1 INFO-VAX	Fri, 16 Mar 2001	Volume 2001 : Issue 149       Contents:' Re: - OpenVMS ever to be on Intel chip?  Re: /PARENT=LNM$SYSCLUSTER Re: Alpha on ABC Evening news 1 Re: Alphastation XP900 / poor Graphic performance  Re: apache performance BASIC DIMENSION and debugging  Calculate Easter in DCL ) Compaq earnings low, 5,000 jobs to be cut  Re: DCL minute of the day  Re: DCL minute of the day  Re: DCL minute of the day P RE: Dumbing Down VMS with UNIX Elements? (was Re: OpenVMS Educati	onal    PrograP RE: Dumbing Down VMS with UNIX Elements? (was Re: OpenVMS EducationalProgram) PrP Re: Dumbing Down VMS with UNIX Elements? (was Re: OpenVMSEducational     ProgramP Re: Dumbing Down VMS with UNIX Elements? (was Re: OpenVMSEducationalProgram) Pro Re: fmsdef.h Re: fmsdef.h* Re: Help!  PASCAL is read-only code to me.* Re: Help!  PASCAL is read-only code to me.* Re: Help!  PASCAL is read-only code to me.* Re: Help!  PASCAL is read-only code to me.% Re: How does one become a VMS guru ?? % Re: How does one become a VMS guru ?? / Re: How to measure system load and availability / Re: How to measure system load and availability  Re: HP Printer on VMS 6.2  Re: HP Printer on VMS 6.2 P Re: I seem to be in a different reality today, one where VMS andMarketing go togP Re: I seem to be in a different reality today, one where VMS andMarketing go tog/ Re: In-memory layout of PACKED ARRAY OF BOOLEAN $ Re: Low cost server - Sun, of course Re: Merging multiple disks RE: Merging multiple disks Re: Merging multiple disks! Re: Obtaining 7.2-1 and TCPIP 5.1 ! Re: Obtaining 7.2-1 and TCPIP 5.1 ! Re: Obtaining 7.2-1 and TCPIP 5.1 ! Re: Obtaining 7.2-1 and TCPIP 5.1  Re: OpenVMS Educational Program  Re: OpenVMS Educational Program  RE: OpenVMS Educational Program  Re: OpenVMS Educational Program  Re: OpenVMS Educational Program   Re: Possible security hole in...  Re: Possible security hole in... Re: Splitting a Large File) Talk to Rich Marcello - Austin Texas area - Re: Talk to Rich Marcello - Austin Texas area - Re: Talk to Rich Marcello - Austin Texas area  TCP/IP SYN flood DoS protection # TDF change problems with V7.3 EFT2?  The Chris and Bill show. Act II * the Northeast Digital Systems Ebay Follies. Re: the Northeast Digital Systems Ebay Follies. Re: the Northeast Digital Systems Ebay Follies. Re: the Northeast Digital Systems Ebay Follies. Re: the Northeast Digital Systems Ebay Follies Re: VMS wanted list.% Re: Was it here?  1GB disk for VS3100 % Re: Was it here?  1GB disk for VS3100   Re: Write CD on PC for OVMS Read Re: [fun] DCL minute of the day  Re: [INFO] RDMS-F-NOMONHOMEDIR Re: [INFO] RDMS-F-NOMONHOMEDIR- Re: [Q] C errors exit handling (and training) ! Re: [Q] skip [vms$common] parsing ! Re: [Q] skip [vms$common] parsing   F ----------------------------------------------------------------------  # Date: Fri, 16 Mar 2001 03:21:04 GMT ; From: Mark Garrett <Mark.Garrett@wedontwantyourspam.com.au> 0 Subject: Re: - OpenVMS ever to be on Intel chip?C Message-ID: <B6D7D04F.146CD%Mark.Garrett@wedontwantyourspam.com.au>   > in article Jrq0TPndqEpX@eisner.encompasserve.org, Rob Young at4 young_r@encompasserve.org wrote on 14/03/2001 00:42:   >>     >>> O >> http://groups.google.com/groups?q=pfister+infiniband+group:comp.arch&hl=en&l ) >> r=&safe=off&rnum=1&seld=924890734&ic=1  >>    K Great infiniband looks like some old concepts ;)   Looks like and update of & an RH20 channel for the new century ;)    
     Cheers         mark ;)    ------------------------------  % Date: Thu, 15 Mar 2001 14:52:47 -0500 7 From: David Beatty <David.Beatty@sasSPAMITUPTHEZZZ.com> # Subject: Re: /PARENT=LNM$SYSCLUSTER 2 Message-ID: <HRyxOn5zWD+919ZcXvpLdd=Qn1PN@4ax.com>  2 That would be correct, David.  Some sample output:  = $ sh log LNM$SYSCLUSTER_TABLE/full/table=lnm$system_directory B    "LNM$SYSCLUSTER_TABLE" [kernel,no_alias,table] = "" [terminal]  (LNM$SYSTEM_DIRECTORY)9 $ sh log LNM$system_TABLE/full/table=lnm$system_directory >    "LNM$SYSTEM_TABLE" [kernel,no_alias,table] = "" [terminal]  (LNM$SYSTEM_DIRECTORY)  + $ sh log * /table=LNM$SYSCLUSTER_TABLE/full   4 (LNM$SYSCLUSTER_TABLE)  [kernel]  [shareable,system]D                         [Protection=(RWC,RWC,R,R)]  [Owner=[SYSTEM]]   ... output removed ...  ' $ sh log * /table=LNM$SYSTEM_TABLE/full   4 (LNM$SYSTEM_TABLE)      [kernel]  [shareable,system]D                         [Protection=(RWC,RWC,R,R)]  [Owner=[SYSTEM]]   ... output removed ...  9 It looks like LNM$SYSCLUSTER_TABLE (pointed at by logical 9 name LNM$SYSCLUSTER) is simply another logical name table ; with the same privileges as LNM$SYSTEM_TABLE.  I don't know * if it's a child table of LNM$SYSTEM_TABLE.  >     Also, the cluster table appears in the search list because: LNM$SYSTEM points to both the system table and the cluster table.  5 On Thu, 15 Mar 2001 16:52:15 +0000, David J. Dachtera ' <donotreply@interbulletin.bogus> wrote:   R >>> DECUS::INFO_VAX                                    24 lines  15-MAR-2001 13:38R >>> ------------------------------------------------------------------------------) >>> Date: Thu, 15 Mar 2001 13:29:31 +0200 5 >>> From: Phillip Helbig <HELBPHI@SYSDEV.EXCHANGE.DE> # >>> Subject: /PARENT=LNM$SYSCLUSTER  >>>   K >>> OK, I'm pretty sure this will get through, so if I see no response I'll 9 >>> have to conclude that everyone is clueless.  :-)  :-)  >>>   
 >>> CREATE >>>    >>>   /NAME_TABLE  >>>    >>>     /PARENT_TABLE  >>>   ! >>>           /PARENT_TABLE=table  >>>   J >>>        Requires either create (C) access to the parent table and writeF >>>        (W) access to the system directory or the SYSNAM privilege. > H >I would think that since "LNM$SYSTEM" .NES. "LNM$SYSCLUSTER", that the O >LNM$SYSCLUSTER table is treated like any other shared logical name table, and  J >the privileges required to modify it would depend upon its ownership and 0 >protection code as well as your process's UIC.  > + >I could be (probably am) wrong, as always.    ------------------------------  # Date: Fri, 16 Mar 2001 03:31:04 GMT ; From: Mark Garrett <Mark.Garrett@wedontwantyourspam.com.au> & Subject: Re: Alpha on ABC Evening newsC Message-ID: <B6D7D33F.146CE%Mark.Garrett@wedontwantyourspam.com.au>   F in article OF0184D257.AC804105-ON80256A0E.005F2F4D@qedi.quintiles.com,L steven.reece@quintiles.com at steven.reece@quintiles.com wrote on 14/03/2001 04:22:   > < > Or DECsystem 20 via a wireless lan and a terminal server ?  B Maybe a DECSYSTEM-20 or a DECsystem-10 but never a DECsystem 20 :)   ------------------------------  # Date: Thu, 15 Mar 2001 20:24:30 GMT 2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman): Subject: Re: Alphastation XP900 / poor Graphic performance0 Message-ID: <2y9s6.35$eE2.1823@news.cpqcorp.net>  F In article <3ab0dafc$1@hcwe67>, "Jakob Erber" <erberj@post.ch> writes:A :  DIGITAL TCP/IP Services for OpenVMS Alpha Version V4.2 - ECO 1 @ :  on a COMPAQ AlphaStation XP900 466 MHP running OpenVMS V7.1-2 : J :On this relativly new WS, the performance of Screen IO is awfully bad. ItK :takes minutes for complex graphics to build up. For example those procuced  :by our case tool.  M :But it seems to be a software problem, since there should be no problem with 4 :VMS 7.2. Does anybody no about a patch for VMS 7.1?  K   I do not follow your logic.  I might well suspect the graphics controller K   is a slow one (you do not indicate which controller), or that the system  I   is memory constrained.  This could also be the application itself, as I J   have seen applications that have faulted themselves into oblivion.  And F   yes, there could be a problem with OpenVMS or DECwindows here, too.   L   ECOs are available for OpenVMS and for DECwindows, particularly apply the I   GRAPHICS ECO kit and the collection of other mandatory ECO kits.  Start *   by applying all current kits for V7.1-2.  H   Your TCP/IP Services version is old, though that may be (but probably J   not) related to the graphics performance you are seeing.  Also consider    an OpenVMS upgrade to V7.2-1.   E   Graphics adapter?  Local or purchased CASE tool?  Is this complaint E   specific to this particular workstation or is this generic across a E   group of workstations?   Are any workstations operating better than -   others?  (If so, what are the differences?)   J   All that said, there is insufficient configuration information, problem H   details, and related and necessary evidence included here for any sortF   of specific recommendations or answer.  To proceed, you will want toF   examine the particular controller involved -- an ELSA GLoria SynergyH   is not a fast controller -- and you will want to carefully examine theG   system activity.  Is it paging?  Is it memory constrained?  Is it CPU I   constrained?  What is going on?  An example of a systematic performance I   investigation is included in the performance documentation, that is the H   sort of methodology that tends to become useful here.  (In addition toH   the system information, a review of what the application is doing alsoF   becomes interesting -- particularly if it is an application that you   support, of course.)  N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Thu, 15 Mar 2001 20:13:01 +0100 2 From: martin@radiogaga.harz.de (Martin Vorlaender) Subject: Re: apache performance ; Message-ID: <3ab1143d.524144494f47414741@radiogaga.harz.de>   3 David Mathog (mathog@seqaxp.bio.caltech.edu) wrote: 1 > "Jan Mnnich" <moennich@greenstream.de> writes: M > >the apache server produces many page faults when executing any cgi-script. M > >on our test-machine (alphaserver 1000) the performance is horrible! i have O > >to wait at least 3 seconds for each generated page! is there any solution to  > >increase the performance? > H > If you're really seeing page paults then it sounds like you're out of  > physical memory.  Buy more.  > G > If that isn't it, then the odds are good that running Apache off of a M > RAMdisk would help.  The problem may be that apache or your script is doing M > a lot of quick file operations - and that tends to be not so quick on VMS.  L > Is your script perl? Does it call module after module after module?  Each F > of those will take time to fetch from disk.  If you can fit your webI > distribution, or at least the CGI related part, entirely into memory it , > will likely run at least 10 times faster.   D Assuming that the perl script is written cleanly, by using mod_perl,G loading and compiling lots of source is not an issue (happens only once  per script).   cu,    Martin --J One OS to rule them all       | Martin Vorlaender  |  VMS & WNT programmer7 One OS to find them           | work: mv@pdv-systeme.de J One OS to bring them all      |   http://www.pdv-systeme.de/users/martinv/> And in the Darkness bind them.| home: martin@radiogaga.harz.de   ------------------------------  $ Date: Mon, 5 Mar 2001 10:15:53 -07005 From: "cstranslations" <cstranslations@email.msn.com> & Subject: BASIC DIMENSION and debugging) Message-ID: <e2DhMfZpAHA.291@cpmsnbbsa09>   3 Here comes the rocket science question of the day -   L What does the debugger do with name of a BASIC array declared with a dynamic bound??   
 Specifically:    [foobar.bas] option type = explicit   declare integer UpperBound    input "UpperBound: "; UpperBound, dimension string FooStrings(1 to UpperBound)  
 print "hello"    $ basic/noopt/debug foobar $ link/debug foobar   L If I set a break on the print and then examine FooStrings the compiler tellsL me the symbol doesn't exist (as opposed to being optimized away). I've spent= the last 3 hours going round in circles. What am I missing???    OpenVMS 7.1-1H2, BASIC 1.2-000   Joe    ------------------------------  % Date: Thu, 15 Mar 2001 14:09:59 -0500 - From: "Peter Weaver" <peter.weaver@stelco.ca>   Subject: Calculate Easter in DCL4 Message-ID: <Ws8s6.155240$Z2.1957787@nnrp1.uunet.ca>  2 Frank da Cruz published some code the other day in? comp.protocols.kermit.misc on how to calculate Easter Sunday in F Kermit. Since it was all integer math and since I wanted to figure outC what I could on Kermit's S-expresion feature I converted it to DCL. F Frank's original post are in the comments. I made a small modificationE at the end to figure out when my wedding anniversary and Easter meet.    Here it is if anyone wants it.   $!B $! This month's Scientific American (March 2001, p.80) includes an article F $! "Easter as a Quasicrystal" by Ian Stewart, in which the calculation ofF $! the date of Easter is explained (and then graphed and compared to a; $! crystalline lattice).  A ten-step algorithm is given for  calculating the F $! Gregorian date of Easter in any given year that is "easy to program on aB $! computer").  Here's an illustration of how to do it in C-Kermit 7.1,0 $! using its new LISP-like S-expression feature: $! $!   #!/usr/local/bin/kermit +@ $!   dcl \&m[] = Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov DecA $!   if ( not def \%1 || not numeric \%1 ) exit 1 Usage: \%0 year E $!   if ( < \%1 1900 || > \%1 2199 ) exit 1 "\0: 1900 <= Year < 2100"  $! $!   (setq x \%1) 8 $ if f$integer(p1) .eq. 0 then p1 = f$cvtime("",,"year")	 $ x = 'p1  $!= $!   (setq a (mod x 19) b (truncate (/ x 100)) c (mod x 100))  $ a = x - (x / 19) * 19 
 $ b = x / 100  $ c = x - (x / 100) * 100  $!, $!   (setq d (truncate (/ b 4)) e (mod b 4)) $ d = b / 4  $ e = b - (b / 4) * 4  $!. $!   (setq g (truncate (/ (+ (* 8 b) 13) 25))) $ g = ((b * 8) + 13) / 25 4 $!   (setq h (mod (+ (* a 19) b (- d) (- g) 15) 30)) $ h = a * 19 + b - d -g + 15 $ h = h - (h / 30) * 30  $!/ $!   (setq m (truncate (/ (+ a (* h 11)) 319)))  $ m = (h * 11 + a) / 3197 $!   (setq j (truncate (/ c 4)) k (truncate (mod c 4)))  $ j = c / 4  $ k = c - (c / 4) * 4 E $!   (setq l (truncate (mod (+ (* 2 e) (* 2 j) m (- k) (- h) 32) 7))) & $ l = 2 * e + (2 * j) + m - k - h + 32 $ l = l - (l / 7) * 7 0 $!   (setq n (truncate (/ (+ h (- m) l 90) 25))) $ n = (h - m + l + 90) / 25 4 $!   (setq p (truncate (mod (+ h (- m) l n 19) 32))) $ p = h - m + l + n + 19 $ p = p - (p / 32) * 32  $!E $!   echo \fday(\m(x)\flpad(\m(n),2,0)\flpad(\m(p),2,0)) \m(p) \&m[n]  \m(x) 	 $!   exit 1 $ month = f$element(n,"-","- - -March-April-May")  $!> $ compare_date = f$cvtime("''p'-''f$extract(0,3,month)'-''x'")' $ if compare_date .lts. "''f$cvtime()'" 	 $    then  $       string := "was"  $       string2 := "fell" 	 $    else  $       string := "will be"  $       string2 := "falls"
 $    endif $! $ suffix = "th" / $ if p .eq. 2 .or. p .eq. 22 then suffix = "nd" / $ if p .eq. 3 .or. p .eq. 23 then suffix = "rd" > $ if p .eq. 1 .or. p .eq. 21 .or. p .eq. 31 then suffix = "st"   $ write sys$output -7     "''f$fao("Easter Sunday in !UL !AS on the !UL!AS of   !AS.",x,string,p,suffix,month)'" $! $!      Customized for PRW $! $ wedding = "" $! $ if n .eq. 4 .and. x .ge. 2001 	 $    then - $        if p .eq. 14 then wedding = "Monday" - $        if p .eq. 15 then wedding = "Sunday" / $        if p .eq. 16 then wedding = "Saturday" - $        if p .eq. 17 then wedding = "Friday"  $! $        if wedding .nes. "" $            then ' $                anniversary = x - 2000  $                suffix = "th"A $                last_two = anniversary - anniversary / 100 * 100 : $                if last_two .lt. 10 .or. last_two .gt. 20 $                    then F $                        last_digit = anniversary - anniversary / 10 * 10@ $                        if last_digit .eq. 1 then suffix = "st"@ $                        if last_digit .eq. 2 then suffix = "nd"@ $                        if last_digit .eq. 3 then suffix = "rd" $                    endif# $                write sys$output - 1    "''f$fao("Our !UL!AS anniversary !AS on Easter * !AS",anniversary,suffix,string2,wedding)'" $            endif
 $    endif $! $!? $! See the article for an explanation of the algorithm and see:  $!4 $!   http://www.columbia.edu/kermit/ckermit3.html#x9 $!@ $! for an explanation of S-Expressions.  Note that the algorithm requiresC $! all arithmetic to be integer, not floating-point, hence the manyN TRUNCATED $! expressions (since S-Expressions never discard fractional parts). $!F $! In UNIX, clip the program, left-justify at least the first line and changeE $! it to point to your C-Kermit 7.1 binary, save it as "easter", then 	 "chmod +xh" $! easter", and then you can type: $! $!   easter 2001 $!E $! (or any other year between 1900 and 2099) to find out the date forl EasterC $! in that year (years outside that range require adjustment of theo algorithm).rE $! On other platforms, type "take easter 2001" (or other year) at the  Kermit
 $! prompt. $!
 $! Exercises:N $!C $!  . Write a similar program for Rosh Hashanah, Passover, Ramadan,m Tet,F $!    Chinese New Year, or any other holiday based on the Lunar cycle. $!+ $!  . Adapt to span a wider range of years.c $!' $!  . Adapt to support other calendars.d $!
 $! - Frank   ------------------------------  % Date: Thu, 15 Mar 2001 16:50:13 -0800e! From: Shane.F.Smith@Healthnet.coma2 Subject: Compaq earnings low, 5,000 jobs to be cutD Message-ID: <OF800C7CB8.2F722EAD-ON88256A11.00045E3E@foundation.com>  % A friend just pointed this out to me:,  A http://dailynews.yahoo.com/h/ap/20010315/tc/compaq_layoffs_1.htmle  G          HOUSTON (AP) - Compaq Computer Corp. is cutting 5,000 jobs andKU          warned Thursday that first-quarter earnings will fall far short of analysts'           estimates.d  J Hmmmm, PCs not doing so well, huh? What was that VMS revenue figure again?- Maybe this is where that VMS ad came from....    Shane    ------------------------------  % Date: Thu, 15 Mar 2001 22:29:17 +0000 ; From: Malcolm MacArthur <malcolmm@rustic-place.demon.co.uk>g" Subject: Re: DCL minute of the day7 Message-ID: <3AB1423D.6758860@rustic-place.demon.co.uk>t   Didier Morandi wrote:e [...]t >  > Question 35 > how can I have DCL ask "_What: " within a procedureT >   > (the reason for this is below) >  > ISLKP1_dmo> ty set.com > $ set = "set"  > $ if p1 .eqs. "" > $ then( > $    assign/user sys$command sys$input > $    goto GO	 > $ endifh > $ p1 = f$edit(p1,"lowercase")o% > $ if p1 .eqs. "dfe" then p1 = "def"e > $GO:4 > $ set 'p1' 'p2' 'p3' 'p4' 'p5' 'p6' 'p7' 'p8' 'p9' > $ exit   Ack!  ! Why not redefine the SET command?a  D Use the freely available VERB package to get a .CLD file for the SET
 command. ThenlF create a "DFE" alias for "DEFAULT" keyword. Then replace DCLTABLES.EXE (test it first
 of course ;-)C  9 I can't remember the exact syntax - I believe you'd want:R   define verb SET  ...I   keyword DEFAULTM     synonym DFEr     ...   C There is a manual about the CDU (Command Definition Utility) on thek OpenVMS website.  D This is the BEST method, because your modified SET command will work everywhere on ther/ operating system. Even in your startup scripts.<  B This could be a problem when you did OS upgrades, and VMS replacedH DCLTABLES.EXE...so I would still make sure your vital command procedures had the correct spelling...e   HTH, Malcolmn   ------------------------------  % Date: Fri, 16 Mar 2001 01:02:21 -050032 From: rdeininger@mindspring.com (Robert Deininger)" Subject: Re: DCL minute of the dayL Message-ID: <rdeininger-1603010102220001@user-2iveb5k.dialup.mindspring.com>  I In article <3AB1423D.6758860@rustic-place.demon.co.uk>, Malcolm MacArthurn* <malcolmm@rustic-place.demon.co.uk> wrote:     > D > This could be a problem when you did OS upgrades, and VMS replacedJ > DCLTABLES.EXE...so I would still make sure your vital command procedures > had the correct spelling...M  E It's not clear that VMS upgrades replace DCLTABLES.EXE anymore.  I'vewI certainly had my private commands survive upgrades.  A particular commandL= might get clobbered if you pick a name that VMS wants to use.k  G I'm pretty sure the whole file got replaced in the old days, since I'veMB trained myself to keep command files that apply my private commandH modifications to the file.  But I haven't had to use these command files after recent upgrades.   -- B Robert Deininger rdeininger@mindspring.com    ------------------------------   Date: 16 Mar 2001 06:15:27 GMT- From: djweath@attglobal.net (Dave Weatherall) " Subject: Re: DCL minute of the day5 Message-ID: <DTiotGxQ0bj6-pn2-Cdd8CLt6ttTb@localhost>s  E On Wed, 14 Mar 2001 07:51:42, Didier Morandi <Didier.Morandi@gmx.ch> M wrote:  & > Why is Brian's post not posted here? >  > D. >  > Peter Weaver wrote:a > > F > > "Brian Schenkenberger, VAXman-" <system@SendSpamHere.ORG> wrote in6 > > message news:009F8F2C.E25962F4@SendSpamHere.ORG... > .../..  F I never saw it either, It's not as if all Brian's stuff gets filtered.A This is the second case that has been apparent in the last three eB weeks. I wonder what else has gone missing. Where and Why too, of  course.d   Cheers - Dave.   ------------------------------  % Date: Thu, 15 Mar 2001 18:16:41 -0600 + From: Christopher Smith <csmith@amdocs.com>-Y Subject: RE: Dumbing Down VMS with UNIX Elements? (was Re: OpenVMS Educati	onal    Progra L Message-ID: <3B55D7F383B0D31197D9009027541CBF0BDD547F@cmiexch1.cmi.itds.com>   > -----Original Message-----2 > From: Christof Brass [mailto:brass@infopuls.com]  > > The job isn't done shitty, the concept is shit (like UNIX).   > What can you make out of shit?  L Fertilizer -- and some aboriginal peoples make entire villages from it... :)   Regards,   Chris   ! Christopher Smith, Perl Developer  Amdocs - Champaign, IL   /usr/bin/perl -e '? print((~"\x95\xc4\xe3"^"Just Another Perl Hacker.")."\x08!\n");r 'r   ------------------------------  % Date: Thu, 15 Mar 2001 20:15:21 -0600N+ From: Christopher Smith <csmith@amdocs.com>lY Subject: RE: Dumbing Down VMS with UNIX Elements? (was Re: OpenVMS EducationalProgram) ProL Message-ID: <3B55D7F383B0D31197D9009027541CBF0BDD5480@cmiexch1.cmi.itds.com>   > -----Original Message-----2 > From: Christof Brass [mailto:brass@infopuls.com]  < > Yes - the differences between the UNIXes are much smaller > > than between VMS and UNIX. What a surprise! VMS isn't UNIX? < > Why isn't UNIX implementing the Windoze API. There is the  > vast majority of apps!  B Softwindows, WABI, Wine, BOCHS, .... you tell me -- why isn't Unix implementing the windows API?q   Seems to me that it is.e   Regards,   Chriso  ! Christopher Smith, Perl Developery Amdocs - Champaign, IL   /usr/bin/perl -e '? print((~"\x95\xc4\xe3"^"Just Another Perl Hacker.")."\x08!\n");t '2  )   ------------------------------  % Date: Thu, 15 Mar 2001 23:58:30 +0000 ) From: Christof Brass <brass@infopuls.com>oY Subject: Re: Dumbing Down VMS with UNIX Elements? (was Re: OpenVMSEducational     ProgramE, Message-ID: <3AB15726.32CD1225@infopuls.com>   Brian Wheeler wrote: > . > In article <3AAFE6FC.D628A3DE@infopuls.com>,5 >         Christof Brass <brass@infopuls.com> writes:O > > Brian Wheeler wrote: > >>1 > >> In article <3AAE5EA9.311BD545@infopuls.com>,W8 > >>         Christof Brass <brass@infopuls.com> writes: > >> > Brian Wheeler wrote:  > >> >>u1 > >> >> In article <3AAE30C5.20D69B88@bbc.co.uk>, ? > >> >>         Tim Llewellyn <tim.llewellyn@bbc.co.uk> writes:o	 > >> >> >t	 > >> >> >2  > >> >> > Jan Vorbrueggen wrote:	 > >> >> > 6 > >> >> >> Christof Brass <brass@infopuls.com> writes:
 > >> >> >>P > >> >> >> > > BTW, in at least one case the use of Unixy C code has resulted inY > >> >> >> > > relaxation of a not-so-useful restriction in a set of VMS system services.sE > >> >> >> > What was this relaxation? Didn't it break existing code?e
 > >> >> >>S > >> >> >> The output format of $ASCTIM for delta times is defined to be limited toeU > >> >> >> 10000 days minus a tick; the binary format itself, useful for calculations R > >> >> >> of all sorts, is of course not so restricted (the limit is around 32000W > >> >> >> years). However, all other time-related system services were also restrictedtX > >> >> >> to that 10000 day limit. This broke when the audit server used the delta timeS > >> >> >> format to compute a time_t - 10000 days after 1-JAN-1970. So this changefD > >> >> >> actually made existing code work, instead of breaking it.
 > >> >> >> > >> >> >>         Jan	 > >> >> >t6 > >> >> > However, 1-JAN-19970 is a unix thing anyway.	 > >> >> >  > >> >>t > >> >> <sarcasm>uL > >> >>         Yeah, nobody but a unix loser would ever use 1-JAN-1970 as a > >> >> base for delta time! > >> >> </sarcasm> > >> >>o; > >> >> Isn't this whole topic getting a little bit stupid?n > >> >>tT > >> >> Its been a constant Unix vs VMS war which, as we all know, is pointless.  IfR > >> >> Compaq decides to add more unix interfaces on top of VMS, what's the harm?P > >> >> How does adding an interface dumb down the system, or lower its quality? > >> >5 > >> > This depends where it is implemented and what.rl > >> > If it's in the kernel then it makes things more complicated and will reduce quality in many respects. > >> > If the UNIX API requires functionality which is in contradiction to VMS system services the view of the system might be inconsistent. > >>R > >> Granted, but since nobody knows how its going to be implemented, then there'sQ > >> no way one can argue that adding unix apis will bring the quality of the VMSm > >> kernel down.- > >- > > This was exactly my point: before the choir applauses it must be clear how it is accomplished. If it is done the wrong way the outcome will be a disadvantage. > O > So, you're saying that just the possibility of VMS engineering doing a shittyhN > job is enough to rag on the concept as a whole?  Your anti-unix prejudice is > clouding your reasoning.  Z The job isn't done shitty, the concept is shit (like UNIX). What can you make out of shit?  
 > >> >> IfR > >> >> it is buggy, its not unix's fault...but the implementation on VMS.  AddingO > >> >> interfaces to an OS doesn't weaken it...it can only strengthen it.  Did!O > >> >> adding a C compiler to VMS lower the quality of the OS?  Did the TCP/IPu3 > >> >> libaries from Tru64 unix lower the quality?r > >> >j > >> > Libraries and tools are not part of the kernel. I wonder what your experience in OS development is. > >>N > >> Again, we don't know how the Unix API would be added to VMS, so it may beL > >> a tools/libraries issue and not touch the kernel at all.  Ok, that is a, > >> bit unlikely, but still, a possibility. > >  > > See above. And there is still the problem of offering the wrong way (the UNIX) to solve things which will complicate the day to day work of VMS admins, programmers and users. > N > The VMS way isn't going away.  What makes you think that adding UNIX APIs isN > going to make life harder for anyone?  Sure, no choice is easier, but before8 > long there won't be anyone around to not make choices.   It makes things more complicated because there are more ways to do things and you have to think about how other people solved things if you have to work on what they created before they left or with it if you work together./ Why would one chose VMSIX? In other words: this COE stuff wouldn't change anything because the miracle wouldn't happen. There is no advantage in running a UNIX like crap app on VMS.t  S > >> I was arguing from the point that there seems to be an overriding feeling thateB > >> anything related to unix is unclean and must not be near VMS. > >  > > Exactly. Dump UNIX which is pure crap (not only from the architecture point of view but also from point of implemetation quality, user interface aso; only the price might be okay if you get payed for using it). > P > You are a troll, there's no other explanation.  You're using Netscape on LinuxO > to post this tripe, yet apparently unix is crap.  Sure, whatever.  Why aren'tn" > you using Netscape on VMS, then?  Sorry for pointing out again that you have a major lack of understanding and logic. Even if there were no app on VMS which is better or even equal to all UNIX apps this is no technical prove that UNIX were better. Of course it would be strange not to conclude that but first there should be an analysis. In our case this analysis reveales that a company whose name is not spelled out here did active damage to the app base, market reputation, business goals and future perspectives of VMS. This lead to the facte that there isn't any decent browser around. The from UNIX ported Navigator is crap as far as I rate the frequent complains in this NG. From what I read only a masochist could try to use VMS as a browser platform. But anyway I didn't try it so far but soon I will because this beloved Navigator on Linux sucks severly too besides that Linux sucks by itself. So it's obvious that you didn't try to put in a resonable statement because you all knew that. It's obvious that you only tried to accuse me. My conclusiong is that you ran out of arguments (to be precise: if haven't read a technical argument from you so far).e8 I really don't understand what you are doing in this NG.  M > I submit to you that the Unix architecture is actually superior to VMS's iniM > many ways.  The simplistic approach where everything is treated as a streameN > of bytes is a powerful concept.  However, does this mean that Unix is betterO > than VMS as a whole?  Nope, as far as I'm concerned, its a matter of choosingsO > the right tool for the job.  Sometimes Unix is the right tool, sometimes VMS.1L > However, to keep this discussion on track, the number of jobs where VMS is  > suitable is rapidly shrinking.  Are USamerican. Sorry I'm not to insult the majority but it's too funny to have the European prejudice of the USamerican stereotype: simplicity is the solution LOL. Albert Einstein said (of course he said it in German and in fact the English version doesn't make much sense but it's too tempting to quote it here): Make it as simple as possible but not too simple. Why not assembly language? The simpler the tools are the better, eh? I don't believe that you really think what you're writing. I also don't understand why the UNIX bytestream advocates don't see that the work of structuring the date has to be done in the apps which is okay for the semantic part. But for several basic types of data organisation its ideal to have it already at hand in the filesystem.h To say it bluntly: having only bytestreams as common denominator in the filesystem or for program data exchange with pipeing is the opposite of superiority. Again: if you really think what you're writing here why are you here?  R > >> >> There are two major APIs out there:  Unix and Windows.  If compaq wants toV > >> >> expand VMS's marketshare, it is going to have to provide a good implementationQ > >> >> of one of those two...and I don't see the Windows API being chosen.  OncelQ > >> >> developers see VMS as "just another target", they're more likely to start R > >> >> writing pieces of code which take advantage of VMS...compared to now where, > >> >> a port is nearly a complete rewrite. > >> >L > >> > Is it that difficult to understand that we don't need another UNIX??? > >>J > >> That's fine...except that unless VMS conforms to something other thanO > >> itself, vendors are not going to bother porting software to it because its K > >> not important (or: profitable) to justify a complete rewrite to make ad > >> VMS application.. > >h~ > > Why isn't that bad? Does any Micro$oft product conform to anything other than itself? To what does a UNIX version conform? > M > Microsoft can do its own thing because marketshare....marketshare which VMS P > does not have.  Unix conforms to POSIX which is open, and anyone can implement > it...witness Linux.t   But no one did it so far. And POSIX is crap also which even most UNIX people concede. Ever read The Mythical Man Month? POSIX is the camel which should have been a horse designed by a committee.  6 > > Why would it be helpful to have UNIX apps on VMS?? > P > Gee, why would anyone want more apps for their OS?  VMS is fun to look at withK > nothing installed...why, I could just type "dir/full" all day!  There are M > many applications for unix which are quite useful that could be compiled on O > VMS if some basic infrastructure was added...which is what this is all about.   - But why on VMS, why not running them on UNIX?h  O > > If you want to take advantage of VMS you have to write a real VMS app. Whatr0 > > is the point in having another UNIX version? > P > The half-assed (but valiant) ports of unix software to VMS have one overridingF > problem:  they're always behind.  It takes so much effort to port anO > application to VMS that trying to keep it up to date is virtually impossible.  >  > >r[ > >> > Think about Apple. Would Steve Jobs argue that Apple should drop their superior API,y& > > UI and so on for UNIX or Windoze?? > >>P > >> Have you seen Mac OS X?  It is _BSD_ with Apple specific APIs built on top.L > >> It is another unix.  You can run unix apps and mac apps...so if someoneJ > >> want to run an open source database it doesn't take months to make it' > >> compile (let alone work correctly)o > >R > > We all know this. Do you know NeXTSTEP? Would any Mac user accept a UNIX app? I know of a long term Mac user who is afraid of MacOS X because he thinks that he might get in touch with the UNIX underneath. > J > It wouldn't have anything to do with you ragging on how bad Unix is, nowL > would it?  "Mr. Mac user, if you use mac OS X and accidentally open up the@ > command line, your soul is as good as gone!"  Give me a break.   ????-Any arguments available? Should I have written that this very same Mac user is a genious in graphics design and that he specifically said that he doesn't want to cope around with the UNIX command line shell crap. Does this make you feel better? Surely he won't die but he would lose his valuable time.n   > > But anyway: MacOS X adds something on top, doesn't change the kernel. MacOS X adds functionality instead of dublicating it. Having UNIX API on VMS is redundant, superflous and completely unecessary. > O > IN YOUR OPINION.  This is what this boils down to.  I (and many others) woulddP > find this a very nice addition to VMS.  If its unnecessary for you, then don'tK > use it.  Its that simple.  Hell, don't even upgrade.  Stay at the currentrK > version because any new feature must be evil and degrade the quality VMS!   Stupid and missing the point. Exactly the opposite is required. Keep the UNIX crap out of VMS that VMS can be developed in a clean and stable fashion with all power focused on the things that are necessary. And that all quality accustomed users can upgrade without risk. If UNIX crap creeps into the kernel than I can't avoid it if I upgrade. Your proposal is a reasonable as your UNIX defense. What are you doing here?"   > >> >Who would like to have another Windoze instead of a Mac? There are other much more efficient and technically much better solutions to the "VMS problem" (how some of this NG would phrase the lack of a few desktop and a few enterprise apps).C > >>N > >> What is that solution?  Compaq (and Digital's) Marketing always seemed toM > >> be focused on preaching to the choir, but that doesn't help.  OutrageousvK > >> costs for OS licenses doesn't help.  No entry-level-priced machines tod& > >> seed the market isn't helping.... > >u > > I published several techical solutions already and I'm getting tired now - some people will be happy to read that, I know ... ; > > You will find them by yourself they are not farfetched.n > O > I've not read them, and judging by your attitude, it'd be a waste of my time.nI > Quick, plow your head into the sand and maybe VMS will become dominant!A  Poor misunderstanding. As I stated several times: people a different. The vast majority of computer people is not VMS people and never will because it likes the challenge of unimportant things like mastering the UNIX command line shell quirks, keeping the irregular command names and options in mind and others. We simply don't need this (and you). We don't need UNIX on VMS. Why don't you get it??   ------------------------------  % Date: Fri, 16 Mar 2001 01:24:26 +0000a) From: Christof Brass <brass@infopuls.com>rY Subject: Re: Dumbing Down VMS with UNIX Elements? (was Re: OpenVMSEducationalProgram) Prof, Message-ID: <3AB16B4A.2B1C5602@infopuls.com>   Brian Wheeler wrote: > . > In article <3AB00BB0.D182A547@infopuls.com>,5 >         Christof Brass <brass@infopuls.com> writes:rD > > Okay you hit a nerve and that's why I responde to your bullshit. > , > That's odd, I was thinking the same thing. >  > >j > > Brian Wheeler wrote: > >>1 > >> In article <3AAEB3E0.285370BF@infopuls.com>,t8 > >>         Christof Brass <brass@infopuls.com> writes: > >> > Fred Kleinsorge wrote:  > >> >>i: > >> >> Not to speak for the US DOD, OpenVMS, or Compaq... > >> >> O > >> >> DII/COE is an attempt to create a magic bullet that allows code re-use.sM > >> >> This would allow code that was written for mission X to be reused foreT > >> >> mission Y.  This hasn't really happened yet.  None the least of the problemsP > >> >> is figuring out how those developing the code for mission X will get anyR > >> >> benefit (read: profit) by allowing their code to be re-used for some otherF > >> >> mission.  So in current use, you end up with Segments that areP > >> >> architecture-specific (which you have anyway, since binary compatabilityR > >> >> isn't a goal), and which use vendor specific features.  BUT these guys areN > >> >> serious, and eventually there will be a lot of COE software in the DOD > >> >> world. > >> >> R > >> >> The unfortunate aspect of DII/COE is that the design center of those doingR > >> >> the development was Sun/Solaris (being practical, they also had to includeM > >> >> Windows - which has it's own features and non-features).  So what COEaO > >> >> provides is a common look & feel for the application environment, whichrQ > >> >> tends to drill down all the way to the UNIX command line interfaces.  You:R > >> >> also have some "standards" - like POSIX and JAVA - which if people writingG > >> >> new segments adhere to, would allow source level compatability.g > >> >>5R > >> >> VMS COE (the product) is not something that will be generally available toS > >> >> *anyone* off the street.  It is a product that includes the OS - as well ase@ > >> >> the COE environment and other required layered products. > >> >> S > >> >> Long term, the OS capabilities that were required, will show up in the basehP > >> >> OS.  The intent is to make the C interfaces on VMS compatable with POSIXT > >> >> (more likely long term - UNIX98 or LINUX).  This will make porting code fromR > >> >> UNIX/Linux a simpler task - no different than any generic UNIX->UNIX port.Q > >> >> In addition, there will also be a shell environment that would allow UNIX S > >> >> users to get something more familiar than DCL.  Lastly are things like file M > >> >> system modifications that allow UNIX semantics and syntax to be used.l > >> >>wT > >> >> This does not "replace" anything that VMS currently does, or even weaken it.S > >> >> All the capabilities are incremental, or supplemental.  But as always, poor P > >> >> UNIX code will work poorly, and good UNIX code will work well.  I expectS > >> >> that when someone wants the code to "mesh" into a VMS environment, they mayfP > >> >> have work to do beyond a simple port - or they can choose to live with a  > >> >> UNIX fish in a VMS pond. > >> >>  > >> >>  _Fred > >> >>e > >> >o > >> > I'm shocked. It's by far worse than I thought. This could be an important step to the real death of VMS. g > >> > - The business effect is *very* doubtful if these COE additions are only there for DoD projects.e > >>O > >> Indeed, it would be, but if it helps port more software to VMS then it canV$ > >> have positive business effects. > >  > > Wrong. Having another UNIX is no advantage. Everybody who can afford VMS can afford another PC to run Linux, Solaris or whatever. No need for VMS to run UNIX crap SW. > Q > Ok, so you're saying that people should buy solaris or linux machines when they,K > need software not available for VMS.  Fair enough...BUT what happens when.I > all of the software functions that VMS is offering become available for>M > Unix at a lower cost?  Oooops!  No more reason to keep that costly VMS box. G > Shut it down and sell it for scrap.  I like VMS, but the one function.O > it has provided us for the last decade (database serving) has been passed offgM > to a much cheaper/faster unix box.  Its only still running until we get the1M > data migrated out and then its going to be decommissioned.  The unix box we P > put it on is a $1500 Dell PC running Linux & Sybase.  And before you go off onJ > "crap unix", etc, its been running non stop since October 1999.  Longish4 > uptimes are no longer the exclusive domain of VMS. > Q > What function should we used the decommissioned VMS machine for?  Great, stablepG > OS...but without software, what's the point of keeping it plugged in?    Good point but not helpful because as you already pointed out: VMS is more expensive anyway and with steadily increasing power of once weak machines you don't need the more expensive machines now where they have been necessary before.tUNIX is crap from its (missing) architecture a.s.o.. I don't repeat it. It will be crap even if it runs forever. If you only put a DB on it, if you don't have any administration to do on this host, if you don't need any VMSish features like clustering and it's unreavaled power to efficiently manage and provide resources to a lot of users there is really no need for VMS.VWhat most people posting to this topic don't understand is that competing with UNIX using quickshot ported UNIX apps is a deadend. Unwillingly you proved what I tried to explain because with that very same example the apps availability wasn't the reason. Instead it was basically what I said: having an app which makes no use of VMS features.   > >> > - Writing SW for a niche is the stupidest thing you can do. SW should be sold to as many customers as possible because copying SW is almost free. > >>P > >> YEP, WRITING SOFTWARE FOR NICHE MARKETS IS STUPID, WHICH IS WHY VMS IS VERYM > >> NEARLY DEAD.  If the COE initiative allows 'commodity' unix source to beyB > >> compiled on VMS then VMS is that much less of a niche player. > >q > > Idiot. *IDIOT*.> > ) > Is this because I don't agree with you?e  No, because you obviously don't think before you write down an argument. Besides that you don't take into account what is related with the COE stuff you exhange the roles of the Q and the SW shops. The COE stuff is the SW which the Q writes and which is aimed for a niche market. Read Fred Kleinsorges very much telling post about this subject. This initiative will help in putting VMS in even more a niche.k  ] > > VMS is not dead. VMS is artificially made to a niche OS by several other idiots like you.  > P > Like ME?  I think not.  VMS has been artificially made into a niche OS becauseN > many proponents (such as yourself) believe with all your heart that anythingL > not VMS is evil and that it must be crap.  VMS is niche because it doesn't& > implement common standards.  Period.  "You are a good example of denying history. Firstly: there other ideas/OSs worth to look at - surely *not* UNIX; that's what I stated. Secondly: the standards argument is stupid because there are several examples in history that having your own solution is not a problem. Ever heard Windoze?  ; > >Study history! Stay away from VMS, you don't deserve it!r > H > Why don't I deserve it?  Is it because I see the value in other tools?P > I have studied history.  I've been using VMS for years...longer than I've beenP > using unix...however I use unix for nearly all of my tasks now.  Why?  Because > that's where the software is.k  / Point well taken. But not VMS' technical fault.    > > Use UNIX! Use UNIX SW. >  > I do, and so do you.   Unfortunately I have to.   > M > [[off topic:  please wrap your lines at 75 chars, these long lines are verye# >   annoying, not to mention rude]]o  [[off topic also: you are the fourth person complaining about my long lines. I decided not to put in artifical line breaks because this should be the task of the rendering engine (like with HTML). Robert Deininger used to post these long lines also and then I thought through and came to the conclusion that I would adjust my news reader to put in these line breakes. Would you like to adjust your reader the same way?]]   > >> > - Offering bad ways like UNIX shells to accomplish tasks is a safe method to kill the productivity of VMS. Today we know how things are solved properly. Tomorrow we will never know. > >>O > >> Get off it.  UNIX != "bad ways", no matter how many times you keep tellingrQ > >> yourself that.  Its different.  That's all.  VMS has one bad side to it:  no Q > >> matter how well designed it is, it is PROPRIETARY.  Unix, as a whole is not, R > >> which gives it a huge advantage in my book over VMS.  With unix I can move to8 > >> another vendor if there are quality or cost issues. > > N > > You don't understand a clue. This proprietary argument has long disproved. > N > Disproved?  How?  VMS is proprietary.  This is a fact.  When you have sourceJ > code on VMS is cannot be compiled anywhere else without massive changes.J > Properly written unix apps (and there are quite a few) can be recompiledJ > with little or no change from one unix to another.  This is also a fact.% > How can this possibly be disproved?a  This can't. Proprietary is a word to denote undocumented (API wise) or niche products. This isn't a technical argument. And as another poster pointed out VMS was one of the first OSs beeing certified by the Open Group (or something similar). There are several standards even open standards that are severely broken. Your proprietary argument isn't worth anything. Do you understand this? Do we agree?W BTW I don't buy your recompilation argument because there are too many counterexamples.s   > > UNIX and Windoze are another form of beeing proprietary with UNIX having the disadvantage of never beeing the same if you change the vendor. > I > Are they as great as when you move from VMS to anything else?  Not evenrL > close.  The difference between unixes is trivial at best for the end user,4 > and only require a bit of effort for the sysadmin.   Yes - the differences between the UNIXes are much smaller than between VMS and UNIX. What a surprise! VMS isn't UNIX? Why isn't UNIX implementing the Windoze API. There is the vast majority of apps!*The differences for using the UNIX commands are painful (for the end user - what is an end user of an OS anyway?) - especially for two widely used commands: "find" and "ps". Compare the option on Linux and Solaris. Are these the same commands? One might find it doubtful if you look at the options.  M > > If 60% of market were owned by VMS nobody would talk about proprietarity.  > M > Ah, but it is not at 60%.  I doubt even 10%.  It is proprietary, regardlessV > of market size.    You again missed the point. Windoze is also proprietary as is UNIX. What makes you claim that UNIX isn't proprietary? What makes you think that this is an advantage?   = > > Read the UNIX-Haters Handbook! Leave VMS alone! Use UNIX!e > O > Ah, the source of your misguidance.  Maybe this quote from the "Anti-Forward"h% > will shed some light on this issue:eL >         Dennis Ritchie writes "You claim to seek progress, but you succeed >         mainly in whining."mM > I have read it.  Its crap.  It bitches and moans about everything, yet many K > of its complaints are "fixed" in most unixes.  In addition, if one wantedsH > to, it would be easy to write a "VMS-Hater's Handbook".  But its not aL > matter of hating anything:  its about choosing the right tool for the job.= > VMS is rapidly becoming less and less often the right tool.  Beause you made it so?  If you really have read it (not only the Anti-Forward) then you would know that it's not about hating for the sake of hating. It clearly shows what sucks and why. And I read it only last year after my painful experiences with UNIX and I found it very true (nothing about the fixes you mentioned generaly). Would you like to go through the book together with me and check what has been fixed? Has the shell command interpretation concept been fixed? No! Have the command and parameter irregularities been fixed?Q No! Has the 'every file is an unstructured stream of bytes' model been fixed? No!oA Please name the three most important things that have been fixed.i  } > >> > I'm too tired to continue this list. Every educated engineer will understand that this is a major attempt to kill VMS.w > >>R > >> If by "educated" you mean "anything but pure VMS is evil and we don't need toL > >> be interoperable with anyone but ourselves", then I guess you're right. > >m
 > > Silly. > J > Yes, your attitude is silly.  You have yet to show that adding unix apisK > kills VMS.  If they started removing VMS apis, then yes, I could see yourvE > point.  But as it stands, they're adding to, and not removing from..  % Do you know Occam's Razor? See below!n   > >> > My hope and wish: the good VMS engineers stay with the normal version. The COE version will be so crappy and full of bugs that it will never be usable. > >>S > >> This is an interesting statement.  So, anything Unixy is so inheritly unstable M > >> that just by implementing it brings the whole OS down?  You're on crack.eQ > >> If they do a shit-awful job of bringing unix services to vms, then its theiru > >> own damned fault.  Period.  > >l> > Stupid. As explained several times: introducing unecessary complexity, superflous or redundant functions and ways to accomplish tasks which are in contradiction to the desing principles of VMS is a safe way to ruin it all. This has nothing to do with implementation quality.  ' > I'll thank you not to call me stupid.h  ' Please take your time before you write.s  / > It has everything to do with implementation! o? I wrote "implementation quality" not implementation in general.t  N >                                              If the UNIX api is added to VMSO > via a set of libraries, then it is no more redundant or more complex than anys > other RTL!    x I wrote this before. But we don't know yet. So it's far too early to applause and we should know how it is accomplished.  N >             If something in the kernel has to change, then it is an issue ofN > the kernel not being flexible enough to accomplish the task at hand.  If theO > task at hand was for anything non-unix-api related, you'd probably be all fori > a change in the kernel.r  The problem with the implementaion in the kernel is that I expect a few UNIX calls to interfere with VMS calls. If you want to access the same resource with the UNIX API from within one program and with the VMS API from within another program. I expect problems that the semantics of these accesses are different and exclude each other. Which results in a severe implementation design problem.o   >  > >eO > >> Attitudes like this is why VMS is dead.  Its a wonder that TCP/IP was everSK > >> added to VMS with attitudes like this....DECNet is the pure networkinge5 > >> protocol!  LAT is the way to true enlightenment!  > >a> > Do you like democracy?? You know, the kind of organising a state in which the people who have money buy the people who simulate decision processes (politicians) or a dumb majority can decide if one plus one is three or four? Is TCP/IP better because the majority is using it?l > F > Is VMS better because nobody is using it?   Quality is not inversely > proportional to usage.   Argh - yes, but no point!! Did I say or imply this? Did you say that DECnet should have to be replaced by TCP/IP because this is the (majority) standard?o  N > >> WAIT A MINUTE!  If VMS is so great, and unix so bad, why is it that I spy > >> this line in your headers:  > >>< > >> X-Mailer: Mozilla 4.73 [en] (X11; U; Linux 2.0.36 i586) > >>M > >> Huh.  Looks like you're just a troll...or a hypocrite.  Not only are your; > >> using a unix to do your mail, you're doing it on a PC.  > >>	 > >> Ugh. 
 > >> Brian > >n > > As I stated several months ago: I'm in a process to moving back to VMS because I'm tired of the wrong promises of UNIX land. UNIX is crap. Believe me! I have 7 years first hand experience mostly with Solaris and Linux, but some others also. > L > Which promises are those?  Why did you switch from VMS in the first place?   Because I changed my working place. I first used CDC and DEC. Then Apple, then PC, then Mac, then PC, then Solaris, Linux, Solaris. My private SW is running on Linux and/or VMS.r  O > In the 7 years did you attempt to learn why Unix does things the way it does,eP > or did you just sit and complain about how "its no VMS"?  I'm serious, I wouldI > like to know.  I (and many others) made a transition from VMS or PCs orwH > whatever to Unix without the obvious hatred you seem to be displaying.   I'm an intested type of person. I learned UNIX with great enthusiams. The promises were it is elegant, simple and reliable. When using the PC I constantly thought about VMS. With UNIX it was different. Only after the mentioned 7 years and reading the mentioned book I became a UNIX hater.  _ > > I also publicly offered to put money into a fund to get a full VMS Opera port. No response.  > >:U > > Do you know on what platform Navigator has been developed? Could it be UNIX crap?u > > Navigator is crap. I use it to avoid Micro$oft although some people say the IE is the better browser. If there is a decent VMS browser available I'll switch to that.  > K > So, we come back to the core of the problem:  lack of available software.lM > Whether something was built on unix doesn't make it crap, anymore than unixo5 > as a whole is crap.  Why don't you write a browser?t  F Good question. Because I have *other* things to do. Do you understand?   > >a6 > > Do you think that your contribution has any value? > L > Yes, in fact I do.   You know, at first I thought you were just a mindlessG > anti-unix troll.  But then I came to a startling realization:  you're K > just close minded.  Computers are just tools.  You pick the one that doesaL > the job the best *FOR YOU*.  If VMS does the job for you (which apparentlyH > it doesn't, as you're using a unix browser) then that's great.  VMS isH > lacking in several areas for me (and others), which is why I think theO > COE initiative is a good thing.  I think VMS is a great operating system, and M > I'd hate to see it go away.  But, that doesn't mean that I think everything>K > else sucks.  There is room in computing for more than one OS...right toolsQ > for the job.  VMS is no longer the right tool for many applications, and addingX= > unix APIs would be a step towards making it the right tool.   Think about the business case. I'm sure it isn't there. The initiative will bring some UNIX apps but they will run unsatisfying and have a lot of bugs which will show up only on VMS. In the mid-term there will no more usable SW on VMS than without this COE initiative, maybe even fewer because the genuine VMS apps might have been changed to the more portable API and will then mainly be supported on UNIXes. Did you ever consider this effect?n  f UNIX isn't the right tool even if it is sometimes the only tool because it lets too much to wish left.  N > I suppose we should just agree to disagree:  I think adding Unix APIs to VMSN > would be a good thing, and you think it'd be a bad thing.  I suppose there'sG > no further need for discussing this, since neither of us are going to  > convince the other.V >  > Brian>   Probably. But I don't see why you think VMS is great and shouldn't vanish if you even like the few most ugly thinks of UNIX like the shell and the API and the simplistic approach (e.g. files as structureless streams of byte). Why are you here in this NG?   ------------------------------  # Date: Thu, 15 Mar 2001 22:34:22 GMTM2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) Subject: Re: fmsdef.hn0 Message-ID: <Orbs6.44$eE2.2025@news.cpqcorp.net>  o In article <98qsfn$ac2$1@reader1.imaginet.fr>, "Jean-Francois Marchal" <jean-francois.marchal@x9000.fr> writes: . :After having installed FMS v2.4 on VMS 7.2-1,A :I'm discovering that fmsdef.h has been installed in sys$examples. :Is there any reason for this ?w' :Could I safely copy it in sys$library,t' :then use it with #include <fmsdef.h> ? = :or should I take my own copy in my application's directory ?t  ,   That would appear to be a bug in FMS V2.4.  F   I'd tend to use it where it was, but there should be little harm in H   moving it into SYS$EXAMPLES: -- though you would then get to maintain K   it and update it each time FMS is upgraded until (if) it gets replicated  K   there by a potential future FMS kit.  The sneakiest approach would be to 0H   searchlist the compiler's include directive logical name (or /INCLUDE)1   through the SYS$EXAMPLES: directory, of course.   N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Thu, 15 Mar 2001 23:45:52 +0100 > From: "Jean-Francois Marchal" <jean-francois.marchal@x9000.fr> Subject: Re: fmsdef.ht. Message-ID: <98rghr$eed$1@reader1.imaginet.fr>  L "Hoff Hoffman" <hoffman@xdelta.zko.dec.nospam> a crit dans le message news:% Orbs6.44$eE2.2025@news.cpqcorp.net...t > . >   That would appear to be a bug in FMS V2.4. >uI >   moving it into SYS$EXAMPLES: -- though you would then get to maintaineL >   it and update it each time FMS is upgraded until (if) it gets replicated( >   there by a potential future FMS kit.  B May we get a future fms version ? FMS has a null price for licence maintenanceo  
 Jean-Franoist   ------------------------------  # Date: Thu, 15 Mar 2001 18:51:42 GMTt= From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)p3 Subject: Re: Help!  PASCAL is read-only code to me.i0 Message-ID: <009F90DF.87ED9831@SendSpamHere.ORG>  b In article <3AB0C2D1.6E091F0E@hiyall.zko.dec.com>, John Reagan <reagan@hiyall.zko.dec.com> writes:% >Brian Schenkenberger, VAXman- wrote:I >> PH >> The dependancy rules all have SYS$LIBRARY:STARLET.PEN listed and MMK,G >> for some reason, wants to rebuild it.  The default rule for .PAS.PEN  >> is: >> a >> .PAS.PEN.& >> $(PASCAL)$(PENVFLAGS) $(MMS$SOURCE); >> PENVFLAGS   = /ENVIRONMENT=$(MMS$TARGET_NAME).ENV/NOLISTu >> o( >> So that is where the .ENV is derived. >BE >I don't know why MMK would want to rebuild it unless the STARLET.PASuD >file is newer than the .PEN.  Hard to do unless somebody has played; >around on your system and copied STARLET.PAS files around.t  F I just installed Pascal prior to executing MMK to build.  I don't play around on my system. ;)t  I >Pascal (starting with V5.6) has conditional compilation.  See Chapter 11dE >in the Compaq Pascal Reference Manual for the %IF, %ELSE, %ELIF, andh@ >%ENDIF directives.  There is also a complimentary /CONSTANT DCLC >qualifier to you let you pass in constants from the command line.  ? >You'll also find %ARCH_NAME, %SYSTEM_NAME, and %SYSTEM_VERSIONh" >directives in Chapter 11 as well.  F Great!  That should got me over and past these hurdles.  I'll pull up ) the doc CD and read through that chapter.t   Thanks much.   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMe            pO city, n., 1. a place where trees are cut down and streets are named after them.?   ------------------------------  # Date: Thu, 15 Mar 2001 19:16:11 GMTh= From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)o3 Subject: Re: Help!  PASCAL is read-only code to me.f0 Message-ID: <009F90E2.F33EA9E8@SendSpamHere.ORG>  p In article <009F90DF.87ED9831@SendSpamHere.ORG>, system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) writes:c >In article <3AB0C2D1.6E091F0E@hiyall.zko.dec.com>, John Reagan <reagan@hiyall.zko.dec.com> writes:t& >>Brian Schenkenberger, VAXman- wrote: >>> I >>> The dependancy rules all have SYS$LIBRARY:STARLET.PEN listed and MMK,oH >>> for some reason, wants to rebuild it.  The default rule for .PAS.PEN >>> is:u >>>  >>> .PAS.PEN' >>> $(PASCAL)$(PENVFLAGS) $(MMS$SOURCE)s< >>> PENVFLAGS   = /ENVIRONMENT=$(MMS$TARGET_NAME).ENV/NOLIST >>> ) >>> So that is where the .ENV is derived.l >>F >>I don't know why MMK would want to rebuild it unless the STARLET.PASE >>file is newer than the .PEN.  Hard to do unless somebody has playede< >>around on your system and copied STARLET.PAS files around. >aG >I just installed Pascal prior to executing MMK to build.  I don't playu >around on my system. ;) > J >>Pascal (starting with V5.6) has conditional compilation.  See Chapter 11F >>in the Compaq Pascal Reference Manual for the %IF, %ELSE, %ELIF, andA >>%ENDIF directives.  There is also a complimentary /CONSTANT DCL D >>qualifier to you let you pass in constants from the command line. @ >>You'll also find %ARCH_NAME, %SYSTEM_NAME, and %SYSTEM_VERSION# >>directives in Chapter 11 as well.c >mG >Great!  That should got me over and past these hurdles.  I'll pull up  * >the doc CD and read through that chapter. > 
 >Thanks much.n >i >--eP >VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COM >           P >city, n., 1. a place where trees are cut down and streets are named after them.    5 OK.  I shouldn't have expected an example in the doc.d  
 How do I say:r   %IF %ARCH_TYPE = ALPHA    ???e   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMt             O city, n., 1. a place where trees are cut down and streets are named after them.e   ------------------------------  % Date: Thu, 15 Mar 2001 14:22:40 -0500e- From: John Reagan <reagan@hiyall.zko.dec.com>)3 Subject: Re: Help!  PASCAL is read-only code to me.t2 Message-ID: <3AB0D030.342CAD2F@hiyall.zko.dec.com>  $ Brian Schenkenberger, VAXman- wrote: > 7 > OK.  I shouldn't have expected an example in the doc.e >  > How do I say:a >  > %IF %ARCH_TYPE = ALPHA    ???W  F You bumped into a known doc error.  It isn't clear from the manual butH the %ARCH_NAME and %SYSTEM_NAME lexicals return strings.  You would code it as:   %IF %ARCH_TYPE = "Alpha" %THENn   WRITELN('On an Alpha');. %ELSEi   WRITELN('Not on an Alpha');i %ENDIF  C I've noted the error in my manual and will get it fixed on the nexti$ revision.  I'll also add an example.   -- s John Reaganl Compaq Pascal Project Leader   ------------------------------  % Date: Thu, 15 Mar 2001 16:40:49 -0500P2 From: norm lastovica <norman.lastovica@oracle.com>3 Subject: Re: Help!  PASCAL is read-only code to me.e* Message-ID: <3AB136E1.53A83AB4@oracle.com>  B While the Compaq Pascal V5.7-82 release notes don't actually give A your example explicitly, they sure seem to be pretty clear to me:    "1.2.3.4  %ARCH_NAME  G  %ARCH_NAME  returns  either  'VAX'  or  'Alpha'   depending   on   the 8  architecture of the system performing the compilation."  C Based on this (and the examples in the release notes), here's what  7 I came up with to demonstrate how to use %IF and %ARCH:p   program x (input,output);s begina!     %if     %arch_name <> 'Alpha't     %then   writeln('foo');      %else   writeln('bar');e
     %endif        %if     %arch_name = 'Alpha'     %then   writeln('Alpha');r!     %else   writeln('Not Alpha');m
     %endif end.     "Brian Schenkenberger, VAXma7 > OK.  I shouldn't have expected an example in the doc.e >  > How do I say:s >  > %IF %ARCH_TYPE = ALPHA    ???m >  > --Q > VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMm > Q > city, n., 1. a place where trees are cut down and streets are named after them.y   -- t> norman lastovica / oracle rdb engineering / usa / 610.696.4685   ------------------------------    Date: 15 Mar 2001 14:28:59 -05009 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)y. Subject: Re: How does one become a VMS guru ??3 Message-ID: <L4cI64Y6Sxtb@eisner.encompasserve.org>s  a In article <5rdTGoMuJTcv@eisner.encompasserve.org>, young_r@encompasserve.org (Rob Young) writes:tg > In article <jLXr6.59587$lj4.1467062@news6.giganews.com>, "Andy Stoffel" <acs@fcgnetworks.net> writes:n >  >> lB >> Is becoming an "Instant" VMS Guru a "worthy" goal ? Personally,> >> I would like to be considered a RPVA (Reasonably Proficient? >> VMS Acolyte") :-). Guru seems too much a "Unixy" designationbB >> or a title you give to someone selling New Age psuedo-religion. >> o > ; > 	After taking a group's VMS test, I was told I was a "VMS D > 	hotshot."  Not a guru, that is reserved for the gurus apparently.I > 	I guess being a hotshot is okay... sounds like someone wants to punch l# > 	you though , so I'm not so sure.o    9 		Well, Andrew wants to punch you.  Does that count ? :-)>  N ==============================================================================N Great Inventors of our time: Al Gore -> Internet; Sun Microsystems -> ClustersN ==============================================================================   ------------------------------  % Date: Thu, 15 Mar 2001 20:01:45 +0000m  From: steven.reece@quintiles.com. Subject: Re: How does one become a VMS guru ??H Message-ID: <OF6BEC9F69.15CA6388-ON80256A10.006DD5D2@qedi.quintiles.com>  I Andrew probably wants to punch anyone that's not indoctrinated to the Sune Borg, I mean Sun God.  :-)i   Larry Kilgallen wrote/quoted :I >>>>      I guess being a hotshot is okay... sounds like someone wants toa punch % >    you though , so I'm not so sure.i    A           Well, Andrew wants to punch you.  Does that count ? :-)d <<<    ------------------------------  # Date: Fri, 16 Mar 2001 00:09:07 GMTx6 From: "Andy Bustamante" <A_C_Bustamante@earthlink.net>8 Subject: Re: How to measure system load and availability? Message-ID: <DQcs6.8$227.1283@newsread2.prod.itd.earthlink.net>   7 Monitor in batch.  This is an extract from a job I use:l  J $ monitor all_classes  /nodisplay/record=db$oper:monitor.dat /end="+01:00"  : $ monitor system/ave, disk/item=que, fcp, io, file, page -I              /nodisp /input=db$oper:monitor.dat /summ=db$oper:monitor.txt   H from within a batch job will produce a summary.  I run these during peakI periods.  If you have smtp enabled you have the report e-mailed.  See theh/ System Management Utilities Reference manual at E http://www.openvms.compaq.com:8000/index.html for documentation about K monitor.  Monitor in batch rates high on my list of remote managment tools. J Disk space will determine the history you can keep.  Another nifty featureD is replaying monitor logs at "warp speed." to isolate problem times.  B You can also get excellent information from F$getsyi and F$getjpi.L Documentation in the DCL Dictionary.  Again with smpt you can be e-mailed or& paged by the system on defined events.       -- Andy Bustamantew Remove the ASCII 95s to replyVG Jean-Francois Marchal <jean-francois.marchal@x9000.fr> wrote in messageo( news:98qtf0$anp$1@reader1.imaginet.fr... > Bonjour  tous ! >l3 > One of my customers is about to let me manage his 3 > AlphaServer 800 remotely,  but he is asking me tot( > produce load and availability figures. >t3 > What standard, out of the box, tools should I usen > for this ? >p > Cordialement > Jean-Franois Marchaln > X9000 - LYON (FR)y >l >    ------------------------------  % Date: Fri, 16 Mar 2001 00:05:41 -0500o' From: Jim Becker <jbecker@ui.urban.org>e8 Subject: Re: How to measure system load and availability, Message-ID: <3AB19F25.8955E8F8@ui.urban.org>   Jean-Francois Marchal wrote: >  > Bonjour  tous !   Bonsoir!  3 > One of my customers is about to let me manage hisr3 > AlphaServer 800 remotely,  but he is asking me tod( > produce load and availability figures. > 3 > What standard, out of the box, tools should I uset > for this ?  C You've already gotten good advice on using MONITOR. I'll add: Don't > fall into the trap of using % CPU Utilization as the sole load measure.  B Regarding availability (uptime percentage?): OOTB tools don't give? this to you directly. You can collect  the necessary data from: 6 - Shutdowns, deliberate or otherwise, in CLUE$HISTORY ' - Startups via ACCOUNTING /TYPE=SYSINIT    --
 Jim Becker+ The Urban Institute (http://www.urban.org/)b' Encompass (http://www.encompassus.org/) 5 ESILUG (http://eisner.encompasserve.org/lugs/esilug/)    ------------------------------  % Date: Thu, 15 Mar 2001 14:45:14 -0500s2 From: "Jason McCormick" <jason.mccormick@lexi.com>" Subject: Re: HP Printer on VMS 6.23 Message-ID: <3ab11bbe$0$17009@wodc7nh1.news.uu.net>C  : "Phil Howell" <phowell@snowyhydro.com.au> wrote in message, news:SFYr6.3550$zW2.152066@ozemail.com.au...= > Jason McCormick <jason.mccormick@lexi.com> wrote in messager. > news:3aafb655$0$5744@wodc7nh0.news.uu.net...E > >   Hello, I'm trying to create some TLBs for an HP printer and wasu > wondering G > > if anyone could help me out.  We have Pager, which is a typesettingv
 > program.C > > It creates Postscript-formatted documents.  These documents aree
 > legal-papernI > > sized (8.5 X 14).  However they will not print to the tray with legalm > papereL > > in it so I'm trying to write a module to be specified at print time with > the K > > /setup() function.  I have many working ASCII modules I wrote using PCLI > thatD > > print legal and duplex but they don't seem to traslate well into wrapping > PSI > > jobs.  In a nutshell, I guess I'm asking for how to make a legal-sizee > paperCK > > formatted PS job print on legal paper from "Tray 3" and also be able tot > turn0 > > on the duplexer as needed.  Can anyone help? > > E > see http://www.hp.com/cposupport/printers/support_doc/bpl01378.html  >k4 > notably the example on switching printer languages. > your setup module should look something like > (where Ec = <27>)m/ > Ec%-12345X@PJL ENTER LANGUAGE = PCL <CR> <LF>y > Ec&l1H - tray 2t > Ec&l3a - legal, > @PJL ENTER LANGUAGE = POSTSCRIPT <CR> <LF>8 > (the postscript job that follows should start with %!)J > you may also want a "reset module" that sets the printer back to a known > statesF > some printers also have options of pcl/postscript/automatic on their controlu > panelu > Phil  J Thanks Phil.  This turned on the duplexer but doesn't make it print out onG legal sized paper.  I entered exactly what you'd showed me and it wouldeL print out one legal sheet that had the "@PJL ENTER LANGUAGE=POSTCRIPT"  So IK added an EcE.  The results of that are that the document prints duplexed oneH letter and no initial legal sheet.  Any further suggestions?  My current
 config is:  ! Ec%-12345X@PJL ENTER LANGUAGE=PCLt Ec&l1H Ec&l3A EcE  @PJL ENTER LANGUAGE=POSTSCRIPT   ------------------------------  % Date: Thu, 15 Mar 2001 15:16:49 -0500g2 From: "Jason McCormick" <jason.mccormick@lexi.com>" Subject: Re: HP Printer on VMS 6.23 Message-ID: <3ab12328$0$17016@wodc7nh1.news.uu.net>e  L   I'm sorry.  I just realized that the file I was working with had duplexingH commands built into the PostScript.  I tried it with another file and no' legal or duplex.  Can anyone else help?    Thanks,n Jasona  = "Jason McCormick" <jason.mccormick@lexi.com> wrote in messagen- news:3ab11bbe$0$17009@wodc7nh1.news.uu.net...p< > "Phil Howell" <phowell@snowyhydro.com.au> wrote in message. > news:SFYr6.3550$zW2.152066@ozemail.com.au...? > > Jason McCormick <jason.mccormick@lexi.com> wrote in message 0 > > news:3aafb655$0$5744@wodc7nh0.news.uu.net...G > > >   Hello, I'm trying to create some TLBs for an HP printer and was 
 > > wonderingtI > > > if anyone could help me out.  We have Pager, which is a typesettingi > > program.E > > > It creates Postscript-formatted documents.  These documents arer > > legal-paper'K > > > sized (8.5 X 14).  However they will not print to the tray with legalw	 > > papertI > > > in it so I'm trying to write a module to be specified at print timef with > > therI > > > /setup() function.  I have many working ASCII modules I wrote using- PCL- > > thatF > > > print legal and duplex but they don't seem to traslate well into
 > wrapping > > PSK > > > jobs.  In a nutshell, I guess I'm asking for how to make a legal-sizet	 > > papernJ > > > formatted PS job print on legal paper from "Tray 3" and also be able to > > turn2 > > > on the duplexer as needed.  Can anyone help? > > >dG > > see http://www.hp.com/cposupport/printers/support_doc/bpl01378.htmln > >s6 > > notably the example on switching printer languages0 > > your setup module should look something like > > (where Ec = <27>) 1 > > Ec%-12345X@PJL ENTER LANGUAGE = PCL <CR> <LF>p > > Ec&l1H - tray 2S > > Ec&l3a - legal. > > @PJL ENTER LANGUAGE = POSTSCRIPT <CR> <LF>: > > (the postscript job that follows should start with %!)L > > you may also want a "reset module" that sets the printer back to a known	 > > statetH > > some printers also have options of pcl/postscript/automatic on their	 > controlb	 > > panele > > Phil >cL > Thanks Phil.  This turned on the duplexer but doesn't make it print out onI > legal sized paper.  I entered exactly what you'd showed me and it wouldaL > print out one legal sheet that had the "@PJL ENTER LANGUAGE=POSTCRIPT"  So IFJ > added an EcE.  The results of that are that the document prints duplexed onJ > letter and no initial legal sheet.  Any further suggestions?  My current > config is: >d# > Ec%-12345X@PJL ENTER LANGUAGE=PCL- > Ec&l1H > Ec&l3A > EcE1  > @PJL ENTER LANGUAGE=POSTSCRIPT >a >r >o   ------------------------------  % Date: Thu, 15 Mar 2001 22:45:58 -0600g7 From: "David J. Dachtera" <djesys.nospam@earthlink.net>"Y Subject: Re: I seem to be in a different reality today, one where VMS andMarketing go togo- Message-ID: <3AB19A86.D4D0DC4C@earthlink.net>u   LJEB wrote:  > J > Just opened the new ComputerWeekly (15th March), and found an advert for; > Alpha & OpenVMS, not only that it is a WHOLE PAGE advert!g > M > Description: Simple red background, a small image, with the following text:  > N > HOW DO YOU FIND OUT WHICH ALPHASERVER OPERATING SYSTEM IS THE MOST RELIABLE? > ASK A SERACH ENGINE. > B > More specifically, ask America's number one rated search engine. >  > Ask nothernlight.com.- > % > They'll give you an instant answer:  > K > Compaq's OpenVMS, the operating system which had made their search engine0C > home the largest text retrieval database ever created. Of course,.K > nothernlight.com is just one of hundreds of thousands of businesses world 6 > wide now profiting from their investment in OpenVMS. >  > Ask them the same question.d# > They'll give you the same answer.  >  > COMPAQ > Inspiration Technology  6 THIS IS THE BIG ONE, ELIZABETH! I'M COMIN' TO JOIN YA!   -- A David J. DachteraS dba DJE Systemsh http://www.djesys.com/  : Unofficial Affordable OpenVMS Home Page and Message Board: http://www.djesys.com/vms/soho/n  F This *IS* an OpenVMS-related newsgroup. So, a certain bias in postings is to be expected.  @ Feel free to exercise your rights of free speech and expression.  F However, attacks against individual posters, or groups of posters, are strongly discouraged.n   ------------------------------  % Date: Thu, 15 Mar 2001 21:50:00 -0800u! From: Koloth <koloth@tmisnet.com> Y Subject: Re: I seem to be in a different reality today, one where VMS andMarketing go tog + Message-ID: <3AB1A988.3A7CDE5A@tmisnet.com>Y   AAAAAAHWWWWOOOOOOOOO!!!!!S   It's about Freaking time!!!!     YEEEAAAAHHH!   "David J. Dachtera" wrote:  
 > LJEB wrote:e > > L > > Just opened the new ComputerWeekly (15th March), and found an advert for= > > Alpha & OpenVMS, not only that it is a WHOLE PAGE advert!u > > O > > Description: Simple red background, a small image, with the following text:t > >aP > > HOW DO YOU FIND OUT WHICH ALPHASERVER OPERATING SYSTEM IS THE MOST RELIABLE? > > ASK A SERACH ENGINE. > >1D > > More specifically, ask America's number one rated search engine. > >j > > Ask nothernlight.com.e > >l' > > They'll give you an instant answer:E > >SM > > Compaq's OpenVMS, the operating system which had made their search enginedE > > home the largest text retrieval database ever created. Of course,CM > > nothernlight.com is just one of hundreds of thousands of businesses worlds8 > > wide now profiting from their investment in OpenVMS. > >d > > Ask them the same question.T% > > They'll give you the same answer.e > >o
 > > COMPAQ > > Inspiration Technology >o8 > THIS IS THE BIG ONE, ELIZABETH! I'M COMIN' TO JOIN YA! >F > -- > David J. DachteraE > dba DJE SystemsN > http://www.djesys.com/ >O< > Unofficial Affordable OpenVMS Home Page and Message Board:! > http://www.djesys.com/vms/soho/r >iH > This *IS* an OpenVMS-related newsgroup. So, a certain bias in postings > is to be expected. >eB > Feel free to exercise your rights of free speech and expression. >jH > However, attacks against individual posters, or groups of posters, are > strongly discouraged.    ------------------------------  $ Date: Tue, 6 Mar 2001 20:18:55 -07005 From: "cstranslations" <cstranslations@email.msn.com>%8 Subject: Re: In-memory layout of PACKED ARRAY OF BOOLEAN) Message-ID: <epNL1UrpAHA.272@cpmsnbbsa09>y  H I haven't used Pascal in ages (and even when I did it was very limited). Speaking in general terms...  J ...you can start with the reference and user manuals for Pascal. These areJ probably available online at Compaq's website. The BASIC and C manuals areF (haven't checked the online C manuals recently as I have the hard copyJ versions). The language manuals will contain information on the underlying argument passing mechanisms.  J Other than that you can create a full compiler listing. You can also run aJ program that declares the array in question under the debugger (specify noJ optimize) getting its address (then looking at the memory at what ever theI address happens to be). If you start working backwards my guess is you'lldK find an array descriptor (possibly noncontigious) or perhaps one related toaL a bit array. You'll also want to take a look at the OpenVMS Calling StandardB manual depending on what your doing in addition to the Programming Interfaces manual.   Joey    2 "Paul Dias" <paul.dias@bbc.co.uk> wrote in message# news:3AA4AFA1.C0C5366E@bbc.co.uk...d > Hi,m > G > Can anyone describe how OpenVMS Pascal would arrange the following in  memory:p >i" > PACKED ARRAY [1..659] OF BOOLEAN >fJ > The "OpenVMS Programming Interfaces: Calling a System Routine" book says that aL > PACKED ARRAY [1..64] OF BOOLEAN is represented as a mask_quadword, but how areE > larger arrays handled? >p	 > Cheers,D >  > Paul.t   ------------------------------  % Date: Thu, 15 Mar 2001 15:50:13 -0600l1 From: "Dave Gudewicz" <david.gudewicz@abbott.com>q- Subject: Re: Low cost server - Sun, of course-8 Message-ID: <98rdfi$2mt$1@fizban.fizban.pprd.abbott.com>  E Yep.  And didn't Scott McNealy wear a DEC badge at one point in time?iI Wasn't he involved back in the daze of the 3 amigos; DecMate, Rainbow andl8 Pro 350/380?  Or might I be thinkin' about someone else.   Dave...e  ? "Terry C. Shannon" <terryshannon@mediaone.net> wrote in message26 news:pwOr6.3789$mH4.1592917@typhoon.ne.mediaone.net... > ; > "Paul Repacholi" <prep@prep.synonet.com> wrote in message ) > news:87elw0s4dw.fsf@prep.synonet.com...i8 > > mathog@seqaxp.bio.caltech.edu (David Mathog) writes: > >eI > > > Compaq is never going to give us the long desired "PC VMS", or evenuJ > > > "PC Alpha", but the days of "PC Solaris" have clearly arrived.  WithI > > > this system and the new Sun Blade 100 the price premium for SolarislI > > > hardware is now only a few hundred dollars more than for a PC.  And I > > > they'll give you Solaris, whereas you have to buy W2K, so the pricexB > > > differential to a running system is really very small - it'sH > > > essentially a dead heat. Ok, the Sun won't be as fast as the IntelI > > > solution of the same cost, but the point is that you can now deployeH > > > a new Sun for the same cost as a new PC.  Would that the same were > > > true for Alpha.i >rI > If you hold your breath waiting for that to happen with Alpha, you will2H > become as cyanotic as the Blue Men spokes-mimes on Intel's Unobtainium siteL > at www.itanium.com (Said mimes are blue because they've been holding their$ > breath waiting for IA-64 to ship). > >o9 > > I think Sun has pulled a brilliant move with the 100.  > >YH > > It get into the PC price bracket, and in many companies is under the. > > purchase approval ceiling. Just go buy it. > >>? > > It is no bleeding edge speedy, but should be 'fast enough'.a > >  >IJ > Indeed. Recall when Sun brought out the original $2K Darwin workstation.K > They met the peecee price point then, and continue to do so. And kudos tooH > Scott McNealy for strong-arming the Unix ISVs into reducing their apps( > prices to ensure Windoze price parity. >hL > Whether you like, dislike, or could care less one way or t'other about SunL > Microsystems, you have to admit that they are not stupid or short-sighted./ > The firm's success is by no means accidental.u >  >. >o   ------------------------------  % Date: Thu, 15 Mar 2001 14:01:51 -0500=- From: "Richard D. Piccard" <piccard@ohio.edu>o# Subject: Re: Merging multiple disksr( Message-ID: <3AB11195.5242EA0E@ohio.edu>  J In some situations I have dealt with rooted logicals by such techniques as  H $       define ohiou 'f$trnlnm("SYS$SYSDEVICE")'[ou.]/transl=(conc,term)  M I don't recall using it with the already-defined logical being a search-list.E  #                                 RDP.     Wayne Sewell wrote:p  n > In article <3AAFAF15.76DF87DA@interbulletin.com>, David J. Dachtera <donotreply@interbulletin.bogus> writes:/ > > steven.reece@quintiles.com wrote in articleoB > > <OF9D6758E8.ECBF7AE8-ON80256A0F.00465426@qedi.quintiles.com> : > >>M > >>Depends what the disks are, how big they are, whether they include systemn< > >>disks or just user data with no alias entries and so on. > >>G > >>If they just contain user files I might be tempted to create rootedeH > >>directory logical names so that if one has DISK$USER1 and DISK$USER2" > >>merging onto DKA200 you'd have@ > >>$ DEFINE/SYSTEM/TRANSLATION_ATTRIBUTE=(CONCEALED) DISK$USER1 > >>DKA200:[USER1.]o@ > >>$ DEFINE/SYSTEM/TRANSLATION_ATTRIBUTE=(CONCEALED) DISK$USER2 > >>DKA200:[USER2.]a > >>M > >>Then you could refer to DISK$USER1:[JOHN.DOE] or DISK$USER2:[JANE.DOE] as.N > >>you would have done previously.  You could also move the DISK$USER1 "disk"O > >>and the DISK$USER2 "disk" to other devices at various times, depending upontI > >>usage, free disk space, speed of access, growth rate etc etc etc etc.p > >>M > >>For a user disk with no aliased file entries I'd transfer the files using N > >>BACKUP, either doing straight disk to disk backups from [000000...] or, ifI > >>this was not feasible, do a backup to a saveset and then transfer ther > >>saveset. > >>2 > >>There are, obviously, other options available.
 > >>Steve. > >> > >>Mark Hemker wrote:K > >>>>>I am going to be migrating to some new disks this weekend and I needeE > >>to merge the contents from multiple disks onto a single new disk.wH > >>Does anyone have any suggestions on the best/fastest way to do this?C > >>Some of the disks have the same directory names and file names.f > > / > > One caveat I would add to Steve's comments:a > >tS > > Some application code (usually DCL proc.'s) will not tolerate or permit the use I > > of rooted logicals. Speaking from experience on a customer site here.s > > N > > Go ahead and try it; but, have a plan B in case that fails. It SHOULD workM > > (famous last words) - I've done it before. Application coders and vendors:> > > have a knack for making useful features unusable, however. >:K > One common mode of failure is to attempt combining rooted logicals.  ManyuD > vendor software packages create their own rooted logicals, such as >oG > $ define/system/nolog/exec/trans=(concealed,terminal) goober_common -r. >                 the_disk:[goober_53.common.] >cC > and then other logicals can be based on the common logical, as in, >o0 > $ define/system goober_exe goober_common:[exe] > $ run goober_exe:yadda_dadda > N > If the_disk is a rooted logical, you get a failure on the define because youO > can't use a rooted logical in the definition of another rooted logical.  As a F > developer of third-party software, I've been bit by this one myself. >nM > However, I found that it is rather easy to get around this problem with theY > no_conceal option of f$parse.a >n > For example: >. > $!* > $ doit: subroutine   !  p1 = target disk > $!
 > $ disk = p1eJ > $ act_disk = f$parse(disk,"[yadda]",,"device", "no_conceal,syntax_only")L > $ act_dir = f$parse(disk,"[yadda]",,"directory", "no_conceal,syntax_only") > $ sho sym act_disk! > $ act_dir = act_dir - "[YADDA]"- > $ sho sym act_dir@ > $ if act_dir .eqs. ""  > $       then) > $               prefix = act_disk + "["  > $       else3 > $               prefix = act_disk + act_dir - "]"n > $       endifoH > $ define/process/nolog/exec/trans=(concealed,terminal) goober_common -, >                 'prefix'goober_53.common.] > $ sho log/ful goober_commonn > $ endsubroutineo >  > Here's an example run: >w > $! > $!? > $! First set up the concealed logical pseudodisk for the teste > $!F > $ define/proc/exec/trans=conc bogus_disk curly_disk2:[junk.garbage.]D > %DCL-I-SUPERSEDE, previous value of BOGUS_DISK has been superseded > $ sho log/ful bogus_diskX >    "BOGUS_DISK" [exec] = "CURLY_DISK2:[JUNK.GARBAGE.]" [concealed] (LNM$PROCESS_TABLE) > $! > $!- > $ call doit curly_disk3         ! real diskl* > $ doit: subroutine   !  p1 = target disk > $!
 > $ disk = p1 J > $ act_disk = f$parse(disk,"[yadda]",,"device", "no_conceal,syntax_only")L > $ act_dir = f$parse(disk,"[yadda]",,"directory", "no_conceal,syntax_only") > $ sho sym act_disk >   ACT_DISK = "CURLY$DKB500:"! > $ act_dir = act_dir - "[YADDA]". > $ sho sym act_dirr >   ACT_DIR = "" > $ if act_dir .eqs. ""  > $       then) > $               prefix = act_disk + "["d > $       else > $       endif H > $ define/process/nolog/exec/trans=(concealed,terminal) goober_common -2 >                 CURLY$DKB500:[goober_53.common.] > $ sho log/ful goober_commonti >    "GOOBER_COMMON" [exec] = "CURLY$DKB500:[GOOBER_53.COMMON.]" [concealed,terminal] (LNM$PROCESS_TABLE)i > $ endsubroutinen> > $ call doit bogus_disk          ! rooted logical pseudo disk* > $ doit: subroutine   !  p1 = target disk > $!
 > $ disk = p1hJ > $ act_disk = f$parse(disk,"[yadda]",,"device", "no_conceal,syntax_only")L > $ act_dir = f$parse(disk,"[yadda]",,"directory", "no_conceal,syntax_only") > $ sho sym act_disk >   ACT_DISK = "CURLY$DKB100:"! > $ act_dir = act_dir - "[YADDA]"h > $ sho sym act_dire >   ACT_DIR = "[JUNK.GARBAGE.]"d > $ if act_dir .eqs. ""  > $       else3 > $               prefix = act_disk + act_dir - "]"T > $       endif H > $ define/process/nolog/exec/trans=(concealed,terminal) goober_common -? >                 CURLY$DKB100:[JUNK.GARBAGE.goober_53.common.]s > $ sho log/ful goober_commonlv >    "GOOBER_COMMON" [exec] = "CURLY$DKB100:[JUNK.GARBAGE.GOOBER_53.COMMON.]" [concealed,terminal] (LNM$PROCESS_TABLE) > $ endsubroutinee > $ exit >oO > An even easier way around the problem is to base the logical on the directoryoC > the command procedure is executing out of, again with no_conceal.n >aQ > For instance, if command procedures for the product (including the one definingb> > the root logical) reside in goober_common:[cmd_procs], then: >y) > $ cmdproc = f$environment ("procedure")3= > $ cmd_procs = f$parse(cmdproc,,,"device", "no_conceal") + -c> >                 f$parse(cmdproc,,,"directory", "no_conceal")  > $ cmd_procs = cmd_procs - "]["- > $ common = cmd_procs - ".CMD_PROCS]" + ".]"n9 > $ define/process/exec/trans=conc goober_common 'common' = > $ define/process goober_cmd_procs goober_common:[cmd_procs]  >tO > If you do it this way, it doesn't matter if the device is a rooted logical or M > not.  Of course, since you have bypassed the rooted logical, you may have a M > problem with directory levels if the rooted logical contains a lot of them.tK > Since your logical is based on the physical disk and directories, it willuN > contain all of the directory levels that had been in the rooted logical.  IfL > your product also has a lot of directory levels, you may exceed the limit.Q > Getting around the directory level limit is one of the reasons for using rootedp > logicals in the first place. >n > Wayne  > --Q > ===============================================================================tO > Wayne Sewell, Tachyon Software Consulting  (281)812-0738  wayne@tachysoft.xxx : > http://www.tachysoft.xxx/www/tachyon.html and wayne.htmlM > change .xxx to .com in addresses above, assuming you are not a spambot  :-)TQ > ===============================================================================TQ > Dean Wormer to Flounder: "Fat, drunk, and stupid is no way to go through life."r   --B ==================================================================B Dick Piccard                           Academic Technology ManagerB piccard@ohio.edu                                 Computer ServicesB http://oak.cats.ohiou.edu/~piccard/                Ohio University   ------------------------------  % Date: Thu, 15 Mar 2001 15:54:36 -0500a+ From: "O'Connor, Marty" <MOConnor@DVFS.COM>t# Subject: RE: Merging multiple disksPF Message-ID: <85C741006DA1D0119CE00000F8752CE304078974@msexc1.dvfs.com>  B In the past I have had to merge disks. On the target disk I set upJ directories off the root of the new disk named for each disks to be copiedJ onto that disk. I then created rooted logicals to point to those disks andE copied the disks to the rooted directories. Then rooted logicals were K created using the same names as the logical names for the old disks and you:L don't have to worry about changing device references. On very rare instancesL some procedure doesn't like the rooted logical instead of a physical device.   Marty O'Connor    I > >>>I am going to be migrating to some new disks this weekend and I needhC > to merge the contents from multiple disks onto a single new disk.u  K Don't forget that if you have users whose SYS$LOGIN is on some of the disksLJ that are to be removed, you may want to update the user authorisation file toG change their /DEVICE, or at the very least, make sure that you create asH logical name that redirects accesses to the old drives to the new drive.  K Also, you'll have to be careful about the queue manager. Entries are storede byJ file-id, so when you move the files to the new disk, the batch queues willK still point to the old files on the old drives. One possible solution is toiJ keep the old drives on-line to let a cycle of batch jobs run and when they@ resubmit themselves, they should be submitting the new location.  I You also have to be quite careful about any software package that has its  own I configuration files or startup that defines logicals to ensure that thosea reflect the new locations.   ------------------------------  % Date: Thu, 15 Mar 2001 18:39:02 -0600t7 From: "David J. Dachtera" <djesys.nospam@earthlink.net>p# Subject: Re: Merging multiple disksr- Message-ID: <3AB160A6.29CC3BDE@earthlink.net>,   Wayne Sewell wrote:d > [snip]M > However, I found that it is rather easy to get around this problem with the- > no_conceal option of f$parse.2 > [snip]  E In our case, that was the problem - F$PARSE( ..., "NO_CONCEAL" ), butdD the surrounding code mishandled the resulting strings. The resulting filespecs were malformed.t   --   David J. DachteraO dba DJE Systemsa http://www.djesys.com/  : Unofficial Affordable OpenVMS Home Page and Message Board: http://www.djesys.com/vms/soho/   F This *IS* an OpenVMS-related newsgroup. So, a certain bias in postings is to be expected.  @ Feel free to exercise your rights of free speech and expression.  F However, attacks against individual posters, or groups of posters, are strongly discouraged.T   ------------------------------  % Date: Thu, 15 Mar 2001 14:09:21 -0500s% From: "Brian Tillman" <tillman_brian>a* Subject: Re: Obtaining 7.2-1 and TCPIP 5.1$ Message-ID: <3ab1139f$1@news.si.com>  I >Actually I was looking at MX last night, and it looks like it would costd $500.h  I MX V4.2 is still free and you can do a _few_ anti-spam tricks with it and./ whiteice.c that comes in the CONTRIB directory.  --A Brian Tillman                   Internet: tillman_brian at si.com A Smiths Aerospace                          tillman at swdev.si.comT= 3290 Patterson Ave. SE, MS      Addresses modified to prevent < Grand Rapids, MI 49512-1991     SPAM.  Replace "at" with "@"8        This opinion doesn't represent that of my company   ------------------------------  # Date: Thu, 15 Mar 2001 20:08:22 GMTp- From: goathunter@goatley.com (Hunter Goatley)w* Subject: Re: Obtaining 7.2-1 and TCPIP 5.10 Message-ID: <3ab1031c.70439716@swen.process.com>  N On Thu, 15 Mar 2001 04:54:46 GMT, "Zane H. Healy" <healyzh@shell1.aracnet.com> wrote:  M >Well, there is also the fact they've finally added XDMCP to TCPIP.  ActuallyrK >I was looking at MX last night, and it looks like it would cost $500.  I'duI >be better off putting that $500 towards the full Condist and OS set.  ItmJ >would be kind of nice if MX or PMDF was available with a Hobbyist license >:^) >1J Unfortunately, our contract with Sun didn't take the Hobbyist program intoK account, and we'd have to pay them royalties for every copy distributed viatK the Hobbyist program.   Paying Sun for copies we give away for free doesn'tl+ make for good financial business sense. ;-)a  L We are looking into it, but I doubt that Sun will be interested in expendingL legal time to change the agreement we have with them so we can give it away.   Hunter ------9 Hunter Goatley, Process Software, http://www.process.com/l9 goathunter@goatley.com     http://www.goatley.com/hunter/    ------------------------------  % Date: Thu, 15 Mar 2001 17:16:48 -0500t  From: Ben Sego <bsego@clark.net>* Subject: Re: Obtaining 7.2-1 and TCPIP 5.1) Message-ID: <3AB13F4F.5478B776@clark.net>i   Hunter Goatley wrote:o  L > Unfortunately, our contract with Sun didn't take the Hobbyist program intoM > account, and we'd have to pay them royalties for every copy distributed viaoM > the Hobbyist program.   Paying Sun for copies we give away for free doesn't - > make for good financial business sense. ;-)A >A  O Gosh, Hunter, I guess you guys never really got the new dot com fever, then diddR you?  'Cause, paying to give things away is exactly what lots of those fine failed (non)businesses did.  N > We are looking into it, but I doubt that Sun will be interested in expendingN > legal time to change the agreement we have with them so we can give it away.    Well, there's always the hope...   Ben Sego   ------------------------------    Date: 15 Mar 2001 15:10:59 -07001 From: nothome@spammers.are.scum (Malcolm Dunnett)j* Subject: Re: Obtaining 7.2-1 and TCPIP 5.1, Message-ID: <4lEMsdSPqJNa@malvm1.mala.bc.ca>  1 In article <3ab1031c.70439716@swen.process.com>, E3     goathunter@goatley.com (Hunter Goatley) writes:c  N >>Well, there is also the fact they've finally added XDMCP to TCPIP.  ActuallyL >>I was looking at MX last night, and it looks like it would cost $500.  I'dJ >>be better off putting that $500 towards the full Condist and OS set.  ItK >>would be kind of nice if MX or PMDF was available with a Hobbyist license  >>:^)8 >>L > Unfortunately, our contract with Sun didn't take the Hobbyist program intoM > account, and we'd have to pay them royalties for every copy distributed viaM > the Hobbyist program.l  3     You have to pay Sun royalties for copies of MX?.     ;-)    ------------------------------  % Date: Thu, 15 Mar 2001 23:11:37 +0000 ) From: Christof Brass <brass@infopuls.com>a( Subject: Re: OpenVMS Educational Program, Message-ID: <3AB14C29.76E1EFBF@infopuls.com>  " Shane.F.Smith@Healthnet.com wrote: > L > Turn it around, Christof. Why don't people use VMS, if Unix sucks so much?  Very good question. My answer: people are different. The vast majority of UNIX advocates is selfsatisfying with mastering unecessary problems like obscure shell substitution rules or very tricky pipes including sed unreadable parameters a.s.o.. By construction in one case or by history of evolution in the other case VMS is designed UNIX is irregular crap. I'm sure only a small minority of UNIX addicts can be relieved to understand the design superiority of VMS. I'm not talking about market share and appsac availability. I'm only talking about design, architecture and implementation technique and quality.p  M > Because that's where the apps run. If you make those same apps available oniJ > VMS then VMS becomes a viable option in many areas where it is currentlyL > not viable now. You try selling a manager VMS for a web server that serves$ > out Real Networks streaming media: > L > Manager: I want to provide streaming Real Networks media to our customers.- > Techie: Can I interest you in a VMS system?e/ > Manager: Does Real Networks server run on it?o- > Techie: Well, no, but it's really robust... L > Manager looks at Techie as if he's grown a second head, and makes a mental4 > note to surf Dice.com looking for a new techie....  Here we are: the business case - I waited ages for someone to present one as to justify this COE stuff besides DoD. But I don't buy it. Why using VMS for an app running alreay on UNIX? A UNIX like app also. A UNIX quality app also. Why bother buying a VMS box? Does really anyone believe that within the next five or ten years there will be in any way close as much apps be around on VMS as on UNIX? As for the business case: the manager selects a UNIX box for sure if sHe is thinking about app availability in any respect. Even if the currently needed apps are there. I personally doubt that a UNIX app will run flawless and efficient on VMS. Can we expect that these quickshot ports of UNIX apps will run in a cluster? Will they use RMS structered files?dAgain back to the business case: What is the advantage in buying a VMS box? I doubt that there is any if you plan running UNIX apps. Will the quality of the system (OS and app) be really better than on the original platform. I seriously doubt that. Given the low quality of the UNIX apps will the hidden bugs not pop up all around? Will that a combination, VMS with a quickshot UNIX app on top, really be more stable than on the original platform? I seriously doubt that.2{ A shop that could afford VMS could afford top liga Intel compatible crap with Linux or Solaris. Where is the business case?a   > Shanev > ? > Christof Brass <brass@infopuls.com> on 03/14/2001 02:27:28 PMU >  > To:   Info-VAX@Mvb.Saic.Comi > cc:e > + > Subject:  Re: OpenVMS Educational Programp > $ > Shane.F.Smith@Healthnet.com wrote: > > L > > From a technical point of view, as you say, we have no concrete details.L > > Let's see what VMS Engineering deliver first. From a real world point ofK > > view, COE certification is intended to garner a larger userbase for VMS  > inL > > the military, guarantee its survival for many years to come, and lead toI > > more apps becoming available. These are all extremely important rightm > now.J > > Keeping VMS "pure" (for want of a better word) is NOT going to keep itK > > alive. You only have to look at Windows to see how much weight an idealsH > > design has in the current market - it's crap but it rules the world. > >iM > > This is a fight for survival. We can't afford the luxury of anyone's ideao@ > > of a perfect environment. Not yours, not mine, not anyone's. > M > It has been written that these COE amendments would for a certain period (IeH > assume at least 5 to 10 years) not be available to the public. So yourI > argument wouldn't fit. But my fear also wouldn't be appropriate. But mynK > point is that technically this is something very risky and done the wrongoM > way will kill VMS technically. Why would one use VMS with UNIX apps insteadsH > of using UNIX? And putting engineering effort in that COE niche marketI > project will reduce the engineering power spent for the public version.m > Insane, isn't it?  > J > I'm sure that even the COE UNIX API would already be there this wouldn'tK > help VMS. Instead it might well kill VMS also from the marketing point ofxJ > view because the real VMS apps will vanish and there will no apps remainK > using the real power of VMS. And there will be a lot of apps spoiling VMSh2 > by not using its features like structured files. > 	 > > Shaned > >yA > > Christof Brass <brass@infopuls.com> on 03/13/2001 03:20:40 PMd > >  > > To:   Info-VAX@Mvb.Saic.Comh > > cc:m > > - > > Subject:  Re: OpenVMS Educational Programo > >m% > > steven.reece@quintiles.com wrote:! > > >tK > > > This is getting to be something of a flame war come religious war.  IiB > > > didn't expect the Spanish Inquisition and all that.......... > > >cG > > > Complexity of code is something that is almost part and parcel of_	 > presentDM > > > product offerings.  This is because people are expecting more functions  > > andrF > > > more displays and in the case of some environments cruddy little > > paperclipsJ > > > that come up when you least expect/need/want them and flying bits of	 > > paper_I > > > that take up CPU time when you should actually be copying the file.b
 > > > Grrrrr!  > > >oI > > > I would expect that stuff like cluster transition capabilities, SCS L > > > connections over switched LANs, memory channel and other interconnects > isM > > > pretty complex to get right, as are a fair number of other parts of theU > > VMSsL > > > operating system (not just the kernel).  Would you say that VMS is low > onK > > > quality?  No.  Would you say that these bells and whistles should not  > be& > > > added in (along with APMP)?  No. > > >eE > > > What you're getting with the work that is being carried out aresL > > > capabilities on VMS to present Unix-like environment features.  Or, as > > FredL > > > Kleinsorge (not speaking for Compaq or OpenVMS or the US DoD) put it : > > >KI > > > >>>So what COE provides is a common look & feel for the applicationIL > > > environment, which tends to drill down all the way to the UNIX command > > lineI > > > interfaces.  You also have some "standards" - like POSIX and JAVA -= > whichdH > > > if people writing new segments adhere to, would allow source level > > > compatability.<<<b > > >o/ > > > then later in the same message he wrote :G > > >3J > > > >>>This does not "replace" anything that VMS currently does, or even
 > > weakenH > > > it. All the capabilities are incremental, or supplemental.  But as > > always, L > > > poor UNIX code will work poorly, and good UNIX code will work well.  IB > > > expect that when someone wants the code to "mesh" into a VMS > environment,K > > > they may have work to do beyond a simple port - or they can choose to  > > live( > > > with a UNIX fish in a VMS pond.<<< > > >C > > > Christof Brass wrote: D > > > >>>Quality is dependent on time and complexity i.e. effort and > > skilfulness.L > > > Having UNIX on VMS raises complexity and hence imposes more difficulty > in > > > achieving quality. > > >U% > > > Why do we want another UNIX?<<<o > > D > > While I'm not satified with these general statements of the well > estimatedAM > > Fred Kleinsorge and while I'm still expecting hard problems with this COEoF > > stuff I have to admit that I can't prove it with concrete details.M > > Unnecessary complexity reduces quality with given time frames/milestones.aK > > The VMS feature rich environment doesn't need any adaption to any thirdcL > > party (DoD) rules - especially if these rules are strongly influenced by > > UNIX crap.M > > The complexity of VMS is needed for delivering its features and thereforeoE > > the tradeoff has been made to implement this quality features anddK > > sacrificing simplicity. UNIX is the opposite. Quality is lowered if youp > add"C > > something you don't need or which destroys the design or offersnL > > possiblities you better don't want to have because they don't fit in the$ > > way things are best done so far. > >rJ > > Why would you like any of this COE stupidity from a technical point ofL > > view? And keep in mind that other things wouldn't be implemented because > of# > > the effort this COE task needs.t   ------------------------------  % Date: Thu, 15 Mar 2001 15:55:30 -0800l! From: Shane.F.Smith@Healthnet.coml( Subject: Re: OpenVMS Educational ProgramD Message-ID: <OF5E33DB02.C9EAA958-ON88256A10.0082DE99@foundation.com>  K I'm tempted to say "hook, line, sinker and copy of angling times...", but If$ didn't really plan it that well. ;-)  H We had a requirement for a web server. VMS came out of the analysis withJ the highest score of the candidate platforms, and would probably have beenJ chosen except for one thing. Some of the necessary apps weren't available.G Showstopper. All those lovely features and good design you like so muchsH couldn't save it, because of a missing piece. Needless to say, we're not using VMS for that project.=  K Now, if we'd had a port of those apps, and that list /was/ short, the story I would have been different. If you think this is an isolated incident, I'm= afraid you're sadly mistaken.    Shane           = Christof Brass <brass@infopuls.com> on 03/15/2001 03:11:37 PM    To:   Info-VAX@Mvb.Saic.Comv cc:   ) Subject:  Re: OpenVMS Educational Program     " Shane.F.Smith@Healthnet.com wrote: >-F > Turn it around, Christof. Why don't people use VMS, if Unix sucks so much?M  I Very good question. My answer: people are different. The vast majority of0H UNIX advocates is selfsatisfying with mastering unecessary problems likeC obscure shell substitution rules or very tricky pipes including sedoJ unreadable parameters a.s.o.. By construction in one case or by history ofG evolution in the other case VMS is designed UNIX is irregular crap. I'm H sure only a small minority of UNIX addicts can be relieved to understandJ the design superiority of VMS. I'm not talking about market share and apps= availability. I'm only talking about design, architecture andp% implementation technique and quality.l  J > Because that's where the apps run. If you make those same apps available onJ > VMS then VMS becomes a viable option in many areas where it is currentlyE > not viable now. You try selling a manager VMS for a web server thatv serves$ > out Real Networks streaming media: > A > Manager: I want to provide streaming Real Networks media to ouro
 customers.- > Techie: Can I interest you in a VMS system?e/ > Manager: Does Real Networks server run on it?e- > Techie: Well, no, but it's really robust....E > Manager looks at Techie as if he's grown a second head, and makes ae mental4 > note to surf Dice.com looking for a new techie....  I Here we are: the business case - I waited ages for someone to present onenK as to justify this COE stuff besides DoD. But I don't buy it. Why using VMScK for an app running alreay on UNIX? A UNIX like app also. A UNIX quality appnI also. Why bother buying a VMS box? Does really anyone believe that withineI the next five or ten years there will be in any way close as much apps betI around on VMS as on UNIX? As for the business case: the manager selects a-> UNIX box for sure if sHe is thinking about app availability inF any respect. Even if the currently needed apps are there. I personallyK doubt that a UNIX app will run flawless and efficient on VMS. Can we expecteH that these quickshot ports of UNIX apps will run in a cluster? Will they use RMS structered files?nK Again back to the business case: What is the advantage in buying a VMS box?AI I doubt that there is any if you plan running UNIX apps. Will the qualitynJ of the system (OS and app) be really better than on the original platform.G I seriously doubt that. Given the low quality of the UNIX apps will thepF hidden bugs not pop up all around? Will that a combination, VMS with aE quickshot UNIX app on top, really be more stable than on the original ! platform? I seriously doubt that.eH A shop that could afford VMS could afford top liga Intel compatible crap2 with Linux or Solaris. Where is the business case?   > Shanes >r? > Christof Brass <brass@infopuls.com> on 03/14/2001 02:27:28 PM1 >: > To:   Info-VAX@Mvb.Saic.Comm > cc:l >n+ > Subject:  Re: OpenVMS Educational ProgramP >5$ > Shane.F.Smith@Healthnet.com wrote: > > C > > From a technical point of view, as you say, we have no concretec details.I > > Let's see what VMS Engineering deliver first. From a real world point  ofK > > view, COE certification is intended to garner a larger userbase for VMS  > inI > > the military, guarantee its survival for many years to come, and lead  toI > > more apps becoming available. These are all extremely important righti > now.J > > Keeping VMS "pure" (for want of a better word) is NOT going to keep itK > > alive. You only have to look at Windows to see how much weight an ideal-H > > design has in the current market - it's crap but it rules the world. > >uH > > This is a fight for survival. We can't afford the luxury of anyone's idea@ > > of a perfect environment. Not yours, not mine, not anyone's. >nJ > It has been written that these COE amendments would for a certain period (IH > assume at least 5 to 10 years) not be available to the public. So yourI > argument wouldn't fit. But my fear also wouldn't be appropriate. But myiK > point is that technically this is something very risky and done the wrongeE > way will kill VMS technically. Why would one use VMS with UNIX appsn insteadrH > of using UNIX? And putting engineering effort in that COE niche marketI > project will reduce the engineering power spent for the public version.p > Insane, isn't it?f >cJ > I'm sure that even the COE UNIX API would already be there this wouldn'tK > help VMS. Instead it might well kill VMS also from the marketing point ofeJ > view because the real VMS apps will vanish and there will no apps remainK > using the real power of VMS. And there will be a lot of apps spoiling VMSo2 > by not using its features like structured files. >e	 > > Shaner > > A > > Christof Brass <brass@infopuls.com> on 03/13/2001 03:20:40 PM0 > >  > > To:   Info-VAX@Mvb.Saic.Com  > > cc:R > >t- > > Subject:  Re: OpenVMS Educational Program4 > >8% > > steven.reece@quintiles.com wrote:e > > > K > > > This is getting to be something of a flame war come religious war.  I B > > > didn't expect the Spanish Inquisition and all that.......... > > >oG > > > Complexity of code is something that is almost part and parcel ofo	 > presentaC > > > product offerings.  This is because people are expecting mores	 functionsn > > andlF > > > more displays and in the case of some environments cruddy little > > paperclipsJ > > > that come up when you least expect/need/want them and flying bits of	 > > paperuI > > > that take up CPU time when you should actually be copying the file.t
 > > > Grrrrr!  > > >tI > > > I would expect that stuff like cluster transition capabilities, SCSe> > > > connections over switched LANs, memory channel and other
 interconnects  > isI > > > pretty complex to get right, as are a fair number of other parts ofP the1 > > VMS-H > > > operating system (not just the kernel).  Would you say that VMS is lowe > onK > > > quality?  No.  Would you say that these bells and whistles should nots > be& > > > added in (along with APMP)?  No. > > > E > > > What you're getting with the work that is being carried out arecI > > > capabilities on VMS to present Unix-like environment features.  Or,a as > > FredJ > > > Kleinsorge (not speaking for Compaq or OpenVMS or the US DoD) put it :  > > > I > > > >>>So what COE provides is a common look & feel for the application D > > > environment, which tends to drill down all the way to the UNIX command  > > lineI > > > interfaces.  You also have some "standards" - like POSIX and JAVA -. > which H > > > if people writing new segments adhere to, would allow source level > > > compatability.<<<< > > >n/ > > > then later in the same message he wrote :o > > >eJ > > > >>>This does not "replace" anything that VMS currently does, or even
 > > weakenH > > > it. All the capabilities are incremental, or supplemental.  But as > > always, I > > > poor UNIX code will work poorly, and good UNIX code will work well.t IiB > > > expect that when someone wants the code to "mesh" into a VMS > environment,K > > > they may have work to do beyond a simple port - or they can choose tos > > live( > > > with a UNIX fish in a VMS pond.<<< > > >r > > > Christof Brass wrote: D > > > >>>Quality is dependent on time and complexity i.e. effort and > > skilfulness.A > > > Having UNIX on VMS raises complexity and hence imposes morea
 difficulty > in > > > achieving quality. > > >d% > > > Why do we want another UNIX?<<<t > >qD > > While I'm not satified with these general statements of the well > estimatedsI > > Fred Kleinsorge and while I'm still expecting hard problems with thisi COEcF > > stuff I have to admit that I can't prove it with concrete details.: > > Unnecessary complexity reduces quality with given time frames/milestones.K > > The VMS feature rich environment doesn't need any adaption to any thirdeI > > party (DoD) rules - especially if these rules are strongly influenced  by > > UNIX crap.C > > The complexity of VMS is needed for delivering its features andn	 thereforeeE > > the tradeoff has been made to implement this quality features andeK > > sacrificing simplicity. UNIX is the opposite. Quality is lowered if you  > addtC > > something you don't need or which destroys the design or offersMH > > possiblities you better don't want to have because they don't fit in thel$ > > way things are best done so far. > >yJ > > Why would you like any of this COE stupidity from a technical point ofD > > view? And keep in mind that other things wouldn't be implemented becauseh > of# > > the effort this COE task needs.e   ------------------------------  % Date: Thu, 15 Mar 2001 17:56:40 -0600n+ From: Christopher Smith <csmith@amdocs.com> ( Subject: RE: OpenVMS Educational ProgramL Message-ID: <3B55D7F383B0D31197D9009027541CBF0BDD547E@cmiexch1.cmi.itds.com>   > -----Original Message-----2 > From: Christof Brass [mailto:brass@infopuls.com]  > > > Because that's where the apps run. If you make those same  > apps available onX@ > > VMS then VMS becomes a viable option in many areas where it  > is currently< > > not viable now. You try selling a manager VMS for a web  > server that serves& > > out Real Networks streaming media: > > @ > > Manager: I want to provide streaming Real Networks media to  > our customers./ > > Techie: Can I interest you in a VMS system?21 > > Manager: Does Real Networks server run on it?a/ > > Techie: Well, no, but it's really robust...r@ > > Manager looks at Techie as if he's grown a second head, and  > makes a mental6 > > note to surf Dice.com looking for a new techie....  @ > Here we are: the business case - I waited ages for someone to > > present one as to justify this COE stuff besides DoD. But I ; > don't buy it. Why using VMS for an app running alreay on e; > UNIX? A UNIX like app also. A UNIX quality app also. Why  ; > bother buying a VMS box? Does really anyone believe that  = > within the next five or ten years there will be in any way  = > close as much apps be around on VMS as on UNIX? As for the o@ > business case: the manager selects a UNIX box for sure if sHe ' > is thinking about app availability int> > any respect. Even if the currently needed apps are there. I 9 > personally doubt that a UNIX app will run flawless and h@ > efficient on VMS. Can we expect that these quickshot ports of F > UNIX apps will run in a cluster? Will they use RMS structered files?< > Again back to the business case: What is the advantage in : > buying a VMS box? I doubt that there is any if you plan < > running UNIX apps. Will the quality of the system (OS and 9 > app) be really better than on the original platform. I t? > seriously doubt that. Given the low quality of the UNIX apps  : > will the hidden bugs not pop up all around? Will that a ? > combination, VMS with a quickshot UNIX app on top, really be uD > more stable than on the original platform? I seriously doubt that.; > A shop that could afford VMS could afford top liga Intel  D > compatible crap with Linux or Solaris. Where is the business case?  7 Let me offer support for the business case in this way:e  F Let's assume that all of this COE is implemented without affecting the= quality of VMS in a negative way (that's another argument...)    Now our situation is this:  L Given the option of two high-end configurations (one Unix, and an equivalent VMS setup),t   Unix will run Unix apps.   VMS will run Unix apps.n  F VMS will _also_ run VMS apps, which may not alone justify buying a VMSI machine, but might be very strong as an "added bonus" to what you can geteJ with the Unix setup.  You may be surprised to know that there are a lot ofL features in VMS that even the most hard-core Unix people would love to have.G A lot of the time, the excuse to go with Unix instead is compatibility..  K I know that during my experience with Unix, I would have certainly liked itl3 to provide things like RMS, decent error reporting,eL _ACCESS_CONTROL_LISTS_!!!, _ABILITY_TO_DROP_OR_ASSUME_PRIVS_INDIVIDUALLY!!!,I etc, etc.  Do you have any idea how many people are paying for add-ons tocK Unix to provide any number of these things that are designed into VMS?  VMS K would give all of this from a single vendor -- there would be no excuse not 
 to use it.  A Now at this point, it may sound like I want VMS to be just a Unixi$ replacement -- but think about this:  E Once VMS has a foothold in these markets, people will start giving itc? software support again. What do they have to lose?  It would be J "compatible," but with more extra features than you could count, and whileL application reliability may be questionable in Unix apps, system reliabilityI need not suffer.  If Compaq handles this right, VMS could possibly _be_ a I Unix replacement, but in the migrational sense. People who don't have theaJ budget to do all of their software development in house might actually end/ up using it again.  That can't be bad, can it? O  L I really do believe that software availability is a big part of the problem.J The excuse used to be that it was "proprietary," but we both know that's aA load of crap, since all of Microshaft's products (which have moreY; marketshare than the 0% they deserve) are also proprietary.c  ? As Andrew Harrison said, it's no magic bullet, but it may help.r   Regards,   Christ  ! Christopher Smith, Perl Developerr Amdocs - Champaign, IL   /usr/bin/perl -e '? print((~"\x95\xc4\xe3"^"Just Another Perl Hacker.")."\x08!\n");  'n      ------------------------------  % Date: Fri, 16 Mar 2001 00:24:18 +0000 ) From: Christof Brass <brass@infopuls.com>w( Subject: Re: OpenVMS Educational Program, Message-ID: <3AB15D32.22E67A65@infopuls.com>  ! steven.reece@quintiles.com wrote:  >  > Christof,  > Calm down.  	 Not easy.b   > Christof commented/quoted:H > >>>> And if english were the only language in the world communications > wouldiL > > be a lot easier.  But reality is different.  And it doesn't make French,1 > > German or Spanish flawed, deficient or wrong.V > ( > I really don't get it. Are you stupid?I > Your example would be correct if you've said that someone invents a newhG > language (which in fact has been done with esperanto a well know widen > spread language).i > L > 1.Introducing unecessary complexity and choices into VMS is a disadvantage	 > per se.tK > 2.Yes, the UNIX way sucks and spoils VMS. We don't need these UNIX moronss > hacking around on VMS. > M > To explain it on kindergarten level: VMS is created by engineers, languagesnL > have been evolved over time without a central authority. I would perfectlyM > agree to have only one language all over the world - but surely it wouldn'tsJ > be English because English has too many flaws. The mismatch of scriptureK > and pronounciation is simply ridiculous. But this shouldn't change into aN0 > thread about natural languages so I stop here. > F > I ask listeners of this thread (if there are any remaining): is this5 > natural language example a well chosen argument?<<<t > . > Yes, it is a well chosen argument/analogy.     Could you please explain it to me on kindergarten level? I explained in detail why I'm sure that the example is complete wrong because there isn't any analogy. Telling me only that it is okay and there is an analogy which I don't see is not very helpful. And, sorry, the Esperanto example was a *joke* but I didn't flag it because I assumed everybody would know that Esperanto is as dead now as VMS never will. I'm sure there are less people fluent with Esperanto than with DCL.p  J >                                            To claim "we don't need theseK > Unix morons hacking around on VMS" is, perhaps, not as broad minded as its > might be.e > L > Techniques which are employed in one or more flavours of Unix may  be veryK > useful.  Techniques which are employed in OpenVMS may be very useful.  Ite= > doesn't necessarily mean that one is better than the other.a   Generaly speaking you are right. In fact almost all UNIX techniques are not well thought through and more important they don't fit within VMS. Do you want become VMS the PL/1 of the OSs?  G > There is no logical reason to suggest that the teams working on Tru64nK > within Compaq are any more or less intelligent, diligent or anything else1K > than their counterparts on the VMS Engineering teams.  They both have thetM > same masters at the end of the day - the Compaq shareholders and customers.h   Honestly this is - especially the last sentence - no argument at all. The room cleaning people also have the same employer and even ashole Palmer was hired by the the company as VMS engineering.   > Christof also commented:I > >>>Great example! The problem as stated above is the lack of standards. E > With VMS you are accustomed to VMS. With UNIX you are accustomed tooL > something vage, ever changing moving target. Where is the hosts file? - it5 > depends. Where is the mount table? - it depends.<<<  > D > So if we want to talk about standards, although OpenVMS was (to myI > knowledge) the first OS to be "branded" by the X/Open consortium (hencenH > OpenVMS), surely it is that which is different rather than the variousJ > flavours of Unix which all use arcane commands like man (for help), grepH > (which is a hell of a way to spell search) and pwd (for show default).  7Attention I got your irony. But this isn't the point. I even accept to certain degree these names. But unecessary irregularities with the options and the broken concept of parameter and option mixup and filename expansion of the shell. If you think this is good or if you even like that why we are talking here?e  J > On VMS where is the page file?  It depends.  Where is the swap file?  It > depends. .   Sorry, give me a break. The page file and the swap file??? Will you touch this any that often as you have to update the system configuration files?D  K >          What is the system startup command procedure?  It depends (since M > the system manager can call his own site-specific startup command procedurelK > from within or without SYSTARTUP_VMS.COM or SYSTARTUP_V5.COM etc).  WherenH > is the hosts file and what is it called?  It depends (you have variousL > options for IP stacks on OpenVMS).  What is the printer called on my floorM > of the building?  It depends.  Is there an easy way to set up my printer soeM > that I print to the one across the hall from me by default every time I loge  > in?  Probably, but it depends. >  > He also mentioned:M > >>>You really don't get the point. This seems to be a severe UNIX sickness.sH > Let me propose a cure: stop using UNIX for a short while (let's say 20 > years) and come back then.H > 1.There is not UNIX shell which is in any respect acceptable. All UNIXD > shells suck severely and not only from the features they offer but: > basically from the concept they handle user interaction.F > 2.The fact itself that there are several shells shows that nobody isL > content with what is there. Is this that difficult to understand? The factH > that there are no standards wrt the shells is a superflous problem.<<< > K > So do you set up symbols for your tasks that you do day to day or week to<B > week on your VMS system (whether it be in a command procedure orM > interactively)?  You do?  Well you're obviously not happy with what's there E > already then, are you?  It's obvious then that DCL is no good and a L > different CLI should be written for OpenVMS.  The fact that DCL ships with# > OpenVMS is a superfluous problem.l  wThe opposite is true. The fact that everybody can adjust the very same DCL to his/her personal pleasure shows that it is exactly the right tool. If I look at this analogy I think you aren't the right guy to judge about analogy correctnes. Setting up your symbols (VMS) is equivalent to defining your aliases (UNIX). Replacing the shell is equivalent of replacing the DCL CLI.w  . > >>>Did you read the UNIX-Haters Handbook?<<<K > No, but that's not relevant here.  You seem to do well in hating Unix and K > anything associated with Unix without any help at all (unless you _wrote_p > the book of course :-)).   This would help because you would know from well known very educated and skilful people why and in which respect UNIX sucks. This would shorten the discussion substantially.S  G > From your quoting and advocacy of the Unix-Haters' Handbook one mightBM > suggest (not that I would) that this is the sole source of your discontent.rL > If this is the case then I would _strongly_ urge you to have a chat with aD > local Unix Sys. Admin if there are any at your site.  you might beK > pleasantly surprised (as might he/she).  Each of you may learn something.c  As I stated before I have deep UNIX exposure since 1996 and I have to use UNIX (Solaris) at work at the moment. And I'll tell you what: it really sucks. More than I ever believed. Sometimes one central NFS server loses its mount points, Navigator dies every third day, Composer sucks by redrawing the complete screen with every single character input, Composer sucks when editing a large (300 lines) file by waiting 3 to 5 seconds until responding for the first character input or deletion after you changedX lines with cursor commands or the mouse. I could continue. The local sys admins are cursing. In fact there is a small but realistic chance right now to drop Solaris and to get rid of UNIX.   > Steve.   ------------------------------  # Date: Fri, 16 Mar 2001 00:46:26 GMT.4 From: "Terry C. Shannon" <terryshannon@mediaone.net>( Subject: Re: OpenVMS Educational Program< Message-ID: <Cnds6.9585$mH4.2460205@typhoon.ne.mediaone.net>  . <Shane.F.Smith@Healthnet.com> wrote in message> news:OF5E33DB02.C9EAA958-ON88256A10.0082DE99@foundation.com... >sK > I'm tempted to say "hook, line, sinker and copy of angling times...", but  I & > didn't really plan it that well. ;-) >uJ > We had a requirement for a web server. VMS came out of the analysis withL > the highest score of the candidate platforms, and would probably have beenL > chosen except for one thing. Some of the necessary apps weren't available.I > Showstopper. All those lovely features and good design you like so much J > couldn't save it, because of a missing piece. Needless to say, we're not > using VMS for that project.t > G > Now, if we'd had a port of those apps, and that list /was/ short, the  storynK > would have been different. If you think this is an isolated incident, I'mt > afraid you're sadly mistaken.  >a  L Of course it's not an isolated incident. I think we'll see Compaq take COE aF little further than the specs would indicate in an effort to ease apps
 availability.    ------------------------------  # Date: Thu, 15 Mar 2001 22:17:00 GMT   From: jlsue <jlsuexxxz@home.com>) Subject: Re: Possible security hole in...n8 Message-ID: <kcf2bt8hhqk12pk8qrd850b0k7jcsi2vaa@4ax.com>  F On Thu, 15 Mar 2001 15:20:09 -0000, "ph0bos" <ph0bos@shady.org> wrote:   >a >>E >>    There may be bugs, but in general most of the components of the D >>    various TCP/IP stacks do not execute with enough privileges toC >>    allow a cracker to do anything more than slightly annoying in J >>    the unlikely event that they actually find a security vulnerability. >> >sG >     Once again, if they're running at all, accept user input from theeH >     net and  are buggy, the 'privilege' people will gain is a shell onD >     your machine, running at the privs of the process so hijacked.E >     Personally, I think that's quite serious, but if you don't mindoF >     everyone and his donkey running round your machine, that's up toD >     you.  Whether they can then perform further attacks to elevateH >     privilege is a moot point - if they weren't in in the first place,E >     they wouldn't get to try.  If there are any elements of the tcp?K >     stack that they can execute locally, accept user input and are buggy,eA >     that run with elevated privs (and there are), then they can  >     gain additional privs. > C >     Quite apart from that, once they have a shell they can likely C >     launch attacks upon other machines, possibly other companies'l> >     machines, or possibly other of your machines on networks> >     that might not have been accessible to them in the first= >     place, taking advantage of any trust relationships that  >     might exist. >e  F This has me thinking, unfortunately, I don't have a newer system handyF right now to really check.  The UCX V4.1 ECO 2 system I did check runsF all UCX processes /detached, and without a CLI.  It sounds to me as ifA the TCP/IP server processes you are working with are running in asF process context that has a CLI, and hence where lib$spawn can actuallyD execute.  If they don't have a CLI, then this vulnerability wouldn't exist (I think).  E Is this really true?  Which servers have you been able to compromise?n What software versions?d   >a   ------------------------------    Date: 15 Mar 2001 17:50:15 -05003 From: malmberg@encompasserve.org (John E. Malmberg)n) Subject: Re: Possible security hole in...m3 Message-ID: <3PIYBPuXNk2F@eisner.encompasserve.org>-  @ In article <4e5s6.216852$Dd3.3036385@monolith.news.easynet.net>,# "ph0bos" <ph0bos@shady.org> writes:r  K > "John E. Malmberg" <malmberg@encompasserve.organization> wrote in messagea/ > news:g8mgftFpT$NB@eisner.encompasserve.org...e >>D >>    You only have access to the user mode stack, nothing that will >>    elevate your privileges. >oL >     Not so.  If you are executing code from remote, you gain local access.J >     If you are exploiting an image that is installed with privileges you. >     don't already have, you gain privileges.  G       Technically possible, however you must realize that in most casesvG       the privileges available from such an exploit are not really thata;       useful to do much more than create a minor annoyance.a  I       More ancient versions of VMS did have more vulnerability there, bute       that was long ago.  H       So I agree that it can be done, assuming that there is an actuallyE       some code that is able to write past an allocated buffer in theaE       shipping product.  There are many reasons that is not likely to        occur.   >>- >>    IIRC The URL is now in the OpenVMS FAQ.i >nG >     All of the 'VMS hacking texts' I've ever seen have been outdated,pI >     inaccurate and for the most part useless.  Not one of them has evers1 >     contained what I would call a real exploit.b  F       That is exactly what the person who posted the last one said andC       he claimed to have updated it.  It did not seem to help much.a  I >     That doesn't mean they don't exist.  VMS needs a decent code audit.X  E       You are presuming that OpenVMS Engineering does not do a decent 8       code audit?  What information do you have on this?  7 >> >     On top of that, for local holes, the procedureuD >> >     is actually considerably simpler than under unix, thanks toH >> >     the lib$spawn() call and the fact that it requires no argumentsD >> >     (unlike execl("/bin/sh", "sh", 0) the standard unix method.H >> >     Under VMS, all you need is a jump to lib$spawn - no 'shellcode' >> >     required. >>E >>    An interesting idea.  Missing a few concepts though.  There are  >>    much more easier ways. >aF >     I'm always interested in easier ways to do that sort of thing :)         That would be telling. >rF >>    And you still will not end up with any more privileges or rights >>    than you started up with.a >cG >      Once again, not so.  Your basic misconception is that people are C >      somehow attempting to manipulate the operating system itselfi? >      into giving them more privileges (a bit like the old VMS F >      store-the-privilege-mask-in-address-space-the-user-can-write-toG >      cock-up :) ).  The point is, use an overflow to hijack processess/ >      that are running at evelated privileges.i  /       I assure you that I had no misconception.a  G       I was assuming that access to a shell prompt or other control waspE       sucessful.  Almost all of the network services run with out anynF       privileges that would allow an intruder to cover their tracks or       to alter sensitive data.  F       In many cases though your initial assumption of what to do wouldB       not work.  For one reason or another OpenVMS knows that manyB       network processes are not allowed access to a command shell.  E       As for the other reasons the overrun methods would probably notM"       work: That would be telling.   <snip>  F >      Personally, I think that's quite serious, but if you don't mindG >      everyone and his donkey running round your machine, that's up tot >      you.e  B        I assure you that everyone and their donkey are not able to:        run arround any OpenVMS machine that I have set up.@        That is also probably the case with most OpenVMS systems,F        unless someone has tampered with the default security settings.  F        Security on a computer system is best used to prevent accidentsC        by authorized users.  Except for the few systems with direct E        non-firewall access to the public internet or a school campus,eA        the likely hood of a deliberate attack is extremely small.a  G        The probability of an internal accident or a hardware failure istD        very much larger, and that is where the most effort should be        done.  >        If that is done properly it is not likely that the rare1        deliberate attack will be able to do much.r  G        <The obvious consequences of getting unauthorized acces snipped>i  F        Basically at this point your discussion seems to be starting toH        consist of repeating the same points over and over, which require.        all the stars to be correctly alligned.  C        Yes, if all those conditions did exist, any operating system F        would be vulnerable.  I never contested the point.  No one can.  H        That does not mean that they have to exist or that they do exist,#        or even are likely to exist.f  D        In the case of OpenVMS, the window for such a bug to occur is        very very small.l  D        And you have not presented anything to convince me otherwise.    -John Personal Opinion Only.   ------------------------------   Date: 15 Mar 2001 22:10 PSTw) From: rankin@eql.caltech.edu (Pat Rankin)u# Subject: Re: Splitting a Large Filer/ Message-ID: <15MAR200122105908@eql.caltech.edu>   B In article <iSRr6.2511$y47.527428@news2-win.server.ntlworld.com>,\3  "robert.glen" <robert.glen@ntlworld.com> writes...y? > I have a large file which I can open up in my editor ( LSE ).  >eM > However I wish to split the file into two parts ( approx 6000 lines each ).n >rH > Is there a command to do this with, similar to the Unix split command?  E      There isn't any split command included with VMS.  This is prettyuD trivial to do with awk or perl, which are both available for VMS, orB probably even with TECO if you have to stick with tools already onC your system.  For splitting a file into only two parts, I'd get the 9 freeware Extract utility (but I'm biased :-) and just usee  0 $ EXTRACT /HEAD=6000 inputfile /OUTPUT=firsthalf2 $ EXTRACT /TAIL=-6000 inputfile /OUTPUT=secondhalf  D That first command would copy the first 6000 records into its outputD file.  The second command would skip the first 6000 records and copyD the remainder of the input into its output file.   It shouldn't have any trouble with long records.  E      You can obtain the Extract utility over the net from ftp.wku.edu C in [.vms.fileserv]extract.zip.  There's also a way to get files vias1 email from that system, but I don't remember how.'  C      For more than two parts, you could still use Extract--with itseB /RECORD=(START=x,END=y) qualifier--but it would be better to use aE tool which can produce multiple output files in a single pass through D the input file.  Unfortunately Extract can't do that.  (Actually youB could achieve the same effect using OPEN/READ followed by repeated@ use of EXTRACT/HEAD on the resulting process permanent file, butA that's a bit complicated for my taste; there won't be any sort of 8 indication once you finally reach the end of the input.)  2                 Pat Rankin, rankin@eql.caltech.edu   ------------------------------  % Date: Thu, 15 Mar 2001 16:18:51 -0500m2 From: "Sue Skonetski" <susan.skonetski@compaq.com>2 Subject: Talk to Rich Marcello - Austin Texas area0 Message-ID: <olas6.39$eE2.1998@news.cpqcorp.net>  J If you are in the Austin Texas area this would be a chance for you to talk5 directly to Richard Marcello VP of the OpenVMS group.s   sue   1 _________________________________________________-  E The Austin Chapter of DECUS/Encompass and Compaq Computer Corporation ) invites you to "Talk with Rich Marcello!"   K Rich Marcello, Vice President of the OpenVMS Systems and Software Group foreB Compaq Computer Corporation, will be our special guest at the nextL DECUS/Encompass Local User Group meeting. Please join us for a discussion of= the exciting developments in OpenVMS both current and future.t   Date: Friday, March 30, 2001   Time: 2:00 - 5:00 p.m.  H Location: Compaq Computer Corporation, 14231 Tandem Blvd., Austin, Texas  J Refreshments will be served. Seats are limited so please RSVP via email to  E Jaime Matute at jaime.matute@austinenergy.com or call (512) 322-6265.(  2 Don't miss this unique opportunity! See you there!    	 Ed Stuartt   DECUS BATLUG Co-Chair   % Manager, Systems and Desktop Servicesa   Information Technology Servicese   City of Austin, Austin Energyp   512-322-6372   Ed.Stuart@austinenergy.com   ------------------------------  % Date: Thu, 15 Mar 2001 15:32:52 -0600e1 From: "Dave Gudewicz" <david.gudewicz@abbott.com>n6 Subject: Re: Talk to Rich Marcello - Austin Texas area8 Message-ID: <98rce8$2hs$1@fizban.fizban.pprd.abbott.com>  F If someone in attendance could summarizer this meeting, I and probably  others here would be interested.   Thanks,f   Dave...   = "Sue Skonetski" <susan.skonetski@compaq.com> wrote in messagef* news:olas6.39$eE2.1998@news.cpqcorp.net...L > If you are in the Austin Texas area this would be a chance for you to talk7 > directly to Richard Marcello VP of the OpenVMS group.  >  > suem >g3 > _________________________________________________c >rG > The Austin Chapter of DECUS/Encompass and Compaq Computer Corporation + > invites you to "Talk with Rich Marcello!"l >?I > Rich Marcello, Vice President of the OpenVMS Systems and Software Groupi for D > Compaq Computer Corporation, will be our special guest at the nextK > DECUS/Encompass Local User Group meeting. Please join us for a discussionl of? > the exciting developments in OpenVMS both current and future.d >  > Date: Friday, March 30, 2001 >r > Time: 2:00 - 5:00 p.m. >iJ > Location: Compaq Computer Corporation, 14231 Tandem Blvd., Austin, Texas >XL > Refreshments will be served. Seats are limited so please RSVP via email to >UG > Jaime Matute at jaime.matute@austinenergy.com or call (512) 322-6265.e >y4 > Don't miss this unique opportunity! See you there! >l >r > Ed Stuartt >e > DECUS BATLUG Co-Chair. > ' > Manager, Systems and Desktop Servicess >l! > Information Technology Servicesi >w > City of Austin, Austin Energyu >t > 512-322-6372 >a > Ed.Stuart@austinenergy.com >  >  >    ------------------------------  % Date: Thu, 15 Mar 2001 14:53:53 -0800 ! From: Shane.F.Smith@Healthnet.comk6 Subject: Re: Talk to Rich Marcello - Austin Texas areaD Message-ID: <OF46C242CE.4F24E085-ON88256A10.007DA11E@foundation.com>  H If somebody in this meeting would be prepared to kidnap him, arrange forK some compromising pictures to be taken with him and a couple of sheep, thena help us blackmail him...... ;-)    Shaner          C Dave Gudewicz <david.gudewicz@abbott.com> on 03/15/2001 01:32:52 PMe   To:   Info-VAX@Mvb.Saic.Comn cc:e  7 Subject:  Re: Talk to Rich Marcello - Austin Texas area     F If someone in attendance could summarizer this meeting, I and probably  others here would be interested.   Thanks,e   Dave...i  = "Sue Skonetski" <susan.skonetski@compaq.com> wrote in messagea* news:olas6.39$eE2.1998@news.cpqcorp.net...G > If you are in the Austin Texas area this would be a chance for you too talk7 > directly to Richard Marcello VP of the OpenVMS group.i >u > suee >a3 > _________________________________________________. > G > The Austin Chapter of DECUS/Encompass and Compaq Computer Corporations+ > invites you to "Talk with Rich Marcello!"u >.I > Rich Marcello, Vice President of the OpenVMS Systems and Software Groupn foroD > Compaq Computer Corporation, will be our special guest at the nextK > DECUS/Encompass Local User Group meeting. Please join us for a discussion  of? > the exciting developments in OpenVMS both current and future.S >  > Date: Friday, March 30, 2001 >. > Time: 2:00 - 5:00 p.m. >sJ > Location: Compaq Computer Corporation, 14231 Tandem Blvd., Austin, Texas >eI > Refreshments will be served. Seats are limited so please RSVP via emaila to >aG > Jaime Matute at jaime.matute@austinenergy.com or call (512) 322-6265.e >v4 > Don't miss this unique opportunity! See you there! >t >N > Ed Stuart  >r > DECUS BATLUG Co-Chairk >y' > Manager, Systems and Desktop Services  >g! > Information Technology Servicesw >  > City of Austin, Austin Energyb >a > 512-322-6372 >s > Ed.Stuart@austinenergy.com >h >c >    ------------------------------  % Date: Thu, 15 Mar 2001 16:32:36 -0800s! From: Shane.F.Smith@Healthnet.coma( Subject: TCP/IP SYN flood DoS protectionD Message-ID: <OF56B02771.1B36925D-ON88256A11.0002AFA0@foundation.com>  G I realise this article isn't about VMS, but it has some good info about0H TCP/IP SYN flooding, and details what looks to be a neat solution to it.H Take a look, if that kind of thing interests you. I guess the idea couldI even be incorporated into UCX or Multinet or whatever, if the maintainerse are interested.A    http://grc.com/r&d/nomoredos.htm   Shanes   ------------------------------  % Date: Thu, 15 Mar 2001 21:51:11 +0100u, From: "Bart Zorn" <B.Zorn@TrueBit.nospam.nl>, Subject: TDF change problems with V7.3 EFT2?* Message-ID: <98ra14$4v4$1@news1.xs4all.nl>  
 Hello all!  " The TDF change is due soon, again.  K I am running OpenVMS G7.3 (EFT2) on my home Alpha, and to my dismay I foundeB out that the rule for the TDF changes in Europe are wrong (again).  I I don't recall which ECO it was, but finally we got it right in DTSS. TheeK rule for western Europe (+0100 standard, +0200 day light saving) should be:a  0 1 3600 MET 0 MET-1MET DST-2,M3.5.0/02,M10.5.0/03  F Essential is, that we  change the time on the last sunday of march and6 october. I thought that is the case is the US as well.  4 HOWEVER, the new time changing code gives us, again:  0 1 3600 MET 0 MET-1MET DST-2,M3.4.0/02,M10.4.0/03  2 just like we had in the early days of DECnet-plus.  % Or am I missing something altogether?e  H Now that I am on the subject, could we have standard timezone names like +0100 and +0200, please?   Regards,  	 Bart Zornf   ------------------------------    Date: 15 Mar 2001 12:52:19 +0800, From: Paul Repacholi <prep@prep.synonet.com>( Subject: The Chris and Bill show. Act II- Message-ID: <871yrzy6po.fsf@prep.synonet.com>a   Summary to date.   choir left:w@ If VMS does not get its act and apps together, it's history. COE will help get apps onto VMS.  
 choir right: y8 Unix is a pile of crud, and we don't want it, don't wantC it, won't have it.  The spooks are trying to subvert us and lure ust$ into a secret cave under Washington.   OK.i  1 Now, perhaps we could look at some of the issues.c  E I can see a BIG plus from COE: It will be a solid "We are not dying!" D peg to stomp on the fudsters with. 20 year is a BIG lump in any part) of the computer world except IBM and VMS.h  9 The down side, is that the POSIX standard has fundamental > incompatabilities with VMS. File Protection/Permissions is oneD example. This *could* impact back into VMS. And a pile of others I'm sure.n  D As to 'quality', remember; this is the military. Quality is the sizeC of the paper pile. And there is little profit in a deployed system, 6 you want it being 'developed' for as long as possible.   --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.o@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------   Date: 15 Mar 2001 13:21:17 CDT= From: wayne@tachysoft.xxx.412538.killspam.015a (Wayne Sewell)m3 Subject: the Northeast Digital Systems Ebay Folliesm. Message-ID: <XQqm2vVImwec@tachxxsoftxxconsult>  N A while back, someone made a comment here that Northeast Digital Systems had aM shitload of alphas available on ebay.  For months now, we've been seeing whati, appeared to be a huge stock of alphas go by.    M However, we have not been seeing a series of alpha sales.   We are seeing theuO *same* alpha systems offered over and over because they were not sold.   If youeI actually track the auctions, you see that they do not complete becasue of-N "reserve price not met."   Bascially, the machines are not selling because theJ reserve price is set at some high value, much higher than anybody has been willing to pay thus far.  @ As far as I can tell, they have not sold *one* of those systems.  H Northeast Digital's feedback rating has been stuck at 23 ever since theyO started posting these alphas months ago.  You have to complete a transaction tofN get feedback.  :-)  Of the two feedback entries since October 2000, neither is for an alpha system.  L They appear to have about a dozen Digital Workstation 433/500/600 models forK sale, both A and AU, though it seems like thousands of units because of thea repetition.s  L As soon as an auction terminates, they immediately resubmit the same machineM in a new one.  I don't know if they adjust the reserve price downward or not.h    E I have absolutely no idea what their reserve prices are, but they arefN apparently far above what I am willing to pay.   I wouldn't mind having one ofN the 600AU units and have tried to probe with the highest price I can bear, butN still get the beloved "you are high bidder, but the reserve price has not beenO met".   Since *all* of the units are still sitting there unsold, it appears theo5 reserve is far above what *anyone* is willing to pay.   N Does anyone know of a Northeast Digital alpha auction that actually completed?  " If so, what model and final price?   -- bO ===============================================================================hM Wayne Sewell, Tachyon Software Consulting  (281)812-0738  wayne@tachysoft.xxxx: http://www.tachysoft.xxx/www/tachyon.html and wayne.html  K change .xxx to .com in addresses above, assuming you are not a spambot  :-)tO =============================================================================== O Dean Wormer to Flounder: "Fat, drunk, and stupid is no way to go through life."    ------------------------------    Date: 15 Mar 2001 15:51:50 -05007 From: hamilton@encompasserve.org (Bradford J. Hamilton) 7 Subject: Re: the Northeast Digital Systems Ebay Folliesn3 Message-ID: <igPKyQUz4Ys$@eisner.encompasserve.org>>   Wayne,  G Your suspicions are correct - they are selling for a very high reserve.S   If you go to their website:$  / http://www.northeastdigital.com/Alphacenter.htm   ) you can see what their prices really are.s  F I bought a laptop from them - good deal, nice people.  Their office is> *hard* to find, but I saved myself $25 on shipping costs...:-)  2 Can't say how good the Alphas really are...YMMV...   Brad  o >In article <XQqm2vVImwec@tachxxsoftxxconsult>, wayne@tachysoft.xxx.412538.killspam.015a (Wayne Sewell) writes:s > P > A while back, someone made a comment here that Northeast Digital Systems had aO > shitload of alphas available on ebay.  For months now, we've been seeing what . > appeared to be a huge stock of alphas go by. >  > O > However, we have not been seeing a series of alpha sales.   We are seeing thecQ > *same* alpha systems offered over and over because they were not sold.   If you K > actually track the auctions, you see that they do not complete becasue oftP > "reserve price not met."   Bascially, the machines are not selling because theL > reserve price is set at some high value, much higher than anybody has been > willing to pay thus far. > B > As far as I can tell, they have not sold *one* of those systems. > J > Northeast Digital's feedback rating has been stuck at 23 ever since theyQ > started posting these alphas months ago.  You have to complete a transaction torP > get feedback.  :-)  Of the two feedback entries since October 2000, neither is > for an alpha system. > N > They appear to have about a dozen Digital Workstation 433/500/600 models forM > sale, both A and AU, though it seems like thousands of units because of the 
 > repetition.  > N > As soon as an auction terminates, they immediately resubmit the same machineO > in a new one.  I don't know if they adjust the reserve price downward or not.M >  > G > I have absolutely no idea what their reserve prices are, but they are.P > apparently far above what I am willing to pay.   I wouldn't mind having one ofP > the 600AU units and have tried to probe with the highest price I can bear, butP > still get the beloved "you are high bidder, but the reserve price has not beenQ > met".   Since *all* of the units are still sitting there unsold, it appears thei7 > reserve is far above what *anyone* is willing to pay.a > P > Does anyone know of a Northeast Digital alpha auction that actually completed? > $ > If so, what model and final price? >  > -- tQ > ===============================================================================nO > Wayne Sewell, Tachyon Software Consulting  (281)812-0738  wayne@tachysoft.xxxs< > http://www.tachysoft.xxx/www/tachyon.html and wayne.html  M > change .xxx to .com in addresses above, assuming you are not a spambot  :-) Q > ===============================================================================dQ > Dean Wormer to Flounder: "Fat, drunk, and stupid is no way to go through life."m   ------------------------------  # Date: Thu, 15 Mar 2001 22:56:13 GMTk2 From: "Zane H. Healy" <healyzh@shell1.aracnet.com>7 Subject: Re: the Northeast Digital Systems Ebay Folliesl6 Message-ID: <hMbs6.11690$a3.26482@typhoon.aracnet.com>  8 Bradford J. Hamilton <hamilton@encompasserve.org> wrote:I > Your suspicions are correct - they are selling for a very high reserve.l   > If you go to their website:c  1 > http://www.northeastdigital.com/Alphacenter.htm1  J Ouch.  You're better off buying from Island Computers!  What gets me aboutK the systems Northeast is selling is that they're really nice sounding boxes  with horrible video cards!   			Zanen   ------------------------------    Date: 15 Mar 2001 20:04:33 -05009 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)d7 Subject: Re: the Northeast Digital Systems Ebay Follies 3 Message-ID: <7iVYVSKc9IPF@eisner.encompasserve.org>T  n In article <XQqm2vVImwec@tachxxsoftxxconsult>, wayne@tachysoft.xxx.412538.killspam.015a (Wayne Sewell) writes: > P > A while back, someone made a comment here that Northeast Digital Systems had aO > shitload of alphas available on ebay.  For months now, we've been seeing whate. > appeared to be a huge stock of alphas go by.  D And I have certainly ignored them because they never offer them withB a VMS license transfer.  There seem to be few sellers on eBay with@ the sense to include a license transfer.  All it costs them is a little paperwork.t  G They don't seem to offer them with Tru64 (nee DEC Unix) (nee DEC OSF/1) B licenses either, and I cannot believe all those used machines were6 bought without an operating system for use with Linux.   ------------------------------  # Date: Fri, 16 Mar 2001 02:53:37 GMT 2 From: "John Fredrickson" <jafred@bellatlantic.net>7 Subject: Re: the Northeast Digital Systems Ebay Folliesi7 Message-ID: <Refs6.3791$vc.2036801@typhoon2.ba-dsg.net>l   Wayne,  J You are quite correct. I myself have participated in several eBay auctionsG for Alphas offered by Northwest and although I was the high bidder in alJ couple of cases, I never reached the reserve price. The maret wants to buyK these machines for no more than $500. Northwest wants to sell them on theirnI web site for about $1800+. They are just using eBay as a free advertisinga source.    John Fredrickson  ! --------------------------------- J "Wayne Sewell" <wayne@tachysoft.xxx.412538.killspam.015a> wrote in message( news:XQqm2vVImwec@tachxxsoftxxconsult... >lJ > A while back, someone made a comment here that Northeast Digital Systems had aaJ > shitload of alphas available on ebay.  For months now, we've been seeing what. > appeared to be a huge stock of alphas go by. >  >_K > However, we have not been seeing a series of alpha sales.   We are seeing/ thepH > *same* alpha systems offered over and over because they were not sold. If youK > actually track the auctions, you see that they do not complete becasue ofeL > "reserve price not met."   Bascially, the machines are not selling because thesL > reserve price is set at some high value, much higher than anybody has been > willing to pay thus far. >rB > As far as I can tell, they have not sold *one* of those systems. >.J > Northeast Digital's feedback rating has been stuck at 23 ever since theyB > started posting these alphas months ago.  You have to complete a transaction toE > get feedback.  :-)  Of the two feedback entries since October 2000, 
 neither is > for an alpha system. >eJ > They appear to have about a dozen Digital Workstation 433/500/600 models forsI > sale, both A and AU, though it seems like thousands of units because ofn theg
 > repetition.3 > F > As soon as an auction terminates, they immediately resubmit the same machineaJ > in a new one.  I don't know if they adjust the reserve price downward or not. >R >TG > I have absolutely no idea what their reserve prices are, but they arenI > apparently far above what I am willing to pay.   I wouldn't mind havingm one ofL > the 600AU units and have tried to probe with the highest price I can bear, butoK > still get the beloved "you are high bidder, but the reserve price has not8 beenE > met".   Since *all* of the units are still sitting there unsold, ith appears thet7 > reserve is far above what *anyone* is willing to pay.g >pE > Does anyone know of a Northeast Digital alpha auction that actuallyc
 completed? >t$ > If so, what model and final price? >C > -- >oL ============================================================================ === : > Wayne Sewell, Tachyon Software Consulting  (281)812-0738 wayne@tachysoft.xxxp: > http://www.tachysoft.xxx/www/tachyon.html and wayne.htmlH > change .xxx to .com in addresses above, assuming you are not a spambot :-)i >gL ============================================================================ ===0J > Dean Wormer to Flounder: "Fat, drunk, and stupid is no way to go through life."   ------------------------------  # Date: Fri, 16 Mar 2001 04:21:03 GMTe; From: Mark Garrett <Mark.Garrett@wedontwantyourspam.com.au>a Subject: Re: VMS wanted list.cC Message-ID: <B6D7DF76.148C7%Mark.Garrett@wedontwantyourspam.com.au>o  H in article 3AAC6121.1C754AE2@rdrop.com, Dean Woodward at deanw@rdrop.com wrote on 12/03/2001 16:39:   > Mark Garrett wrote:h >> yB >> in article rUslzcAE$7$5@eisner.encompasserve.org, Bob Kaplow at >>  G >>> I'd still settle for having Palmer be the next CEO of Microsoft :-(- >> aJ >> What a great idea :) He could manage the split up of Microsoft so well,N >> with his wealth of experience at screwing DEC/Digital. He could maybe break? >> it down to bits small enough for some PeeCee vendor to buy:)  > @ > How would Compaq owning the M$ OS's make our situation better?L I wasn't suggesting CPQ buy them, just that Mr. P is well suited to reducingJ otherwise large fish down to size that smaller fish can eat ;) And I mightL have been thinking one of the other fun PC companies bought them eg. DELL :)   ------------------------------  % Date: Thu, 15 Mar 2001 19:39:22 +0000d4 From: John Laird <john@laird-towers.freeserve.co.uk>. Subject: Re: Was it here?  1GB disk for VS31008 Message-ID: <5h52btoun4eindcsjrii5q2rkqt6dtfide@4ax.com>  @ On Thu, 15 Mar 2001 18:14:00 GMT, system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) wrote:s  C >>Since he said Seagate 5.25 disks, I assume that they are ST41200N 6 >>drives, which work beautifully as VS3100 boot disks. >_ >Give that man a cheroot!  _  F Is there a 5.25" drive bay in a VS3100 ?  I can only (dimly) recall anG optional drive plate which had about 3 slots for low-profile or at most, half-height 3.5" devices.V  F This side of the pond, external SCSI cases run about 1 arm plus leg...  	 	John ;-)l -- t
 John Laird   ------------------------------  # Date: Thu, 15 Mar 2001 21:02:17 GMTs= From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)n. Subject: Re: Was it here?  1GB disk for VS31000 Message-ID: <009F90F1.C5E5577C@SendSpamHere.ORG>  o In article <5h52btoun4eindcsjrii5q2rkqt6dtfide@4ax.com>, John Laird <john@laird-towers.freeserve.co.uk> writes: A >On Thu, 15 Mar 2001 18:14:00 GMT, system@SendSpamHere.ORG (Briani  >Schenkenberger, VAXman-) wrote: > D >>>Since he said Seagate 5.25 disks, I assume that they are ST41200N7 >>>drives, which work beautifully as VS3100 boot disks.s >> >>Give that man a cheroot!   >oG >Is there a 5.25" drive bay in a VS3100 ?  I can only (dimly) recall an H >optional drive plate which had about 3 slots for low-profile or at most >half-height 3.5" devices. >,G >This side of the pond, external SCSI cases run about 1 arm plus leg...:  E Get yourself a 5V/12V extension and run it outside of the box.  Plug 6D into the drive and put the drive on the SCSI chain.  Just be carefulC not to spill your pint on it if you get too pissed on POETS day. ;)w   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMu             O city, n., 1. a place where trees are cut down and streets are named after them.n   ------------------------------  % Date: Thu, 15 Mar 2001 17:38:54 -0600-7 From: "David J. Dachtera" <djesys.nospam@earthlink.net>1) Subject: Re: Write CD on PC for OVMS Readi- Message-ID: <3AB1528E.FBC6EE9B@earthlink.net>    Merlier Franck wrote:9 > @ > Phil <phil.judi.maslin@clear.net.nz> a crit dans le message : > 3aa899ee$1@clear.net.nz...M > > Anyone know of any special software available or special requirements for  > > writing ofM > > CD's on PC (Win2000 or NT4) for reading by CD drive on Alpha running OVMSyM > > 7.1. Long file names are the priority as ISO9660/Joliet only seem to gived4 > > n.3. Required SAS files have long... extensions. > >, > >0 > 
 > 1st Step > " > Make a good file structure on PC > 4 > like c:\user\data for Mydisk:[user.data] directory > < > Do an iso file system with Fireburner (bulid track option.7 > Fireburner create an .iso file with long file system.0 > [snip]  ( Can you tell us more about "Fireburner"?  G ...including where to get it? Is it freeware? ...shareware? ...payware?n   -- b David J. Dachtera  dba DJE Systems  http://www.djesys.com/  : Unofficial Affordable OpenVMS Home Page and Message Board: http://www.djesys.com/vms/soho/2  F This *IS* an OpenVMS-related newsgroup. So, a certain bias in postings is to be expected.  @ Feel free to exercise your rights of free speech and expression.  F However, attacks against individual posters, or groups of posters, are strongly discouraged.l   ------------------------------   Date: 16 Mar 2001 06:15:25 GMT- From: djweath@attglobal.net (Dave Weatherall).( Subject: Re: [fun] DCL minute of the day5 Message-ID: <DTiotGxQ0bj6-pn2-ZUDoQfRPumLH@localhost>   D On Wed, 14 Mar 2001 20:00:26, paddy.o'brien@zzz.tg.nsw.gov.au wrote:   > Carl,O > > > >Uh, if you have a sytem running VMS then you ahve this doc: > >w
 > >INITIALIZE  > S > Thanks, Carl, I forgot the obvious place -- when in doubt use HELP.  I suppose I hB > had forgotten because I am on 7.2 and "dis-remembered" the help. > H > >(This would be for an old style ODS-2 disk, not the new bigger bitmap# > >ODS-2 disks or the ODS-5 disks.)  > S > As a developer, I have enough problems with some of the directory paths my users A= > use if I have to pick up their run data for debug purposes.e > R > As an admin on my systems, I knew that me as a developer did not want the ODS-5 R > hassle within my programs.  I compromised with myself, and only installed ODS-2  > :-)  >  > Regards, Paddy  D I know the pain. You run your image from your development directory D taking the data from the user. They are very organised and have big D directory tree many levels deep and meaningful filenames. A pain to F type. I normally just point a logical name at the long one and specifyF that to whatever I'm testing. Saves typing but you have  to be careful. if you're checking one query after another :-)   Cheers - Dave.   ------------------------------  # Date: Thu, 15 Mar 2001 19:59:58 GMTh, From: Didier Morandi <Didier.Morandi@gmx.ch>' Subject: Re: [INFO] RDMS-F-NOMONHOMEDIRs& Message-ID: <3AB11F1C.E8A3DA9C@gmx.ch>  O Richard, sys$sysroot is a concealed logical name, not a member of a search liste afaik.   D.   "Richard D. Piccard" wrote:b >  > That won't get you far:m >  > $ sh log sys$manager> >    "SYS$MANAGER" = "SYS$SYSROOT:[SYSMGR]" (LNM$SYSTEM_TABLE) > $ sh log sys$sysroot9 >    "SYS$SYSROOT" = "$1$DKA0:[SYS0.]" (LNM$SYSTEM_TABLE)  >         = "SYS$COMMON:"dB > 1  "SYS$COMMON" = "$1$DKA0:[SYS0.SYSCOMMON.]" (LNM$SYSTEM_TABLE)   ------------------------------  % Date: Thu, 15 Mar 2001 15:30:34 -0500 - From: "Richard D. Piccard" <piccard@ohio.edu>d' Subject: Re: [INFO] RDMS-F-NOMONHOMEDIRi( Message-ID: <3AB1265E.EBD5C5BB@ohio.edu>  Q I always interpreted the continuation line as evidence of it being a search list;f% in our case, quoted below, it is boths       $1$DKA0:[SYS0.]s   and also       $1$DKA0:[SYS0.SYSCOMMON.]a  7 Have I been misinterpreting the output of Show Logical?c  #                                 RDPe     Didier Morandi wrote:d  Q > Richard, sys$sysroot is a concealed logical name, not a member of a search liste > afaik. >e > D. >a > "Richard D. Piccard" wrote:  > >u > > That won't get you far:  > >  > > $ sh log sys$manager@ > >    "SYS$MANAGER" = "SYS$SYSROOT:[SYSMGR]" (LNM$SYSTEM_TABLE) > > $ sh log sys$sysroot; > >    "SYS$SYSROOT" = "$1$DKA0:[SYS0.]" (LNM$SYSTEM_TABLE)i > >         = "SYS$COMMON:"sD > > 1  "SYS$COMMON" = "$1$DKA0:[SYS0.SYSCOMMON.]" (LNM$SYSTEM_TABLE)   --B ==================================================================B Dick Piccard                           Academic Technology ManagerB piccard@ohio.edu                                 Computer ServicesB http://oak.cats.ohiou.edu/~piccard/                Ohio University   ------------------------------  % Date: Thu, 15 Mar 2001 17:16:38 +0000l- From: Tim Llewellyn <tim.llewellyn@bbc.co.uk>u6 Subject: Re: [Q] C errors exit handling (and training)) Message-ID: <3AB0F8F6.EC959148@bbc.co.uk>=   Didier Morandi wrote:e   > Tim Llewellyn wrote: >n#> > if you want to change the default signal handling behaviour (ie not exit on -E- and -F- severity), you will need to declare your own condition handler (LIB$ESTABLISH, but there is probably a C RTL to do it too) and do your own message logging with $PUTMSG, $GETMSG, or best of all $FAOL.  >r > Yep, I know it.r > I > > Disclaimer, I have done this before and it worked well, but not in C.$ > G > Disclaimer, I have done this in Fortran before and it worked well :-)r >i  ) Suprise suprise, I did it in Fortran too.e  --o6 Tim Llewellyn, OpenVMS Infrastructure, Remarcs Project0 MedAS at the BBC, Whiteladies Road, Bristol, UK.A Email tim.llewellyn@bbc.co.uk. Home tim.llewellyn@cableinet.co.ukr  A I speak for myself only and my views in no way represent those ofk MedAS or the BBC.    ------------------------------  # Date: Thu, 15 Mar 2001 20:13:05 GMTr, From: Didier Morandi <Didier.Morandi@gmx.ch>* Subject: Re: [Q] skip [vms$common] parsing& Message-ID: <3AB1222D.D9987F8E@gmx.ch>  & SLKP1_dmo> dir [vms$common.sysmgr]/tot+ Directory SYS$SYSDEVICE:[VMS$COMMON.SYSMGR]p& Total of 199 files, 8112/10620 blocks.    ISLKP1_dmo> dir sys$manager:/tot Directory SYS$SYSROOT:[SYSMGR]' Total of 516 files, 70095/77400 blocks.u Directory SYS$COMMON:[SYSMGR]b& Total of 199 files, 8112/10620 blocks.< Grand total of 2 directories, 715 files, 78207/88020 blocks.  + ISLKP1_dmo> dir [sys0.syscommon.sysmgr]/tott/ Directory SYS$SYSDEVICE:[SYS0.SYSCOMMON.SYSMGR] & Total of 199 files, 8112/10620 blocks.  P How can you say "VMS$COMMON" .NES. "SYSCOMMON" ? Of course it is. There is a setH file/enter if I remember well. And the total of 199 files for 8112/10620M demonstrates that we are talking about the same sys$manager directory, aren't & we? (or I missed something somewhere).   D.   "David J. Dachtera" wrote:> > On the other hand, remember: "VMS$COMMON" .NES. "SYSCOMMON".   ------------------------------  % Date: Thu, 15 Mar 2001 21:31:14 -0600u7 From: "David J. Dachtera" <djesys.nospam@earthlink.net>e* Subject: Re: [Q] skip [vms$common] parsing- Message-ID: <3AB18902.E828DA64@earthlink.net>    Didier Morandi wrote:d > ( > SLKP1_dmo> dir [vms$common.sysmgr]/tot- > Directory SYS$SYSDEVICE:[VMS$COMMON.SYSMGR] ( > Total of 199 files, 8112/10620 blocks. > " > ISLKP1_dmo> dir sys$manager:/tot  > Directory SYS$SYSROOT:[SYSMGR]) > Total of 516 files, 70095/77400 blocks.b > Directory SYS$COMMON:[SYSMGR] ( > Total of 199 files, 8112/10620 blocks.> > Grand total of 2 directories, 715 files, 78207/88020 blocks. > - > ISLKP1_dmo> dir [sys0.syscommon.sysmgr]/totr1 > Directory SYS$SYSDEVICE:[SYS0.SYSCOMMON.SYSMGR]s( > Total of 199 files, 8112/10620 blocks. > 2 > How can you say "VMS$COMMON" .NES. "SYSCOMMON" ?   Well...    DJAS01::DDACHTERA$ sh sym say=   SAY == "WRITE SYS$OUTPUT"=7 DJAS01::DDACHTERA$ say ("vms$common" .nes. "syscommon")a 1o DJAS01::DDACHTERA$ :-)  ! > Of course it is. There is a set   > file/enter if I remember well.  F Seriously, yes - [SYS%.SYSCOMMON] and/or [SYS%%.SYSCOMMON] are ALIASes/ for [VMS$COMMON] (at, least we HOPE they are!).=   However, ...  + > And the total of 199 files for 8112/10620dO > demonstrates that we are talking about the same sys$manager directory, aren't.( > we? (or I missed something somewhere).  A ...remember, or perhaps I should say understand how the DIRECTORYeG command works. DIRECTORY deals with - you guessed it - directories, andlD the entries there in, not FIDs. As we've discussed in other threads,D there is currently no easy way to find all the aliases for any file.D DIRECTORY and the services which support it look at the names of theH files as entered in directories, not their FIDs. FIDs and paths are only? remotely related. Therefore, the string [mis]match explains the  phenomenon (sort of).o  G Remember also that BACKUP is a separate animal unto itself. It goes outlA of its way to identify aliases and track files by FID (when usinge3 /IMAGE), something DIRECTORY does not bother to do.d  A You can test this yourself. If you have a common system disk in alE cluster, for example, assuming you have only roots 10 and 11, then...o  $ $ SET DEFAULT SYS$SYSDEVICE:[000000]7 $ DIRECT/EXCLUDE=[SYS10.SYSCOMMON...]*.*;* [*...]/GRANDW7 $ DIRECT/EXCLUDE=[SYS11.SYSCOMMON...]*.*;* [*...]/GRANDc6 $ DIRECT/EXCLUDE=[SYS*.SYSCOMMON...]*.*;* [*...]/GRAND2 $ DIRECT/EXCLUDE=[VMS$COMMON...]*.*;* [*...]/GRAND  H ...all of the above DIRECT[ORY] commands will produce different results.   Try it for yourself. :-)   -- a David J. Dachterar dba DJE Systems  http://www.djesys.com/  : Unofficial Affordable OpenVMS Home Page and Message Board: http://www.djesys.com/vms/soho/e  F This *IS* an OpenVMS-related newsgroup. So, a certain bias in postings is to be expected.  @ Feel free to exercise your rights of free speech and expression.  F However, attacks against individual posters, or groups of posters, are strongly discouraged.-   ------------------------------   End of INFO-VAX 2001.149 ************************