes.   ------------------------------  # Date: Thu, 03 Mar 2005 14:30:49 GMT 3 From: hammond@not@peek.ssr.hp.com (Charlie Hammond)o, Subject: Re: Fiorina and The Peter Principle0 Message-ID: <tcFVd.891$I75.342@news.cpqcorp.net>  C In article <1109805482.b48929dfad008cf48321d11904dd52b3@teranews>, c/ JF Mezei <jfmezei.spamnot@teksavvy.com> writes:   G >Carly has been a staunch republican supporter and had been rumoured ton0 >get a government appointment for some time. ...  ; For those who may have a full stomach, I appologize, but...y  5 Carly vs. Hillary in '08 is not all that unthinkable.T   -- EJ       Charlie Hammond -- Hewlett-Packard Company -- Ft Lauderdale  FL  USAF           (hammond@not@peek.ssr.hp.com -- remove "@not" when replying)J       All opinions expressed are my own and not necessarily my employer's.   ------------------------------  $ Date: Thu, 3 Mar 2005 09:31:14 -0500# From: "John Smith" <a@nonymous.com>u, Subject: Re: Fiorina and The Peter Principle, Message-ID: <g_6dnaxNBq0pgrrfRVn-oQ@igs.net>   Charlie Hammond wrote:D > In article <1109805482.b48929dfad008cf48321d11904dd52b3@teranews>,1 > JF Mezei <jfmezei.spamnot@teksavvy.com> writes:e >wF >> Carly has been a staunch republican supporter and had been rumoured5 >> to get a government appointment for some time. ...o > = > For those who may have a full stomach, I appologize, but...t >o7 > Carly vs. Hillary in '08 is not all that unthinkable.     " I can see her campaign slogan now:  7 "I screwed Lucent. I screwed HP. Now let me screw you."e     --J OpenVMS - The classics never go out of style.....but we don't want them to become popular again.    ------------------------------  % Date: Thu, 03 Mar 2005 11:01:42 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com>i, Subject: Re: Fiorina and The Peter PrincipleB Message-ID: <1109865026.960d8f79aa727db9291468e9e334828f@teranews>   Charlie Hammond wrote:  u7 > Carly vs. Hillary in '08 is not all that unthinkable.D  D I was discussing this very thing with a friend recently, except that% both were running as Vice Presidents.s  H Carly has very good presentation skills, but she also excells in talkingE without saying anything, and in a VP/presidential race, I am not suret1 this is such an asset. (but once elected, it is).   > It would make for interesting debates with very high ratings.   F Since she wasn't included in the Bush staff in January, I figured that= perhaps what she'd do is run for senate/congress seat in 2006,@ representing Kalifornia. Then run for presidential race with theB goal/asset of being able to delivery Kalifornia to the Rep. party.  B If she accepts a World Bank appointment, that would make it nearlyG impossible for her to be considered as VP for 2008 and if the democrats C are elected, her chances of getting any good jobs would dwindle and ! she's be stuck at the world bank.n   ------------------------------  # Date: Thu, 03 Mar 2005 17:31:22 GMT-3 From: hammond@not@peek.ssr.hp.com (Charlie Hammond)o, Subject: Re: Fiorina and The Peter Principle0 Message-ID: <KRHVd.914$165.579@news.cpqcorp.net>  C In article <1109865026.960d8f79aa727db9291468e9e334828f@teranews>, -/ JF Mezei <jfmezei.spamnot@teksavvy.com> writes:: >Charlie Hammond wrote:g > 8 >> Carly vs. Hillary in '08 is not all that unthinkable. >oE >I was discussing this very thing with a friend recently, except that0& >both were running as Vice Presidents.  G It has been clear for some time that unless she screws up really badly,oG Hillary will be the 2008 Democrat nominee for president.  i.e., NOT VP.nG The Democrats were very deliberate -- if somewhat subtle! -- in LOOSING)D the 2004 election so that there would NOT be a Democrat incumbent toF block Hillary in 2008.  Unless the Repblicans pull of a small -- maybeF not so small? -- miracle, Hillary will be our next president.  I'm notF saying I like the idea, just that this is the way the wind is blowing.  D As for Carly running for VP -- hard to imagine.  I'm not so sure sheF has anything more that a remote chance at the presidential nomination,D but if she gets it, she'll probably try to find a way to run without
 a VP!  <grin>I   -- aJ       Charlie Hammond -- Hewlett-Packard Company -- Ft Lauderdale  FL  USAF           (hammond@not@peek.ssr.hp.com -- remove "@not" when replying)J       All opinions expressed are my own and not necessarily my employer's.   ------------------------------  % Date: Thu, 03 Mar 2005 12:34:58 -0500e- From: JF Mezei <jfmezei.spamnot@teksavvy.com>:, Subject: Re: Fiorina and The Peter PrincipleB Message-ID: <1109870556.ca6938549ea71070ea512ddc9149a4c4@teranews>   Charlie Hammond wrote:  F > As for Carly running for VP -- hard to imagine.  I'm not so sure sheH > has anything more that a remote chance at the presidential nomination,F > but if she gets it, she'll probably try to find a way to run without > a VP!  <grin>c  ( ROFL ! :-) :-) :-) Very good one !!!!!!!  H makes you wonder how a woman of her calibre would have refused to have aG COO when she should have known that the other option was getting fired.   F Perhaps her contract was such that had she voluntarily left to anotherG job, her compensation package would have been tiny, so it may have beenoD to her own personal advantage to get fired. 20+ million is not smallK change. (or near 40  when you consider the other amenities she is getting).i   ------------------------------  % Date: Thu, 03 Mar 2005 13:32:33 -0500r( From: Bill Todd <billtodd@metrocast.net>, Subject: Re: Fiorina and The Peter Principle= Message-ID: <9tGdnRB5S4LcxbrfRVn-1A@metrocastcablevision.com>e   Charlie Hammond wrote:   ...E  I > It has been clear for some time that unless she screws up really badly,a: > Hillary will be the 2008 Democrat nominee for president.  I I'd advise you not to count on that, but I admit (regretfully) that it's p a distinct possibility.h      i.e., NOT VP.I > The Democrats were very deliberate -- if somewhat subtle! -- in LOOSINGaF > the 2004 election so that there would NOT be a Democrat incumbent to > block Hillary in 2008.  H Hard to say.  My usual inclination is not to blame on conspiracies what F can reasonably be blamed on simple (even if monumental) incompetence. H And once the Democratic leadership managed to eliminate the interesting E (and potentially more viable) candidates, what they had left was not , very inspiring.m  A I would, however, agree that the leadership seemed to prefer the aG *possibility* of losing the presidency to that of losing their control oF over the party.  Whether they expected additional losses in the House  and Senate is less clear.h  1    Unless the Repblicans pull of a small -- maybeo? > not so small? -- miracle, Hillary will be our next president.   D I don't think so.  Some of us refused to vote for Kerry even with a F second Bush term as the possible outcome.  A lot more would refuse to E vote for Hillary with even a somewhat less objectionable alternative.   B If it takes yet another loss for Democratic party to be forced to 6 provide a nominee we're willing to vote for, so be it.   - bill   ------------------------------  # Date: Thu, 03 Mar 2005 18:48:02 GMT 3 From: hammond@not@peek.ssr.hp.com (Charlie Hammond) , Subject: Re: Fiorina and The Peter Principle0 Message-ID: <CZIVd.928$Ii5.363@news.cpqcorp.net>  C In article <1109870556.ca6938549ea71070ea512ddc9149a4c4@teranews>, -/ JF Mezei <jfmezei.spamnot@teksavvy.com> writes:0 >Charlie Hammond wrote:e >pG >> As for Carly running for VP -- hard to imagine.  I'm not so sure sheeI >> has anything more that a remote chance at the presidential nomination,dG >> but if she gets it, she'll probably try to find a way to run without  >> a VP!  <grin> >u) >ROFL ! :-) :-) :-) Very good one !!!!!!!   A Glad you liked it.  (Heck, I'm just glad somebody understood it!)   I >makes you wonder how a woman of her calibre would have refused to have atH >COO when she should have known that the other option was getting fired. >tG >Perhaps her contract was such that had she voluntarily left to anothera6 >job, her compensation package would have been tiny...   Possible.  nI It also isn't difficult to imagine that her ego is bigger than her brain.h  D But frankly, I'm much more interested in what the future will bring.   -- cJ       Charlie Hammond -- Hewlett-Packard Company -- Ft Lauderdale  FL  USAF           (hammond@not@peek.ssr.hp.com -- remove "@not" when replying)J       All opinions expressed are my own and not necessarily my employer's.   ------------------------------  % Date: Thu, 03 Mar 2005 13:34:52 -0500:( From: Bill Todd <billtodd@metrocast.net>, Subject: Re: Fiorina and The Peter Principle= Message-ID: <9tGdnRN5S4JRxbrfRVn-1A@metrocastcablevision.com>a   JF Mezei wrote:    ...i  H > Since she wasn't included in the Bush staff in January, I figured that? > perhaps what she'd do is run for senate/congress seat in 2006tB > representing Kalifornia. Then run for presidential race with theD > goal/asset of being able to delivery Kalifornia to the Rep. party.  B I think you may be seriously overestimating Carly's popularity in I California, though perhaps many of the celebrity-conscious people in the i# southern half would vote for her...w   - bill   ------------------------------  # Date: Thu, 03 Mar 2005 18:36:54 GMTy# From: Beach Runner <bob@nospam.com>h, Subject: Re: Fiorina and The Peter Principle; Message-ID: <aPIVd.128146$qB6.5728@tornado.tampabay.rr.com>s   Charlie Hammond wrote:  E > In article <1109870556.ca6938549ea71070ea512ddc9149a4c4@teranews>, o1 > JF Mezei <jfmezei.spamnot@teksavvy.com> writes:n >  >>Charlie Hammond wrote: >> >>G >>>As for Carly running for VP -- hard to imagine.  I'm not so sure sheiI >>>has anything more that a remote chance at the presidential nomination,sG >>>but if she gets it, she'll probably try to find a way to run without) >>>a VP!  <grin> >>* >>ROFL ! :-) :-) :-) Very good one !!!!!!! >  > C > Glad you liked it.  (Heck, I'm just glad somebody understood it!)r >  > J >>makes you wonder how a woman of her calibre would have refused to have aI >>COO when she should have known that the other option was getting fired.q >>H >>Perhaps her contract was such that had she voluntarily left to another7 >>job, her compensation package would have been tiny... >  > 
 > Possible.  K > It also isn't difficult to imagine that her ego is bigger than her brain.P > F > But frankly, I'm much more interested in what the future will bring. > H  From what I hear, she know a LOT about Ken Olsen and the VMS engineers # were originally impressed with her.H   ------------------------------   Date: 3 Mar 2005 11:50:33 -0600P- From: Kilgallen@SpamCop.net (Larry Kilgallen)h Subject: Re: GBLSECTIONS loop 3 Message-ID: <NL5nRqwmrh72@eisner.encompasserve.org>w   In article <rdeininger-0203052028460001@user-uinj5jj.dialup.mindspring.com>, rdeininger@mindspringdot.com (Robert Deininger) writes:K > In article <rXrqF4r4puRp@eisner.encompasserve.org>, Kilgallen@SpamCop.nete  I >>The offer by AUTOGEN sometimes works (it has for me - never say never).e >>H >>I think the flaw is in the underlying design of the ADD_xxx mechanism.$ >>Not that I have a better design... > K > I saw no mention of the VMS version.  Substantial work was done in V7.3-2 L > to fix problems around AUTOGEN and some of the layered products and greedy > IO adapters.  D That is good to hear.  I have never run VMS 7.3-2 except for a brief period to upgrade to VMS 8.2.t   ------------------------------   Date: 3 Mar 2005 06:58:55 -0800i& From: tlywong@gmail.com (Terence Wong)6 Subject: Re: Identify the process locking a RMS record= Message-ID: <87af9e83.0503030658.22d7aa47@posting.google.com>f  + Thanks to everyone responded to my query.   J I'll pass these info to our internal technical support staff to follow up.   Regards, Terence.   "Main, Kerry" <kerry.main@hp.com> wrote in message news:<FD827B33AB0D9C4E92EACEEFEE2BA2FB53F928@tayexc19.americas.cpqcorp.net>...m > Hello Terry, > I > As a suggestion, you may want to consider the Availability Manager that-; > is available for free download at the following location:" > @ > http://h71000.www7.hp.com/openvms/products/availman/index.html > - > Client piece can run on OpenVMS or Windows.  > 	 > Regards0 >  > Kerry Main > Solutions Architectt > Hewlett-Packard (Canada) Co.# > Consulting & Integration Servicesi > Voice: 613-592-4660  > Fax   : 613-591-4477 > Email: Kerry.Main@hp.com > =20d >  > > -----Original Message-----4 > > From: Terence Wong [mailto:tlywong@gmail.com]=20  > > Sent: March 1, 2005 10:44 AM > > To: Info-VAX@Mvb.Saic.Como6 > > Subject: Identify the process locking a RMS record > >=20D > > We had problem to identify application process locking a record. > >=20H > > Before any program change, is there any VMS utility show the process > > locking the record ? > >=20G > > Also how to identify the process locking the record through ANALYZEm
 > > /SYSTEM ?t > >=20$ > > Thanks for any input in advance, > >=20 > > Terence. > >=20   ------------------------------  % Date: Thu, 03 Mar 2005 13:51:13 +0200n4 From: Mike Rechtman <michael.rechtman.nospam@hp.com> Subject: Re: Logfile name & Message-ID: <42271650.419FB35A@hp.com>   David Gray wrote:n > 	 > Hi all,  > F > From within a batch job how do I find out what the value of the /LOG > qualifier was at submit time?  >  > e.g. > @ > $ submit/log=gen_logs:a_log_file_with_a_long_name.log    X.COM > D > When X.COM is running I want to put the value specifed on the /LOG > into a symbol. >  > $ show symbol LogFile ; >      LogFile = "gen_logs:a_log_file_with_a_long_name.log"  > F > I was fairly sure I could do this with a lexical function but cannot > find/remember which one. >  > Cheers >         Dave.1     Do you mean:  C $ LogFile = f$getqui("DISPLAY_JOB","LOG_SPECIFICATION",,"THIS_JOB")o   HTH, Mike   -- sE --------------------------------------------------------------------- E Usual disclaimer: All opinions are mine alone, perhaps not even that.l? Mike Rechtman                            *rechtman@tzora.co.il* F Kibbutz Tzor'a.                          Voice (home): 972-2-9908337  B   "20% of a job takes 80% of the time, the rest takes another 80%"E ---------------------------------------------------------------------  -----BEGIN GEEK CODE BLOCK-----  Version: 3.1: GCM/CS d(-)pu s:+>:- a++ C++ U-- L-- W++ N++ K? w--- V+++$6 PS+ PE-- t 5? X- tv-- b+ DI+ D-- G e++ h--- r+++ y+++@ ------END GEEK CODE BLOCK------h   ------------------------------  % Date: Thu, 03 Mar 2005 08:09:04 -0500i- From: JF Mezei <jfmezei.spamnot@teksavvy.com>-/ Subject: Re: new Itanium after Tukwila: PoulsonsB Message-ID: <1109854600.2bce5a10019a7a7ae22a1d8e3dae1252@teranews>   Dave Weatherall wrote:C > presume it was a controbutor to Fred's belief that there would be  > IA-64 workstations for VMS.   E One has to remember that back then, it was tought that the 8086 would H run out of steam and a new architecture needed (same arguments as VAX to@ Alpha). So there was the gola of having IA64 replace the 8086 asE industry standard chip, and this is why they initially incorporated a B pentium 3 emulation/core inside the IA64 thing, but it was quicklyD realised that the 8086 emulation was way too slow compared to modernG 8086 chips (especially since it took so long for Intel to get the firstu commercial chip out (merced).=  C At the time Curly murdered Alpha, it had been known for a number of~C years that IA64 wouldn't replace the 8086 as industry standard. ButEF there was still a chance that it might become a very common enterpriseC chip. Yep compaq still claimed it would become a commodity industryn standard chip.  B If it was such a wonderful chip, perhaps Compaq and HP should haveF announced the end of 8086 based desktops and servers. They didn't. Yet) they force us to move to that IA64 thing.m    F Funny how "oh, it is simple port, just recompile" argument is given toD us, but "you can't have desktops on IA64 due to lack of software" is- used to justify NOT moving from 8086 to IA64.u   ------------------------------  $ Date: Thu, 3 Mar 2005 19:42:30 +0100, From: "Dr. Dweeb" <5msg0h202@sneakemail.com>/ Subject: Re: new Itanium after Tukwila: Poulsons- Message-ID: <d07lqt$2fu7$1@news.cybercity.dk>h   JF Mezei wrote:t > Dave Weatherall wrote:D >> presume it was a controbutor to Fred's belief that there would be >> IA-64 workstations for VMS. > G > One has to remember that back then, it was tought that the 8086 wouldaG > run out of steam and a new architecture needed (same arguments as VAXsE > to Alpha). So there was the gola of having IA64 replace the 8086 as-G > industry standard chip, and this is why they initially incorporated arD > pentium 3 emulation/core inside the IA64 thing, but it was quicklyF > realised that the 8086 emulation was way too slow compared to modernC > 8086 chips (especially since it took so long for Intel to get then% > first commercial chip out (merced).r >   L Which reminds me - is that emulation core (or whatever its correct name was)H still taking space on the die, or has it actually been removed?  I neverK seem to hear *anything* about IA64 - 8086 emulation any more.  It was a hotaI topic for a while with the inbuilt being a dog, then there was some FX!32 G style emulation - I guess that was going to be part of the IA64 Windoze K kits.  But it seems years since I heard anything about it - not that I have " been looking too hard I might add.  	 Dr. Dweebr  E > At the time Curly murdered Alpha, it had been known for a number of-E > years that IA64 wouldn't replace the 8086 as industry standard. ButrH > there was still a chance that it might become a very common enterpriseE > chip. Yep compaq still claimed it would become a commodity industrya > standard chip. >rD > If it was such a wonderful chip, perhaps Compaq and HP should haveH > announced the end of 8086 based desktops and servers. They didn't. Yet+ > they force us to move to that IA64 thing.  >s >fH > Funny how "oh, it is simple port, just recompile" argument is given toF > us, but "you can't have desktops on IA64 due to lack of software" is/ > used to justify NOT moving from 8086 to IA64.e   ------------------------------  % Date: Fri, 04 Mar 2005 01:20:38 +0800  From: prep@prep.synonet.com./ Subject: Re: new Itanium after Tukwila: Poulsonn- Message-ID: <87r7iwr6hl.fsf@prep.synonet.com>s  ) Dave Froble <davef@tsoft-inc.com> writes:k   > icerq4a@spray.se wrote:d  E >> What I try to say is that the main reasons for large L3 caches area8 >> to alleviate for a less powerful system architecture.   > I must confess to curosity.e  C > In the above you're admitting that you think the itanic isn't the  > best of designs.  E That is putting it very mildly. Ignoring the total lack of deliveringD> near on time or performance, the itanic digs swamp appon hole.  E It *requires* a fundamental breakthrough in compiler theory, and thengD real compilers to implement it. This has not happened, and no one is
 supprised.  F Speculative execution adds even more pressure to the memory system andD blows out latency. Bandwidth can be fixed with enough $$, latency is forever!  @ It requires the compiler to schedule the code stream to get goodF performance, but this will change in very many cases with the DATA you
 process...  C The amount of data in a context switch is humungous, and thus slow.t  F > Another post claimed that when HP's technical people saw the successB > of DEC and IBM with out-of-order designs they advised management> > that EPIC should be dropped, and that sticking with it was a' > management, not a technical decision.   1 That is what the scuttle-butt just post EV6 said.e  1 > How many applications run pretty much in cache?o   Quite a few, but not RDBs..   < > How many users tweek and optimize the compilation of their > applications?,  	 2, maybe.s  1 > How many users buy computers to run benchmarks?o  A > Itanic is a rather poor design that due to state of the art CPUdE > manufacturing will allow most people to run their applications withr6 > acceptable performance, in spite of the poor design.  E Intels designers have done a huge job keeping the power in check, bute pearls before swine.  D > So my question is, just what is it about this CPU design that gets# > people to advocate and defend it?c  E Billions of dollars. And a non-stop stream of crap about how fast andd wonderfull it will be.  F The itanic seems to have also invented a new, original hole. It is notD only latency, bandwidth and thermally limited, it is design limited.D Other than moderate re-shuffles of the core, and adding huge amountsD of cache the RTL design has gone almost nowhere since it crawled out of the dark.  E The sooner it is killed of for good, the better off everyone will be.e( Including, or possibly, espicialy intel.  @ Its only plus is commercial; it has killed off at least 3 rivalsC without even shipping a single chip. Not even Big BLue at it finestl managed that sort of coup.     -- m< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.2@                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  # Date: Thu, 03 Mar 2005 08:02:52 GMTeL 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: <00A4030C.CE8EC387@SSRL.SLAC.STANFORD.EDU>  R In article <Aa2dnfVyab13G7vfRVn-tw@igs.net>, "John Smith" <a@nonymous.com> writes:- >Alan Winston - SSRL Central Computing wrote: < >> In article <7dqdnQEBgc-iWbjfRVn-iQ@igs.net>, "John Smith" >> <a@nonymous.com> writes:  >>> Neil Rieck wrote:>+ >>>> <bob@instantwhip.com> wrote in messagemA >>>> news:1109550872.067891.88350@o13g2000cwo.googlegroups.com...uH >>>>> Oracle would create a band of very unhappy customers if they tried >>>>> to eol rdb >>>>> F >>>> The following interesting Oracle RDB points were mentioned at the
 >>>> seminar:h >>>>H >>>> 1) 50% of the code currently in RDB was put there since the sale of, >>>> the RDB division from Digital to Oracle >>>  >>>aH >>> No surprise there.... Code that big always has bugs that need fixingH >>> 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.2 >>E >> That seems unnecessarily dismissive.  Rdb development has suffered0 >> somewhatmH >> from diversions like Tru-64 development (nobody wanted to buy it) andD >> Windows development, but there a lot of new features, performanceF >> enhancements, new syntax, etc, etc, since the Oracle near-giveaway.F >> Multifile databases, row cache, two flavors of JDBC server, SQL*Net5 >> support, etc, etc - lot of new features since 6.2.d >>D >> (And it's worth remembering that Oracle got just about all of theE >> Digital Rdb development team and built them a workplace near theircA >> old one, rather than requiring all development move across thee >> country.) >>G >> I wouldn't argue with a charge  that Oracle's posture toward Rdb hasi >> beentE >> benign neglect - leaving the Rdb team to get on with it so long as A >> they keep making their revenue goals - when it could have beenm >> promoted moreF >> aggressively.  But I reject what I read as your imputation that Rdb: >> has essentially been in maintenance mode for ten years. >t >pL >I know that there has been a lot of work on it, and VMS too. My glibness inJ >my post was directed more towards statements of performance that were not1 >presented with factual evidence to back them up.0  N Okey-doke.  You'll note that I cut out your remarks about the speed claims - I- had no disagreement with them, and don't now.T   -- Alan    ------------------------------  $ Date: Thu, 3 Mar 2005 07:26:05 -0500) From: "Neil Rieck" <n.rieck@sympatico.ca>oA Subject: Re: OpenVMS Seminar in Toronto (2005-02-24) a few points ; Message-ID: <loDVd.43207$Vf6.1205620@news20.bellglobal.com>r  / "John Smith" <a@nonymous.com> wrote in message  & news:Aa2dnfVyab13G7vfRVn-tw@igs.net... >n [...snip...] >tK > I know that there has been a lot of work on it, and VMS too. My glibness   > inK > my post was directed more towards statements of performance that were note2 > presented with factual evidence to back them up. > 9 > Fastest commecial database can mean a lot of things....#, > - fastest Rdb database on a single machineK > - fastest rdbms on a cluster of 1280's maxed out with memory, cpu's, and 0 > the  > fastest storage going G > - fastest rdbms on a single processor tied to a Patriot 2 missile 15 3	 > seconds  > after launch > @ > No presentation of experimental method, no claim to greatness. >   K I, too, would like to know the hardware and software parameters associated eM with such a claim. I didn't see any slide backing up such a claim, but Terry  M Shannon mentioned that he was looking into IBM's TPCC benchmark (re: special  A DB2 on R6000) and would be publishing something about it shortly.t  L p.s. I was making notes as fast as possible (for me) and hope that accuracy M wasn't sacrificed doing so. That said, my transcription of the comment about )M a special version of DB2 (which no customer may purchase) on R6000 is pretty  F much 99% accurate. Now we need to discover if the speaker was factual.  
 Neil Rieck Kitchener/Waterloo/Cambridge,- Ontario, Canada.8 http://www3.sympatico.ca/n.rieck/links/cool_openvms.html   ------------------------------   Date: 2 Mar 2005 23:41:22 -0800s' From: aniranshu@yahoo.co.in (aniranshu)n Subject: Oracle RDB0= Message-ID: <36ad2016.0503022341.3588d24c@posting.google.com>    Hi,   J I am new to this product. Can anyone provide me with links to start with?   
 Thanks a lot,p Anirban    ------------------------------  % Date: Thu, 03 Mar 2005 10:19:49 +0100u9 From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com>  Subject: Re: Oracle RDBn' Message-ID: <4226D6B5.7F9CE3A9@aaa.com>d   http://www.oracle.com/rdbr   aniranshu wrote: >  > Hi,: > K > I am new to this product. Can anyone provide me with links to start with?1 >  > Thanks a lot, 	 > AnirbanY   ------------------------------  % Date: Thu, 03 Mar 2005 18:29:19 +0100e0 From: Keith Cayemberg <keith.cayemberg@arcor.de> Subject: Re: Oracle RDB.B Message-ID: <4227496f$0$24945$9b4e6d93@newsread2.arcor-online.net>   Jan-Erik Sderholm wrote:  > http://www.oracle.com/rdb  >  > aniranshu wrote: >  >>Hi,t >>K >>I am new to this product. Can anyone provide me with links to start with?h >> >>Thanks a lot,i	 >>Anirbany   Here's a few more Links...  - Oracle Rdb - Wikipedia, the free encyclopedia,' http://en.wikipedia.org/wiki/Oracle_Rdbi  ( Oracle Rdb (quick overview) - Neil Rieck< http://www3.sympatico.ca/n.rieck/docs/rms_rdb_notes.html#rdb   Documentation URLs7 http://www.oracle.com/technology/documentation/rdb.html @ http://www.oracle.com/technology/products/rdb/rdb_doc_index.html  6 Software Downloads for Oracle Rdb and Related ProductsG (please see licensing conditions for commercial and non-commercial use) B http://www.oracle.com/technology/software/products/rdb7/index.html  ) Oracle-L: Re: Oracle RDB vs. Plain Oraclet; http://www.orafaq.com/maillist/oracle-l/2002/05/16/1530.htm    Cheers!e   Keith Cayembergr   ------------------------------   Date: 3 Mar 2005 02:54:30 -0800  From: macro81@sms.at( Subject: Running OpenVMS on DEC PWS 500AC Message-ID: <1109847270.003260.199970@f14g2000cwb.googlegroups.com>f   I found a used Alpha PWS 500a.F Think i'll use it for C coding, etc. A nice little alpha for home use.: I would be interested running OpenVMS on this workstation., Or FreeBSD if there is no change to run VMS.   Following specs:- 500 MHz Alpha 21164, 2x SCSI HDD, 256 MB RAM,>1 CD-ROM, 3,5" Floppy, On-Board Sound, SCSI and LAN> System with AlphaBios 5.4h  < I found out that there are two series (500 A and AU) of thisC DEC alpha PWS Miata. "A" for Win NT preinstalled; "AU" for OpenVMS,e Tru64 Unix.   , Is it possible to run OpenVMS on a PWS 500a?@ Or on a modified 500a? Changing some peripheral components, etc?  D I've to check what SCSI controller and what graphic chipset is used." HDD and CDROM are SCSI components.  3 As far as I know some components should be replacedi' in order to successfully bootstrap VMS. 1 e.g. a QLogic SCSI controller, other graphic card+1 Elsa Gloria with Permedia chip set or powerstorm?r  $ Think a frimware change is required.! SRM Console instead of AlphaBIOS?>  ) Updates or new firmware versions from the  old digital ftp server?    TIA!!.   Greets, Marcus   ------------------------------  % Date: Thu, 03 Mar 2005 13:22:57 +0100uE From: Bitnissen <bitnissen#SPAMMER_FORMAT_YOUR_HARDDISK#@hotmail.com>x, Subject: Re: Running OpenVMS on DEC PWS 500A8 Message-ID: <ovvd21l5jh010ouvk68m2j1vuarna7t15u@4ax.com>  	 Hi Marcus   B I am running OpenVMS 7.3-1 on my PWS 500A with the latest firmware
 (SRM 7.2).  ; You can get the latest firmware from HP alpha support pages   / To install OpenVMS you will need a SCSI cd-rom.   ? If you want to run x-windows you will need a OpenVMS compatibleu" graphics card - se the openVMS FAQ   Best regards   Jimmii    3 On 3 Mar 2005 02:54:30 -0800, macro81@sms.at wrote:4   >I found a used Alpha PWS 500a.nG >Think i'll use it for C coding, etc. A nice little alpha for home use. ; >I would be interested running OpenVMS on this workstation..- >Or FreeBSD if there is no change to run VMS.  >c >Following specs:b. >500 MHz Alpha 21164, 2x SCSI HDD, 256 MB RAM,2 >CD-ROM, 3,5" Floppy, On-Board Sound, SCSI and LAN >System with AlphaBios 5.4 >:= >I found out that there are two series (500 A and AU) of this>D >DEC alpha PWS Miata. "A" for Win NT preinstalled; "AU" for OpenVMS, >Tru64 Unix. >#- >Is it possible to run OpenVMS on a PWS 500a? A >Or on a modified 500a? Changing some peripheral components, etc?. > E >I've to check what SCSI controller and what graphic chipset is used.5# >HDD and CDROM are SCSI components.6 >04 >As far as I know some components should be replaced( >in order to successfully bootstrap VMS.2 >e.g. a QLogic SCSI controller, other graphic card2 >Elsa Gloria with Permedia chip set or powerstorm? >o% >Think a frimware change is required.i" >SRM Console instead of AlphaBIOS? >o* >Updates or new firmware versions from the >old digital ftp server? >f >TIA!! >n >Greets, Marcus    ------------------------------   Date: 3 Mar 2005 01:54:11 -0800e2 From: filip.de.block@proximus.net (Filip De Block)8 Subject: setting default protection for CDD+ directories= Message-ID: <8366afbe.0503030154.7ad92947@posting.google.com>A  ; I have a whole of (sub) directories in a massive CDD+ tree. B Creating a new directory is no problem, setting protection on that@ directory is no problem, but I fail to convince CDD+ to take theD directory protection when creating new element/entries what so ever.  @ DEFINE PROTECTION for directory cdd$top.pni.v065.task identifier  PNI_RW access read+write. --> ok etc...  : DEFINE FIELD cdd$top.pni.v065.task.test datatype longword.  5 SHOW PROTECTION FOR FIELD cdd$top.pni.v065.task.test t
 TEST_FIELD(1)iK           (IDENTIFIER=[DEBLOCK],ACCESS=READ+WRITE+MODIFY+ERASE+SHOW+DEFINE+t7           CHANGE+DELETE+CONTROL+OPERATOR+ADMINISTRATOR) J           (IDENTIFIER=[SYSTEM],ACCESS=READ+WRITE+MODIFY+ERASE+SHOW+DEFINE+7           CHANGE+DELETE+CONTROL+OPERATOR+ADMINISTRATOR)lN           (IDENTIFIER=[*,*],ACCESS=READ+WRITE+MODIFY+ERASE+SHOW+DEFINE+CHANGE+0           DELETE+CONTROL+OPERATOR+ADMINISTRATOR)  E Anyone can tell me where these (silly) ACE's come from? How I can gettF my own ACE's, just like using the OPTION=DEFAULT with vms directories.  
 advTHANKSance    ------------------------------   Date: 3 Mar 2005 05:42:25 -0800l$ From: juno10000@hotmail.com (Triger)$ Subject: Re: sysgen and decw problem= Message-ID: <d969c33c.0503030542.2a932fa7@posting.google.com>F  o peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER) wrote in message news:<newscache$jcmqci$q242$1@news.sil.at>...mf > In article <1109772682.233003.97210@l41g2000cwc.googlegroups.com>, martinkirby12@yahoo.co.uk writes:@ > >It is a known issue with DECwindows, fixed in later releases. > I > Yup, it seems fixed now. But it took some (not to say many) versions...c > J > >Problem is that DECwindows calculates an increment for GBLSECTIONS thatI > >is incorrect. You can add your own ADD_GBLSECTIONS to MODPARAMS.DAT or-A > >you can disable the DECwindows Autogen check using a define ofa > >DECW$IGNORE_AUTOGEN.  > O > I prefer MIN_* (in my case MIN_GBLSECTIONS=1024 just for DECW) way over ADD_*8 > M > And don't forget to add this line to SYS$MANAGER:AGEN$NEW_NODE_DEFAULTS.DAT57 > and SYS$MANAGER:AGEN$NEW_SATELLITE_DEFAULTS.DAT also.   . I tried to add  min_gblsections but no change.H Martin, where I should put DECW$IGNORE_AUTOGEN ? In sysstartup_vms.com ?I Can you say a little more about modparams.dat and gblsections, please....   	 Thank youi   ------------------------------  * Date: Thu, 3 Mar 2005 14:44:47 +0000 (UTC)6 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER)$ Subject: Re: sysgen and decw problem0 Message-ID: <newscache$ma6sci$md3$1@news.sil.at>  d In article <d969c33c.0503030542.2a932fa7@posting.google.com>, juno10000@hotmail.com (Triger) writes:p >peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER) wrote in message news:<newscache$jcmqci$q242$1@news.sil.at>...g >> In article <1109772682.233003.97210@l41g2000cwc.googlegroups.com>, martinkirby12@yahoo.co.uk writes:8A >> >It is a known issue with DECwindows, fixed in later releases.m >> tJ >> Yup, it seems fixed now. But it took some (not to say many) versions... >>  K >> >Problem is that DECwindows calculates an increment for GBLSECTIONS thataJ >> >is incorrect. You can add your own ADD_GBLSECTIONS to MODPARAMS.DAT orB >> >you can disable the DECwindows Autogen check using a define of >> >DECW$IGNORE_AUTOGEN. >> wP >> I prefer MIN_* (in my case MIN_GBLSECTIONS=1024 just for DECW) way over ADD_* >> .N >> And don't forget to add this line to SYS$MANAGER:AGEN$NEW_NODE_DEFAULTS.DAT8 >> and SYS$MANAGER:AGEN$NEW_SATELLITE_DEFAULTS.DAT also. >i/ >I tried to add  min_gblsections but no change.o  I If you also have a GBLSECTIONS= hardcoded in MODPARAMS.DAT then this take9H precedence over ADD_GBLSECTIONS= and MIN_GBLSECTIONS= values. Check your? MODPARAMS.DAT and also SYS$SYSTEM:AGEN$PARAMS.REPORT for hints.i  I >Martin, where I should put DECW$IGNORE_AUTOGEN ? In sysstartup_vms.com ?s  6 The usual place would be in SYS$STARTUP:SYLOGICALS.COM  J >Can you say a little more about modparams.dat and gblsections, please....    http://www.hp.com/go/openvms/doc   -- n Peter "EPLAN" LANGSTOEGERw% Network and OpenVMS system specialistv E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------  % Date: Thu, 03 Mar 2005 10:40:10 +0100r! From: Soterro <soterroatyahoocom> : Subject: Re: This is the End... My Only Friend, The End...: Message-ID: <4226dac2$0$6574$4d4ef98e@read.news.ch.uu.net>   David J Dachtera wrote: , > So, what you're saying is, its inevitable.  G Decus in Germany (in April) is expected to have an attendance of about bG 600 people. That's half from what it was just a couple of years ago if - I'm not mistaking.   S    ------------------------------  $ Date: Thu, 3 Mar 2005 08:26:35 -0500# From: "John Smith" <a@nonymous.com>T: Subject: Re: This is the End... My Only Friend, The End..., Message-ID: <wvCdnbgtz7AQjbrfRVn-pg@igs.net>   Soterro wrote: > David J Dachtera wrote:<- >> So, what you're saying is, its inevitable.e >mH > Decus in Germany (in April) is expected to have an attendance of aboutH > 600 people. That's half from what it was just a couple of years ago if > I'm not mistaking.    K And immediately afterwards the HP apologists will be saying how wonderfully I attended the event was, which is like saying how interesting the Titianic.7 looked at that steep angle to the surface of the water.m   --- OpenVMS - The classics never go out of style.a   ------------------------------  % Date: Thu, 03 Mar 2005 12:36:12 +0000s- From: Roy Omond <Roy.Omond@BlueBubble.UK.Com>oE Subject: Re: What is the Difference between Shadow and Mirrored disk?g, Message-ID: <38oeqnF5o7hh6U1@individual.net>   Phillip Helbig wrote:f  8 > In article <38lkqfF5os0vvU1@individual.net>, Roy Omond( > <Roy.Omond@BlueBubble.UK.Com> writes:  >aH >>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* theh$ >>probability of losing the cluster. >>. >>Node A (1 vote) + local quorum disk (1 vote) >>Node B (1 vote)e >> >>Node B goes down ... >>C >>Now you have introduced an unnecessary probability of the cluster I >>being lost, equal to the probability of the quorum disk "disappearing".e >tH > In the situation above, if the quorum disk disappears, then there are I > still 2 out of 3 votes there, so the cluster will survive.  The reason oG > this is pointless, though, is since if node A goes down, so does the aK > disk, so the above is equivalent to giving node A two votes and node B 1 f > vote.   < No Phillip, you've not read my posting with sufficient care.  2 Note the line "Node B goes down ...", and re-read.  = And, no it's not equivalent.  The quorum disk *increases* then probability of cluster failure.g  J > With a two-node cluster without a "proper" quorum disk (which is really J > a three-node cluster with the disk taking the place of the third node), I > there is NO WAY to have a real cluster, whether or not a "local quorum hJ > disk" is used and however the votes are distributed.  Either you have a K > 50-50 situation: cluster freezes if either node goes down, or you have a rK > master-slave situation, where one node must ALWAYS be up for the cluster i9 > to be up and it doesn't matter what the other one does.i  F Unless you've got separate SYSUAFs, Rightslists etc. (not a good idea,B IMHO), then the master-slave is much better than the other method.   Hint: MVTIMEOUTo  H > I really fail to see why people these days even bother with less than H > optimal solutions.  Just get a cheap VAX and use it as a quorum node;  > problem solved.r  F Agreed.  Even better, get a cheap (preferably free) ES47 or GS1280 :-)   ------------------------------  $ Date: Thu, 3 Mar 2005 12:38:07 -0500 From: norm.raphael@metso.comE Subject: Re: What is the Difference between Shadow and Mirrored disk?dQ Message-ID: <OF03E538B6.FD387423-ON85256FB9.0060CD80-85256FB9.0061226A@metso.com>m  J helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply)  wrote on 03/02/2005 05:09:10 PM:  8 > In article <38lkqfF5os0vvU1@individual.net>, Roy Omond' > <Roy.Omond@BlueBubble.UK.Com> writes:f >sJ > > Sorry to harp on about this, but since it's coming from an HP address,H > > I'd just like to add the comment that using a *local* quorum disk asC > > suggested is not only unnecessary, but actually *increases* thea& > > probability of losing the cluster. > >i0 > > Node A (1 vote) + local quorum disk (1 vote) > > Node B (1 vote)w > >  > > Node B goes down ... > >bE > > Now you have introduced an unnecessary probability of the clustersK > > being lost, equal to the probability of the quorum disk "disappearing".D >cG > In the situation above, if the quorum disk disappears, then there are,H > still 2 out of 3 votes there, so the cluster will survive.  The reasonF > this is pointless, though, is since if node A goes down, so does theJ > disk, so the above is equivalent to giving node A two votes and node B 1 > vote.. >cI > With a two-node cluster without a "proper" quorum disk (which is reallysI > a three-node cluster with the disk taking the place of the third node),oH > there is NO WAY to have a real cluster, whether or not a "local quorumI > disk" is used and however the votes are distributed.  Either you have aiJ > 50-50 situation: cluster freezes if either node goes down, or you have aJ > master-slave situation, where one node must ALWAYS be up for the cluster9 > to be up and it doesn't matter what the other one does.d >!  G > I really fail to see why people these days even bother with less thanvG > optimal solutions.  Just get a cheap VAX and use it as a quorum node;i > problem solved.h >c  ; ..but no VAX can be in a supported cluster with an Itanium!b   ------------------------------   End of INFO-VAX 2005.124 ************************t to imagine that her ego is bigger than her brain.h  D But frankly, I'm much more interested in what the future will bring.   -- cJ       Charlie Hammond -- Hewlett-Packard Company - @ H !  H`
H
J !H"@  
H  #s PF    P , c h    $-C +i !    	hZ h6W @4!         @P-  H(P         h @
X:  $`P.`             C H  $	  H@
@         HgX   $ E%@	@   !                  fX X"T  DB`#X
H ^ !$,Y@` -X  !  H H	 ,  
hZ    8  B`
(  8  B 	HZ  !     'P H" !!  H
 X 
J !'3,
@X D !0 $1         ` `   ! `  C
H@p $q           P  C  $  B (fXh@ DB D   @     (  @  @        H , ,H   $  $
q  XP e!B   ( !!( B C  h"PT, 
 J !0EB &P X  $4-  C  
J !D4B
 X D !@ ,      41         `H `    DB             4 P         X @
h  $@P6`            0 C H2  $	  H@
@         dX0( 0P@	q     @8     C
X  $ E-@             X4 P          @
h  $@P6`             C  $  B XdX  8 DB C@P !,    X6 !,$B\        P C