0 INFO-VAX	Sat, 02 Feb 2002	Volume 2002 : Issue 64      Contents:& Re: "C" written os's spell "disaster"!& Re: "C" written os's spell "disaster"!& Re: "C" written os's spell "disaster"!& RE: "C" written os's spell "disaster"!& Re: "C" written os's spell "disaster"!& RE: "C" written os's spell "disaster"!& Re: "C" written os's spell "disaster"!D Re: Be prepared, Encompass is using SBA to SPAM folks for MembershipD Re: Be prepared, Encompass is using SBA to SPAM folks for MembershipD Re: Be prepared, Encompass is using SBA to SPAM folks for Membership/ Re: Capellas wants IBM model, but does reverse! & compaq logo on DECwindows login screen Re: curses on OpenVMS  Re: curses on OpenVMS # Re: Default terminal printer in CDE # Re: Default terminal printer in CDE  LICENSE/CHARGE Re: OpenVMS CD to 5.25 disksP Re: Proliant better than Alphas  was (Re: Alphaserver 4100: physical memory) mem Re: setting the record straight E RE: SGI solicits Alpha customers to trade in their "orphaned" systems + Re: Storageworks RAIDArray 310 INVLID CACHE  tcpip 	 Re: tcpip 6 The Inquirer says : "Compaq thrashes IBM, HP, Fujitsu": Re: The Inquirer says : "Compaq thrashes IBM, HP, Fujitsu"; Re: Using the 'latest and greatest' (was:VMS 7.3 stability) 2 Re: VMS booted on Itanic according to The Inquirer  F ----------------------------------------------------------------------  % Date: Sat, 02 Feb 2002 04:44:55 +0100 2 From: martin@radiogaga.harz.de (Martin Vorlaender)/ Subject: Re: "C" written os's spell "disaster"! ; Message-ID: <3c5b60b7.524144494f47414741@radiogaga.harz.de>   * Patrick Young (P.Young@unsw.EDU.AU) wrote:7 > martin@radiogaga.harz.de (Martin Vorlaender) wrote... . > > Patrick Young (P.Young@unsw.EDU.AU) wrote:> > > > "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> wrote...M > > > > The reason that VMS system services require a length, is to make that M > > > > probe possible without too high an overhead.  But it *is* possible to L > > > > do it with a null terminated string *if* you really wanted to do it.O > > > > You just need to probe for each byte before reading/testing/copying it.  > > > R > > > if (__PAL_PROBER(c_string_addr, c_string_len, ps_reg & PR$M_PS_PRVMOD) == 0) > > >     return(SS$_ACCVIO);  > > N > > Errr ...and you get the c_string_len from where (if not from a parameter)? > D > That's where the "just need to probe for each byte before reading/> > testing/copying it" comes in. The c_string_len becomes 1 andE > you test if the byte is null if you had access to it, if not null - 6 > use it as data and move on to probing the next byte.  $ Now I get it. Thanks for clarifying.  C If I wanted to read up on the __PAL_* stuff: where would I find it?   3 > Though in most cases you would do something like:  > F > int my_priv_routine(char *null_termed_string, int bufsize_of_string)  I Now that's the VMS way - though not the way C code or *ix normally works.    cu,    Martin --  G So long, and thanks        | Martin Vorlaender  |  VMS & WNT programmer 4 for all the books...       | work: mv@pdv-systeme.deK In Memoriam Douglas Adams  |       http://www.pdv-systeme.de/users/martinv/ ;             1952-2001      | home: martin@radiogaga.harz.de    ------------------------------  # Date: Sat, 02 Feb 2002 10:37:46 GMT  From: system@SendSpamHere.ORG / Subject: Re: "C" written os's spell "disaster"! 0 Message-ID: <00A08F34.34FE939A@SendSpamHere.ORG>  i In article <55f85d77.0202011559.7926603a@posting.google.com>, P.Young@unsw.EDU.AU (Patrick Young) writes: v >martin@radiogaga.harz.de (Martin Vorlaender) wrote in message news:<3c5ae124.524144494f47414741@radiogaga.harz.de>...- >> Patrick Young (P.Young@unsw.EDU.AU) wrote: = >> > "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> wrote... L >> > > The reason that VMS system services require a length, is to make thatL >> > > probe possible without too high an overhead.  But it *is* possible toK >> > > do it with a null terminated string *if* you really wanted to do it. N >> > > You just need to probe for each byte before reading/testing/copying it. >> > > > Q >> > if (__PAL_PROBER(c_string_addr, c_string_len, ps_reg & PR$M_PS_PRVMOD) == 0)  >> >     return(SS$_ACCVIO); >>  M >> Errr ...and you get the c_string_len from where (if not from a parameter)?  >>   > C >That's where the "just need to probe for each byte before reading/ = >testing/copying it" comes in. The c_string_len becomes 1 and D >you test if the byte is null if you had access to it, if not null -5 >use it as data and move on to probing the next byte.  > 2 >Though in most cases you would do something like: > E >int my_priv_routine(char *null_termed_string, int bufsize_of_string)  > ' >and a well behaved user could then do:  >  >char buf[80]; >.... ' >status = my_routine(buf, sizeof(buf));  > ; >or if they wanted to break their own code (but not yours):  > ' >status = my_routine(buf, strlen(buf));     G The probing of every byte is not necessary.  Only the page in which the F byte resides.  Some simple math can determine which byte would requireH a probe of the page to see if it is accessible.  The extra little bit ofH math is far less costly than the probe of every byte -- especially on an0 Alpha where the probe is a PALcode routine call.   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COM              J   "And of course, I'm a genius, so people are naturally drawn to my fiery I   intellect.  Their admiration overwhelms their envy!" -- Calvin & Hobbes    ------------------------------  # Date: Sat, 02 Feb 2002 12:55:35 GMT - From: "John E. Malmberg" <wb8tyw@qsl.network> / Subject: Re: "C" written os's spell "disaster"! * Message-ID: <3C5BE5CB.1060104@qsl.network>   Martin Vorlaender wrote:   > & > Now I get it. Thanks for clarifying. > E > If I wanted to read up on the __PAL_* stuff: where would I find it?  >      See the Compaq C User's Guide.  ; http://www.openvms.compaq.com/commercial/c/index_alpha.htm>      -John  wb8tyw@qsl.network Personal Opinion Only    ------------------------------  $ Date: Sat, 2 Feb 2002 06:46:45 -0800# From: "Tom Linden" <tom@kednos.com> / Subject: RE: "C" written os's spell "disaster"! 9 Message-ID: <CIEJLCMNHNNDLLOOGNJIMEBPEBAA.tom@kednos.com>   H Very strange, if you got to http://www-legacy.digital.com/semiconductor/C you get the right page, for about 3 to 4 seconds then it changes to ? http://www.compaq.com/alphaoem/  and says PAGE NOT FOUND.  Odd!   9 There used to be a wealth of technical info at this site.    > -----Original Message-----4 > From: John E. Malmberg [mailto:wb8tyw@qsl.network]+ > Sent: Saturday, February 02, 2002 4:56 AM  > To: Info-VAX@Mvb.Saic.Com 1 > Subject: Re: "C" written os's spell "disaster"!  >  >  > Martin Vorlaender wrote: >  > > ( > > Now I get it. Thanks for clarifying. > > G > > If I wanted to read up on the __PAL_* stuff: where would I find it?  > >  >  >   > See the Compaq C User's Guide. > = > http://www.openvms.compaq.com/commercial/c/index_alpha.htm>  >  >  > -John  > wb8tyw@qsl.network > Personal Opinion Only  >    ------------------------------  * Date: Sat, 2 Feb 2002 14:51:19 +0000 (UTC), From: "Richard Maher" <maher_rj@hotmail.c0m>/ Subject: Re: "C" written os's spell "disaster"! 1 Message-ID: <a3gud6$npi$1@knossos.btinternet.com>    Hi,   4 Patrick Young <P.Young@unsw.EDU.AU> wrote in message7 news:55f85d77.0202010634.22eea2da@posting.google.com...  > G > The "register" keyword above is a little for "wank" value - BTW - the E > C compiler might also believe this to be true of the programmer who 
 > uses it.   This from sys$examples:uwss.c   # **     <snip>Furthermore, since the < **      __PAL_READ_PS built-in has type uint64 (not directlyD **      compatible with the bit-field definition) it is difficult to@ **      avoid causing a store if one decides to use the bitfieldC **      designations, which is contrary to one of the goals in this  **      routine (see below). <snip>> **     The macro capabilities of C are limited, by comparison, <snip>3 **      ps_temp should not, and likely will not, be B **      materialized other than in a register (since we don't know= **      where we can safely put it) but there is currently no F **      enforceable compiler guarantee of this, or of similar behaviorD **      for a "full-sized" user-written system service.  The link ofE **      the image should fail if allocation of storage results in a C  **      RTL call.   H You *could* drop the /PROTECT linker qualifier but the VMS documentationL will tell you (as confirmed by Stephen Hoffman recently in this very forum!)B that you have just opened up a gaping hole in your now UNprotectedD sub-system. (Actually I no longer believe this to be true but Hoff'sD refusing to tuck me in and tell me everything's gonna be alright :-)  B > "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> wrote in message- news:<R9f68.276$am1.4883@news.cpqcorp.net>... I > > The reason that VMS system services require a length, is to make that  probe I > > possible without too high an overhead.  But it *is* possible to do it  with aJ > > null terminated string *if* you really wanted to do it.  You just need to: > > probe for each byte before reading/testing/copying it.  J This is just typical of the bullshit compromises and contortions one wouldL have to go through in order to program in C. Why talented VMS engineers haveJ spent years socially engineering (nay compelling) the client base to use CI is beyond me. Having said that can anyone please tell me if the following 
 are correct:-   J . Page protection and $setprt work on CPU specifically sized pages and not pageletsF . PROBEx only checks the protection on the first and last pages in the range. (Page or Pagelet?) K . If my address and length span more than two pages then I can't assume the C middle one has the same protection or is accessible by the mode I'm 	 checking. L . EXE$PROBEx checks *every* page (pagelet?) in a longword range of addresses   Brian wrote:  H >The probing of every byte is not necessary.  Only the page in which theG >byte resides.  Some simple math can determine which byte would require I >a probe of the page to see if it is accessible.  The extra little bit of I >math is far less costly than the probe of every byte -- especially on an 1 >Alpha where the probe is a PALcode routine call.   < I thought the problem was that the "string", or should I sayJ stream-of-bytes, could span many pages and in the abscence of a descriptor+ or max len every byte had to be probed. EG:    . Have we hit a page boundry  . Yes - Check access to new page . No - Check next byte for null   J On VMS it is probably reasonable (use to be) to put an arbitrary 65k limitL on the string size. Divide the CPU page size into that and then loop throughK the pages to find the first non-accessible or 65k. Then search sequentially L through the accessible page range untill a null is located or ss$_accvio it.I What happens with those devices that now support 64bit lengths to $qio? I > can't wait to see the io$m_stream or io$m_look_for_null flags.   Regards Richard Maher.  H PS. I hope the IPF port with result in the return of "The Missing" Macro? examples to sys$examples! Fred's probably not up for a Pinochet  reference right now :-)   4 Patrick Young <P.Young@unsw.EDU.AU> wrote in message7 news:55f85d77.0202010634.22eea2da@posting.google.com... B > "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> wrote in message- news:<R9f68.276$am1.4883@news.cpqcorp.net>... I > > The reason that VMS system services require a length, is to make that  probe I > > possible without too high an overhead.  But it *is* possible to do it  with aJ > > null terminated string *if* you really wanted to do it.  You just need to: > > probe for each byte before reading/testing/copying it. > >  > 5 > This is **exactly** my understanding to the letter.  > C > I've accepted NULL terminated UNIX C type strings using executive = > mode routines, for example (also written in C), on Alpha...  > K > if (__PAL_PROBER(c_string_addr, c_string_len, ps_reg & PR$M_PS_PRVMOD) ==  0) >     return(SS$_ACCVIO);  >  > With ps_reg being: > 3 >     register unsigned int ps_reg = __PAL_RD_PS();  > D > In order to find out which mode you were previously in. OpenVMS isG > cunning in that the last two bits of the status register are reserved H > for software use - in this particular case (PRVMOD == defines previous > access mode).  > G > The "register" keyword above is a little for "wank" value - BTW - the E > C compiler might also believe this to be true of the programmer who 
 > uses it. > J > It's not that hard, and C is not that evil. Trust Fred and I on this ;-) > H > The evil C type O/Ses (UNIX/Window(tm)) are all a poor implementation. > G > Now, if you are on VAX - I've forgotten - and don't have a clue! my C H > programming days came at around the end of my VAX days, and I told you > already I have a bad memory.   ------------------------------  $ Date: Sat, 2 Feb 2002 06:52:49 -0800# From: "Tom Linden" <tom@kednos.com> / Subject: RE: "C" written os's spell "disaster"! 9 Message-ID: <CIEJLCMNHNNDLLOOGNJIGECAEBAA.tom@kednos.com>   D Of course, if you coded in a high-level language with range checkingJ you wouldn't have any of these problems, nor would you care which page the byte was on.      > -----Original Message-----@ > From: system@SendSpamHere.ORG [mailto:system@SendSpamHere.ORG]+ > Sent: Saturday, February 02, 2002 2:38 AM  > To: Info-VAX@Mvb.Saic.Com 1 > Subject: Re: "C" written os's spell "disaster"!  >  > ? > In article <55f85d77.0202011559.7926603a@posting.google.com>, - > P.Young@unsw.EDU.AU (Patrick Young) writes: @ > >martin@radiogaga.harz.de (Martin Vorlaender) wrote in message9 > news:<3c5ae124.524144494f47414741@radiogaga.harz.de>... / > >> Patrick Young (P.Young@unsw.EDU.AU) wrote: ? > >> > "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> wrote... A > >> > > The reason that VMS system services require a length, is  > to make thatB > >> > > probe possible without too high an overhead.  But it *is*
 > possible to C > >> > > do it with a null terminated string *if* you really wanted  > to do it. 4 > >> > > You just need to probe for each byte before > reading/testing/copying it.  > >> > > > > = > >> > if (__PAL_PROBER(c_string_addr, c_string_len, ps_reg &  > PR$M_PS_PRVMOD) == 0)  > >> >     return(SS$_ACCVIO); > >>C > >> Errr ...and you get the c_string_len from where (if not from a 
 > parameter)?  > >> > > E > >That's where the "just need to probe for each byte before reading/ ? > >testing/copying it" comes in. The c_string_len becomes 1 and F > >you test if the byte is null if you had access to it, if not null -7 > >use it as data and move on to probing the next byte.  > > 4 > >Though in most cases you would do something like: > > G > >int my_priv_routine(char *null_termed_string, int bufsize_of_string)  > > ) > >and a well behaved user could then do:  > >  > >char buf[80]; > >.... ) > >status = my_routine(buf, sizeof(buf));  > > = > >or if they wanted to break their own code (but not yours):  > > ) > >status = my_routine(buf, strlen(buf));  >  > I > The probing of every byte is not necessary.  Only the page in which the H > byte resides.  Some simple math can determine which byte would requireJ > a probe of the page to see if it is accessible.  The extra little bit ofJ > math is far less costly than the probe of every byte -- especially on an2 > Alpha where the probe is a PALcode routine call. >  > --4 > VAXman- OpenVMS APE certification number: AAA-0001 > VAXman(at)TMESIS(dot)COM > K >   "And of course, I'm a genius, so people are naturally drawn to my fiery K >   intellect.  Their admiration overwhelms their envy!" -- Calvin & Hobbes  >    ------------------------------  * Date: Sat, 2 Feb 2002 15:26:17 +0000 (UTC), From: "Richard Maher" <maher_rj@hotmail.c0m>/ Subject: Re: "C" written os's spell "disaster"! / Message-ID: <a3h0en$och$1@helle.btinternet.com>    Hi,   L On second read I realized that it should be "The Disappeared" Macro examplesJ rather than "The Missing" but then I thought that a Pol Pot Killing FieldsL analogy might be more appropriate given the similarities between "C" and his "Year Zero Program".  < No Toilets - No Strings. There's a certain symmetry to that.   Regards Richard Maher.  L PS. Somethin' tells me that Fred's still not laughing but then he's got bits3 of instruction set strewn over his veggie patch :-)e  L (No disrespect meant to any freinds or family of those who suffered and died under those regimes.)   5 Richard Maher <maher_rj@hotmail.c0m> wrote in messagel+ news:a3gud6$npi$1@knossos.btinternet.com...s > Hi,  >S6 > Patrick Young <P.Young@unsw.EDU.AU> wrote in message9 > news:55f85d77.0202010634.22eea2da@posting.google.com...c > >sI > > The "register" keyword above is a little for "wank" value - BTW - theuG > > C compiler might also believe this to be true of the programmer whop > > uses it. >s > This from sys$examples:uwss.cl >i% > **     <snip>Furthermore, since thet> > **      __PAL_READ_PS built-in has type uint64 (not directlyF > **      compatible with the bit-field definition) it is difficult toB > **      avoid causing a store if one decides to use the bitfieldE > **      designations, which is contrary to one of the goals in this0 > **      routine (see below). > <snip>@ > **     The macro capabilities of C are limited, by comparison, > <snip>5 > **      ps_temp should not, and likely will not, be D > **      materialized other than in a register (since we don't know? > **      where we can safely put it) but there is currently no-H > **      enforceable compiler guarantee of this, or of similar behaviorF > **      for a "full-sized" user-written system service.  The link ofG > **      the image should fail if allocation of storage results in a Co > **      RTL call.r > J > You *could* drop the /PROTECT linker qualifier but the VMS documentationF > will tell you (as confirmed by Stephen Hoffman recently in this very forum!)iD > that you have just opened up a gaping hole in your now UNprotectedF > sub-system. (Actually I no longer believe this to be true but Hoff'sF > refusing to tuck me in and tell me everything's gonna be alright :-) >oD > > "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> wrote in message/ > news:<R9f68.276$am1.4883@news.cpqcorp.net>... K > > > The reason that VMS system services require a length, is to make that_ > probenK > > > possible without too high an overhead.  But it *is* possible to do itr > with aL > > > null terminated string *if* you really wanted to do it.  You just need > to< > > > probe for each byte before reading/testing/copying it. >oL > This is just typical of the bullshit compromises and contortions one wouldI > have to go through in order to program in C. Why talented VMS engineerst haveL > spent years socially engineering (nay compelling) the client base to use CK > is beyond me. Having said that can anyone please tell me if the followings > are correct:-e >nL > . Page protection and $setprt work on CPU specifically sized pages and not
 > pageletsH > . PROBEx only checks the protection on the first and last pages in the > range. (Page or Pagelet?)nI > . If my address and length span more than two pages then I can't assume  theoE > middle one has the same protection or is accessible by the mode I'm  > checking..D > . EXE$PROBEx checks *every* page (pagelet?) in a longword range of	 addressesg >a > Brian wrote: >-J > >The probing of every byte is not necessary.  Only the page in which theI > >byte resides.  Some simple math can determine which byte would require-K > >a probe of the page to see if it is accessible.  The extra little bit of9K > >math is far less costly than the probe of every byte -- especially on anr3 > >Alpha where the probe is a PALcode routine call.s >e> > I thought the problem was that the "string", or should I sayL > stream-of-bytes, could span many pages and in the abscence of a descriptor- > or max len every byte had to be probed. EG:h >V > . Have we hit a page boundry" > . Yes - Check access to new page! > . No - Check next byte for nulld >BL > On VMS it is probably reasonable (use to be) to put an arbitrary 65k limitF > on the string size. Divide the CPU page size into that and then loop throughg@ > the pages to find the first non-accessible or 65k. Then search sequentiallyJ > through the accessible page range untill a null is located or ss$_accvio it. K > What happens with those devices that now support 64bit lengths to $qio? I @ > can't wait to see the io$m_stream or io$m_look_for_null flags. >g > Regards Richard Maher. >sJ > PS. I hope the IPF port with result in the return of "The Missing" MacroA > examples to sys$examples! Fred's probably not up for a Pinocheti > reference right now :-)o >  >m   ------------------------------  % Date: Sat, 02 Feb 2002 11:47:47 +0100_/ From: Paul Sture <paul.sture@bluewin.delete.ch>aM Subject: Re: Be prepared, Encompass is using SBA to SPAM folks for Membershipu4 Message-ID: <VA.00000528.841d6edf@bluewin.delete.ch>  : In article <3C57C5A0.5E2ED461@127.0.0.1>, Nic Clews wrote:  J > I suffer a lot of spam both at work and at home. The actual subject line > used in the email: >  [snip]  I I'm heartily sick of them too. I'm currently trying out Kmail on my LinuxgH box and it has a nice option called "Bounce". I've no idea if it's worth/ doing or even works, but it _is_ satisfying :-)o  H FWIW I've long been a believer that replying to spam with a "remove" (orK whatever) simply serves to confirm your email address as valid and active,  : and in all likelihood won't get you removed from anywhere.  I However, a new breed of this hit me this week. It welcomed me to a stream K of porn, which it would keep on sending UNLESS I clicked on the url, which 1D quite clearly contained my email address (this would have been less F obvious in an HTML mail reader). IMO that was an attempt to get me to   confirm my address as working... ___:
 Paul Sture Switzerlandh   ------------------------------   Date: 2 Feb 2002 09:30:49 -0600f- From: Kilgallen@SpamCop.net (Larry Kilgallen)pM Subject: Re: Be prepared, Encompass is using SBA to SPAM folks for Membershipm3 Message-ID: <U1TxKxTzw9wY@eisner.encompasserve.org>O  f In article <VA.00000528.841d6edf@bluewin.delete.ch>, Paul Sture <paul.sture@bluewin.delete.ch> writes:  K > However, a new breed of this hit me this week. It welcomed me to a streamLM > of porn, which it would keep on sending UNLESS I clicked on the url, which wF > quite clearly contained my email address (this would have been less H > obvious in an HTML mail reader). IMO that was an attempt to get me to " > confirm my address as working...  D In addition to the paid filtering service, SpamCop does offer a free* reporting service at http://spamcop.net/ .   ------------------------------  # Date: Sat, 02 Feb 2002 15:35:55 GMTo4 From: "Terry C. Shannon" <terryshannon@mediaone.net>M Subject: Re: Be prepared, Encompass is using SBA to SPAM folks for Membership = Message-ID: <vHT68.22946$%h1.8317152@typhoon.ne.mediaone.net>p  : "Larry Kilgallen" <Kilgallen@SpamCop.net> wrote in message- news:U1TxKxTzw9wY@eisner.encompasserve.org...wA > In article <VA.00000528.841d6edf@bluewin.delete.ch>, Paul Stureh& <paul.sture@bluewin.delete.ch> writes: >nF > > However, a new breed of this hit me this week. It welcomed me to a streamH > > of porn, which it would keep on sending UNLESS I clicked on the url, whichnG > > quite clearly contained my email address (this would have been lessmI > > obvious in an HTML mail reader). IMO that was an attempt to get me tof$ > > confirm my address as working... >cF > In addition to the paid filtering service, SpamCop does offer a free, > reporting service at http://spamcop.net/ .  L Indeed they do. They also offer a PAID service ($12.50 for 25MB of SPAMmail)I that gives you priority access over the unpaid community. I use it, but InJ really have no way of knowing how effective it is, since the SPAMMERS just keep on coming.c  K Here in the USA the problem is our weenie elected officials, who don't have F the guts to pass effective anti-SPAM legislation. The Direct MarketingK Association lobbyists would be unhappy if said legislation materialized. So H what we get is SPAM with replete with disclaimers like "This is NOT SPAML since we have included a Remove Link." Either the Remove Link does not work,H or it verifies that your email address is active so you can be sent even' more high-quality offers from SPAMMERS.i   ------------------------------  % Date: Sat, 02 Feb 2002 11:47:46 +0100l/ From: Paul Sture <paul.sture@bluewin.delete.ch>o8 Subject: Re: Capellas wants IBM model, but does reverse!4 Message-ID: <VA.00000524.841d6aeb@bluewin.delete.ch>  C In article <87bsfajc57.fsf@prep.synonet.com>, Paul Repacholi wrote:n) > JF Mezei <jfmezei@videotron.ca> writes:e > C > > Now, since then, I did notice IBM becoming more open to non-IBMnG > > technologies (TCPIP, ethernet etc). Remember that IBM didn't want teD > > adopt ethernet and went for its proprietary Token Ring technolgyC > > that was inferior and a real pain.  And IBM had brainwashed itscG > > customers into thinking that ethernet was very inefficient etc etc.s > G > IBM saw standard networking as death... Once it is in place, then thes3 > data can escape the castle. Not good for lock in.i > E > IBM *did* do an ethernet controller, it was called the DACU, DevicebF > Attachment Control Unit. It was a channel to unibus addapter, plus aE > DEUNA!! To order one and get it, you needed the blessing of the IBMe > board in NY. >WE Now _that_ reminds me of the DEC attitude back sometime in the 1980s.o  D I had a salesman/marketing type tell a potential customer that "DEC E does NOT support EBCDIC. We support the international ANSI and ASCII tD standards". I couldn't shut the idiot up. I could easily prove that C ANSI <--> EBCDIC tape transfers were quite easy, and had plenty of tE working examples to prove it, but the customer was really put off by a  DEC's stubbornness on the issue.  @ > Now, IBM seems to go for what ever will get them the long term
 > revinue. >  ___X
 Paul Sture Switzerlanda   ------------------------------  $ Date: Sat, 2 Feb 2002 18:35:57 +0100/ From: Roland Barmettler <roland@barmettler.net>g/ Subject: compaq logo on DECwindows login screenc; Message-ID: <20020202183557.0e62343c.roland@barmettler.net>t   --=.vZGLR?PYLLrCrU* Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bite   Hi  * This may be a stupid question, but anyway:A Up to VMS 7.2-x there was a nice "digital" logo on the DECwindows"> login screen. Since 7.3 it became a rather ugly "Compaq" logo.C Does anybody know how to change that back ? Is it an bitmap of some % sort or hardcoded into the X server ?u   Thanks for any help :-)t Roland  E ---------------------------------------------------------------------(4  FreeStone Systems    |   EMail : roli@freestone.net5  Roland Barmettler    |           roli@barmettler.netp2  Zuercherstrasse 16   |           roli@DeLorean.chA  CH-8107 Buchs  ZH    |   WWW   : http://www.freestone.net/~roli/g   --=.vZGLR?PYLLrCrU' Content-Type: application/pgp-signature    -----BEGIN PGP SIGNATURE-----b! Version: GnuPG v1.0.6 (GNU/Linux)t  @ iD8DBQE8XCOAOZCvraXJF/gRAjb4AKCFCOZFJzgDG0TeIxEn2qowY4r2zwCfePv+ eaZsjVV3xu16w2UwbgQjzKM= =Zdmrr -----END PGP SIGNATURE-----f   --=.vZGLR?PYLLrCrU--   ------------------------------  % Date: Sat, 02 Feb 2002 09:48:13 +0100i From: Dirk Munk <munk@home.nl> Subject: Re: curses on OpenVMS& Message-ID: <3C5BA7CD.7020407@home.nl>   Tony Cheung wrote:   >Hi, >lC >I am trying to create a text-console application on the OpenVMS. I. >would be using C/C++. >eD >I've read about the curses implementations of OpenVMS and know thatC >it's not 100% compatible with other curses. Is that still true for  >OpenVMS v7.3? > D >Is there any other easier way to create a text-console with hilightD >menus, text-entry form? Or any other higher level libraries to use? >cG Yes, many. Look at DecForms as a product you can use for this purpose. rI It will give you a text-console front-end for your application. There is  E also a library with these kind of routines, but I will have to ask a  J collegue about its name. Compaq Cobol has text console statements as well.       > & >Any pointer will be much appreciated. >n >Thank you very much.u >g >Tony Cheung >_   ------------------------------  % Date: Sat, 02 Feb 2002 12:53:45 -0500n' From: Howard S Shubs <howard@shubs.net>t Subject: Re: curses on OpenVMS< Message-ID: <howard-6FE67B.12534502022002@enews.newsguy.com>  F In article <3C5BA7CD.7020407@home.nl>, Dirk Munk <munk@home.nl> wrote:  I > Yes, many. Look at DecForms as a product you can use for this purpose. vK > It will give you a text-console front-end for your application. There is dG > also a library with these kind of routines, but I will have to ask a tL > collegue about its name. Compaq Cobol has text console statements as well.  $ You're thinking of the SMG$ library.   -- c Howard S ShubsD "Run in circles, scream and shout!"  "I hope you have good backups!"' Aren't there any networked SJFs around?-   ------------------------------  % Date: Sat, 02 Feb 2002 11:47:47 +0100L/ From: Paul Sture <paul.sture@bluewin.delete.ch>n, Subject: Re: Default terminal printer in CDE4 Message-ID: <VA.00000527.841d6ed5@bluewin.delete.ch>  W In article <e319d183.0202010050.5dba3245@posting.google.com>, Christoph Gartmann wrote:o > rdeininger@mindspring.com (Robert Deininger) wrote in message news:<rdeininger-3101021826480001@1cust172.tnt2.nashua.nh.da.uu.net>...mT > > In article <a3bevb$llt$1@n.ruf.uni-freiburg.de>, gartmann@immunbio.mpg.de wrote: > >  > > >Hello,: > > >"O > > >how do I define a default printer for my DECterm windows under CDE? I have0O > > >a logical SYS$PRINT defined jobwide that points to my preferred queue. Buti, > > >this doesn't help. I tried to to put a 4 > > >  $ DEFINE decw$printer_format_default "LASER7"T > > >in SESSIONETC.COM, no luck either. Setting the printer from within the terminalM > > >window and saving the settings in DECW$TERMINAL_DEFAULT.DAT doesn't helpsO > > >either. The printer setting doesn't seem to get saved in there. Thus, whatu > > >else could I try? > > > 
 > > >Regards,e > > >   Christoph Gartmann > > I > > in DECwindows Motif, you can define logical names which control whicheI > > queues are listed in the Print... dialog boxes.  Something like this:  > > 7 > > $ define decw$printer_format_text "queue_1,queue_2"e? > > $ define decw$printer_format_ps   "queue_3,queue_4,queue_5"  > > M > > Only the queues you list will show up in the dialog boxes.  (Don't ask me)0 > > where this is documented, I don't remember.) > > M > > I don't know if the CDE print mechanism honors these logical names.  I'vee > > never clicked that icon. > G > I tried various possibilities and found that the following line in my  > LOGIN.COM did the trick:2 >    $ DEFINE decw$printer_format_default "LASER7"B > Of course, I had to log out from CDE and login again in order to > activate the settinig. > L OTOH, I've been happily working for ages with the following LOGIN.COM entry:  +  $ DEFINE SYS$PRINT generic_dcps_queue_name) ___ 
 Paul Sture Switzerlandt   ------------------------------   Date: 2 Feb 2002 15:55:25 GMTi3 From: gartmann@immunbio.mpg.de (Christoph Gartmann)y, Subject: Re: Default terminal printer in CDE0 Message-ID: <a3h25d$jg1$1@n.ruf.uni-freiburg.de>  f In article <VA.00000527.841d6ed5@bluewin.delete.ch>, Paul Sture <paul.sture@bluewin.delete.ch> writes:X >In article <e319d183.0202010050.5dba3245@posting.google.com>, Christoph Gartmann wrote: >> rdeininger@mindspring.com (Robert Deininger) wrote in message news:<rdeininger-3101021826480001@1cust172.tnt2.nashua.nh.da.uu.net>...U >> > In article <a3bevb$llt$1@n.ruf.uni-freiburg.de>, gartmann@immunbio.mpg.de wrote:c >> > i >> > >Hello, >> > >P >> > >how do I define a default printer for my DECterm windows under CDE? I haveP >> > >a logical SYS$PRINT defined jobwide that points to my preferred queue. But- >> > >this doesn't help. I tried to to put a "5 >> > >  $ DEFINE decw$printer_format_default "LASER7" U >> > >in SESSIONETC.COM, no luck either. Setting the printer from within the terminalaN >> > >window and saving the settings in DECW$TERMINAL_DEFAULT.DAT doesn't helpP >> > >either. The printer setting doesn't seem to get saved in there. Thus, what >> > >else could I try?- >> > > >> > >Regards, >> > >   Christoph Gartmanne >> > iJ >> > in DECwindows Motif, you can define logical names which control whichJ >> > queues are listed in the Print... dialog boxes.  Something like this: >> > s8 >> > $ define decw$printer_format_text "queue_1,queue_2"@ >> > $ define decw$printer_format_ps   "queue_3,queue_4,queue_5" >> > wN >> > Only the queues you list will show up in the dialog boxes.  (Don't ask me1 >> > where this is documented, I don't remember.)  >> > 5N >> > I don't know if the CDE print mechanism honors these logical names.  I've >> > never clicked that icon.k >> uH >> I tried various possibilities and found that the following line in my >> LOGIN.COM did the trick: 3 >>    $ DEFINE decw$printer_format_default "LASER7"nC >> Of course, I had to log out from CDE and login again in order to  >> activate the settinig.  >> oM >OTOH, I've been happily working for ages with the following LOGIN.COM entry:s > , > $ DEFINE SYS$PRINT generic_dcps_queue_name  O I had that, too but suddenly it stopped working and I couldn't find out why. Itt( might be due to some patch I installed.    Regards,    Christoph Gartmann   H -- --------------------------------------------------------------------+H | Max-Planck-Institut fuer      Phone   : +49-761-5108-464   Fax: -452 |H | Immunbiologie                                                        |H | Postfach 1169                 Internet: gartmann@immunbio.mpg.de     |H | D-79011  Freiburg, FRG                                               |H +--------- http://www.immunbio.mpg.de/home/english/menue.html ---------+   ------------------------------  $ Date: Sat, 2 Feb 2002 07:10:31 -0800# From: "Tom Linden" <tom@kednos.com>a Subject: LICENSE/CHARGEf9 Message-ID: <CIEJLCMNHNNDLLOOGNJIMECAEBAA.tom@kednos.com>o  - Is there a site which displays the results of  $ SHOW LICENSE/CHARGE   for all the differnet platforms?  F In particular, I am interested in ES40,45. which if someone could post would be much apreciated.i   ------------------------------  % Date: Sat, 02 Feb 2002 11:47:47 +0100m/ From: Paul Sture <paul.sture@bluewin.delete.ch>a% Subject: Re: OpenVMS CD to 5.25 disks 4 Message-ID: <VA.00000526.841d6e35@bluewin.delete.ch>   In article aE <2795B75EF003D311801A00A0C906B511011C6C5A@cucexec.gbc.getronics.nl>, i Jeroen van Dijk wrote:H > For the MicroVAX II and the vaxstation 2000 the max VMS version is 7.2/ > The MicroVAX 3100 has no max VMS version yet.o > 7 Hmm. Isn't Charon-VAX an emulation of the MicroVAX II? t ___t
 Paul Sture Switzerlandm   ------------------------------   Date: 2 Feb 2002 07:15:40 GMTg- From: djweath@attglobal.net (Dave Weatherall)eY Subject: Re: Proliant better than Alphas  was (Re: Alphaserver 4100: physical memory) memr5 Message-ID: <DTiotGxQ0bj6-pn2-lmuXp4X0i8ZQ@localhost>   D On Fri, 1 Feb 2002 11:09:53 UTC, Nic Clews <sendspamhere@127.0.0.1>  wrote:   > Fabio Cardoso wrote: > > 6 > > You cannot deny that in a few years OpenVMS, Linux/ > > and WNT or similar will run in the same boxa- > > concurrently. Sharing the same hardware !i0 > > Intel problems were not their processors but( > > the bad developed operating systems. > C > After some while hacking about with my home AMD based system, andp> > conducting my own research (and other observations) into theJ > instabilities I'm experiencing, I think this conclusion is correct. I'llE > spare you the gruesome details of having to follow procedures whicheG > included clicking my heels three times, and resulted in corrupting myc > system disk. > F > I know at home I need a board and CPU upgrade, but I'm steering wellI > clear of VIA chipsets and AMD processors. Perhaps on paper AMD may lookaC > better, but the field tells a different story. Whilst I'm not the B > world's biggest fan of M$ operating systems, in my case they areC > absolved. There is software, hardware, and what I now describe as  > "sufferware".a > J > It's a P4 and a mobo with an Intel chipset for me, I personally will not# > consider AMD for my next upgrade.d  & Well, I can't let that implication go.  E I've not had an Intel CPU in any of my machines since I replaced the  F Intel 8088 in my Rainbow with a NEC V20 (over 15years ago). I've used D mainly AMD and occasionally Cyrix. (386,486,M1,586,K6-2,K6-3,K7 and C finally Athlon). All with a mix of Ali, Via, SiS and Intel support  F chips.They are all still working and, from the 5/686 on,  still giving5 service under Os/2, Win9x, NT and occasionally Linux.$  F The only _real_ problem I had was with the Cyrix, it used to BSOD NT aD lot, but that turned out later to be because I was mixing Fast Page E Ram and EDO simms!! Strangely enough though, OS/2 _never_ crashed it jC !! Since sorting the memory out I don't remember a Bsod either but s. then its mainly a Linux machine these days :-)  E I have had trouble with Interrupt allocation but again that was only  D under W9x where the Plug and Pray behaviour of Windows and the Bios ? would get into violent disagreement about which card got which tC interrupt. That was eventually overcome by telling W9x to but out. a Then I had stability.u  E There are recent stories of bugs in Via chipsets but they, according yD to C'T magazine, have more to do with Creative Labs not adhering to  the PCI spec. than VIA bridges.s  F I use my machine for music and the occasional Rally game. I don't use  Creative labs cards any more.r  B So I would not have any problem with AMD machines running VMS. In C fact, as and when I get around to obtaining the Charon-VAX demo, I I; missed it the first time around,  I might well be doing so!l   -- s Cheers - Dave.   ------------------------------  % Date: Sat, 02 Feb 2002 11:47:46 +0100e/ From: Paul Sture <paul.sture@bluewin.delete.ch> ( Subject: Re: setting the record straight4 Message-ID: <VA.00000525.841d6af5@bluewin.delete.ch>  < In article <3C5AD30B.B4FBE467@videotron.ca>, JF Mezei wrote:   [snip] > L > Fine, I will leave you alone. But don't expect ANY sympathy in a few yearsM > when you are sold to Microsoft. I won't even bother telling you "I TOLD YOU,L > SO". And meanwhile I will silenty become in the second class of people whoC > have "flaked off" and do tell customers to avoid anything Compaq.h > K Please go silently then. Your repeated cries of "I told you so" are wearingo( very thin, and don't help your argument.  E Your constant negativity quite frankly gets me down, and I'm sure I'm D not the only one. At times your purpose in posting seems to me to beH solely to spread Fear, Uncertainty and Doubt, all in the name of "I told you so".  F Stop trying to drag us down to your level. In my experience Pessimists3 don't get anywhere, and I don't intend to join you.a ___e
 Paul Sture Switzerlandr   ------------------------------  $ Date: Sat, 2 Feb 2002 09:28:06 -0500+ From: "Main, Kerry" <Kerry.Main@Compaq.com>pN Subject: RE: SGI solicits Alpha customers to trade in their "orphaned" systemsT Message-ID: <BE56C50EA024184DAF48F0B9A47F5CF401AB1C1A@kaoexc01.americas.cpqcorp.net>  D mmm... re: SGI loyalty to its users - amazing how marketeers like to rewrite history eh ?  . >>> SGI introduces the concept of loyalty. <<<  $ Reference - articles from the past :  3 http://www.zdnet.com/eweek/news/0413/14msgi.html=20o< "SGI charts new business course, forges alliance with Intel"A " NEW YORK -- Silicon Graphics Inc. will work with Intel Corp. to ? incorporate the latter's processors in its systems, the companyi announced today.",  7 http://news.com.com/2100-1001-229246.html?legacy=3Dcnetp@ "Workstation maker SGI has released its first Intel-based serverE computers with both Microsoft Windows NT and Linux operating systems,a5 but Linux is getting top billing in the long run... "n   :-)o   Regards,  
 Kerry Main Senior Consultante Compaq Canada Corp.e Professional Services  Voice: 613-592-4660  Fax  :  819-772-7036 Email: Kerry.Main@Compaq.com     -----Original Message-----7 From: Zane H. Healy [mailto:healyzh@shell1.aracnet.com]  Sent: January 31, 2002 7:58 PM To: Info-VAX@Mvb.Saic.ComhF Subject: Re: SGI solicits Alpha customers to trade in their "orphaned" systems     5 Carl Karcher <karcher@thuria.waisman.wisc.edu> wrote:s > Reason 1:a  H > SGI introduces the concept of loyalty. Like many Compaq users, you may: > be disturbed by the news that Compaq abandoned its Alpha microprocessors4@ > in favor of Intel(R) Itanium(R) architecture. Well, don't feel	 orphaned.lA > SGI is well known for loyalty to our users, and we've created a1 specialeE > offer to let you trade in your old Alpha systems and get a terrificp dealG > on the new SGI(R) Origin(R) 300 server. But don't delay--the Februaryt 28 > deadline rapidly approaches.  F I find this statement from SGI a little odd.  A few years ago, weren't theyF going to dump IRIX in favor of Windows NT and MIPS in favor of x86?  I seelC they did just release some interesting new MIPS based workstations,t	 but...=20tH Then there is that issue of all the patents that they apparently sold to; Microsoft just recently (or did they simply license them?).e   			Zanea   ------------------------------   Date: 2 Feb 2002 09:08:37 -0800a From: germanhd@usa.net (Holger)l4 Subject: Re: Storageworks RAIDArray 310 INVLID CACHE= Message-ID: <54239d74.0202020908.303d2d24@posting.google.com>a  C Uwe, many thanks for your detailed instructions and immediate help!   C Meanwhile the array is up and everything seems to be fine. I worked A through all the CLI commands and thing I now got it all together. B There is one very strange behaviour left, I meanwhile got used to:@ After certain commands, like adding units and enabling writebackC caching, or after sending commands very quickly after each other, It2 usually get no correct response. Usually it shows:  > Copyright Digital Equipment Corporation 1993, 1997. All rights	 reserved.e4 HSZ20 Firmware version V31Z-2, Hardware version  E01   Last fail code: 20090010    Press " ?" at any time for help.  ) Cache battery is now sufficiently charged:  E >>I forced this by decharging (simply unplugging) the battery severalaC times, and now the battery is fine again and was able to backup the4) cache for 4 hours and probably longer!!>>L   Power Supply failure cleared.h  8 >>I unplugged one powersupply to see what would happen<<   HSZ20>  C !! But if one is patient enough to wait a few seconds with the nexta command, it works. !!w  < The last question would be if there is any update, you would recommend, for the firmware.   Holger  G zessin@decus.de wrote in message news:<00A08E05.B1F63674.1@decus.de>...  > Holger wrote:lI > > Thank you for the hint with the HSZ40. On that page I found only real-= > > manuals for the HSZ50 but the CLI manual helped somewhat.n > J > Oops, sorry, I didn't notice it was just the SPD. I'll see if I can find > a manual in our library. > P > > Your command sequence was approved by the HSZ50 CLI manual, but didn't help. >  > Did you get any errors?o > L > > I first removed all drives, than I decided to use the screwdriver methodN > > and removed the battery from the controller. I expected to clear the cache' > > that way but it didn't really help.S > K > The HS controllers leave a signature in the cache and are very picky when & > you change components the wrong way. > I > > It just helped me to start searching for a place to get that battery.. > > I found it here:3 > > http://www.portable-energy.com/prod/cell50.htmlaI > > BUT now the command sequence worked and the controller seems to be upI > > and running! > $ > Strange. Did you log your session?  C >>>Explanation is probably the before described behavior, regarding  the speed of command input<<<h   ( > > So now I will start to configure it.H > > IS there any COMMAND that will reset the whole configuration and let. > > me start from scratch with some new disks? > * > You might have seen that the HSZ50 uses:/ >     SET THIS_CONTROLLER INITIAL_CONFIGURATIONp  
 >>>Correct<<<s  D > Recent controllers have the 'CONFIGURATION RESET' command and someG > controllers can be reset with a combination of push buttons - I don'twG > know about the HSZ20. Unless someone else can help you should be ablerE > to delete any objects starting from the UNIT level and then workinge; > through the storage-set level down to the phsyical disks.a   ------------------------------  $ Date: Sat, 2 Feb 2002 07:46:42 -0800# From: "Tom Linden" <tom@kednos.com>  Subject: tcpip9 Message-ID: <CIEJLCMNHNNDLLOOGNJIKECBEBAA.tom@kednos.com>   I I often access nodes in our cluster using putty on a W2K box.  One of thet nodesiI runs UCX (TCPIP V4.2) on 7.1-2 and another 5.1 on 7.3  I thought I was onl theyC latter and typed TCPIP, but I was on the former .  It printed out aa
 TCPIPtraceL of XMT and RCV packets.  $ TC also works and a search of the system revealed2 nothing, I looked for logicals or files named tc*.  L Where does this come from?  Something like the unix 'which' command would be nice.F   ------------------------------  % Date: Sat, 02 Feb 2002 12:32:21 -050072 From: rdeininger@mindspring.com (Robert Deininger) Subject: Re: tcpipK Message-ID: <rdeininger-0202021232220001@1cust178.tnt3.nashua.nh.da.uu.net>.  F In article <CIEJLCMNHNNDLLOOGNJIKECBEBAA.tom@kednos.com>, "Tom Linden" <tom@kednos.com> wrote:e  J >I often access nodes in our cluster using putty on a W2K box.  One of the >nodesJ >runs UCX (TCPIP V4.2) on 7.1-2 and another 5.1 on 7.3  I thought I was on >theD >latter and typed TCPIP, but I was on the former .  It printed out a >TCPIPtrace M >of XMT and RCV packets.  $ TC also works and a search of the system revealedh3 >nothing, I looked for logicals or files named tc*.o >sM >Where does this come from?  Something like the unix 'which' command would be  >nice.  G I don't have access to a UCX 4.2 system at the moment, so I'm guessing.i  & I also don't know what 'which' does...  @ At the DCL level, there are at least 3 ways to invoke a command:  H 1. A regular DCL verb in the process's command table.  These are createdJ (by a product kit or end user) via SET COMMAND.  The freeware VERB utilityJ can reverse the process, producing a .CLD file representing a piece of theG command table.  VERB is stupid about abbreviations.  If there's a TCPIPiI command defined, an abbreviation will invoke it from DCL, but you'll havetI to spell it out completely for VERB.  (Perhaps this has changed in recente versions of VERB.)  F 2. A DCL symbol.  Something like TC*PIP :== whatever would do what youE describe. The "whatever" part could involve either a command file, an'F executable image, or a built-in verb.  The "*" character in the symbolE definition indicates the shortest abbreviation that DCL will accept. nB Anything between TC and TCPIP would be recognized in this example.  G 3. The DCL$PATH logical name mechanism.  I never use it, so I shouldn't I say much.  My vague recollection is that is does NOT do abbreviations, so'& it likely isn't involved in this case.  J The last two mechanisms should be described in the VMS User's guide and/or the DCL dictionary.t  D Logical names are not directly involved in invoking commands in DCL.  2 I have no idea if that answers your question.  :-)   ------------------------------  % Date: Sat, 02 Feb 2002 11:56:36 +0100$9 From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com>5? Subject: The Inquirer says : "Compaq thrashes IBM, HP, Fujitsu"i' Message-ID: <3C5BC5E4.997F73FC@aaa.com>b  ' http://www.theinquirer.net/01020215.htmn  6 WHEN WE VISITED Compaq's swanky City HQ near St Paul's: Cathedral yesterday, Richard George, who does an extremely@ good job evangelising Alpha and the servers which carry us, gave& us a copy of the latest TPC-C results.  < And very impressive they are too, with the ES45 spanking the pants off Fujitsu, HP, and IBM.w  < George also claimed that Compaq had managed to increase Unix8 market share in the UK following the announcement of the9 Intel-Compaq technology swap, and the HP-Compaq takeover.n  = He said that Compaq was on track to produce several new Alphat? servers during this year using the "copper whopper" 1GHz Alpha.a  ? And there are speed ramps ahead for the Alpha too, with figures>; like 1.2GHz and 1.5GHz being mentioned in the conversation.t  6 You can find the Compaq TPC-C results by clicking here& (http://www.theinquirer.net/tpcc.jpg).  B Meanwhile, the folks at Openvms.org have a fresh survey for you to	 sample. ,     Jan-Erik Sderholm.b   ------------------------------  % Date: Sat, 02 Feb 2002 12:13:30 -0500 2 From: rdeininger@mindspring.com (Robert Deininger)C Subject: Re: The Inquirer says : "Compaq thrashes IBM, HP, Fujitsu"rK Message-ID: <rdeininger-0202021213310001@1cust178.tnt3.nashua.nh.da.uu.net>t  0 In article <3C5BC5E4.997F73FC@aaa.com>, Jan-Erik1 =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com> wrote:d  ( >http://www.theinquirer.net/01020215.htm >n7 >WHEN WE VISITED Compaq's swanky City HQ near St Paul's ; >Cathedral yesterday, Richard George, who does an extremelyfA >good job evangelising Alpha and the servers which carry us, gavee' >us a copy of the latest TPC-C results.i >h= >And very impressive they are too, with the ES45 spanking thes  >pants off Fujitsu, HP, and IBM.  H There's no question the ES45 is a very nice system.  The designers did aJ great job, and deserve more credit for improving Compaq's bottom line than they'll probably get.1   ------------------------------  # Date: Wed, 30 Jan 2002 13:29:36 GMTc$ From: "Ed Wilts" <ewilts@ewilts.org>D Subject: Re: Using the 'latest and greatest' (was:VMS 7.3 stability)8 Message-ID: <4zS58.455$Sc.32947@typhoon.mn.mediaone.net>  K > Unless you have a compelling reason for using a new version (for example,e XFCuE > functionality that you *must* have :-)), you might want to considern
 > waiting.  7 XFC *must* currently be disabled, so that wasn't it :-)c   -- Ed Wilts Mounds View, MN, USA mailto:ewilts@ewilts.org   ------------------------------  % Date: Sat, 02 Feb 2002 12:01:25 -0500h2 From: rdeininger@mindspring.com (Robert Deininger); Subject: Re: VMS booted on Itanic according to The Inquirer K Message-ID: <rdeininger-0202021201250001@1cust178.tnt3.nashua.nh.da.uu.net>d  J In article <3C5B3E83.109952A2@rcn.com>, Jonas Lindholm <jlindholm@rcn.com> wrote:  I >does Itanium support all four access modes in hardware or do you need tol >simulate that instead in SW ?   Itanium has 4 access modes.n   ------------------------------   End of INFO-VAX 2002.064 ************************