1 INFO-VAX	Sat, 05 Apr 2003	Volume 2003 : Issue 187       Contents:' Re: Another day without VMS advertising P Re: BOF sessions at CUO/Interex/DECUS European Users Conference & Expo, May 19-2 Re: COV Sponsors Re: COV Sponsors; Re: DCL anomoly - documented, but I think it's really a bug ; Re: DCL anomoly - documented, but I think it's really a bug $ Re: DECwindows on multiple VMS hosts  Re: Disappearing CR/LF in editor  Re: Disappearing CR/LF in editor. Re: Do OpenVMS Alpha have a Year 2057 problem? Re: Fortran Guru requested# Re: HSD50 Firmware - Latest Version 3 Re: IBM bugs shut down bank!  Proof IBM is garbage! 0 Re: Include an automated date within a printout.> Re: Inquirer: HP's Alpha RetainTrust programme a complete bust> Re: Inquirer: HP's Alpha RetainTrust programme a complete bust. Re: Legato vs TSM for VMS backup pros and cons. Re: Legato vs TSM for VMS backup pros and cons' Re: OpenVMS and Security - REALLY !!!!! , Re: Palmer killed Alpha so AMD could get it! Re: pathworks administer issue Re: tcpip Programming  Re: tcpip Programming ( Re: The Space Invaders minute of the day( Re: The Space Invaders minute of the day2 Re: Trying something dangerous with cma* libraries+ Re: VMS XTERM to Sun Solaris: Keyboard Map?  We buy NEWER Alpha Systems& Re: who is procedure of backup for vms  F ----------------------------------------------------------------------  # Date: Sat, 05 Apr 2003 04:56:43 GMT # From: "John Smith" <a@nonymous.com> 0 Subject: Re: Another day without VMS advertisingH Message-ID: <fmtja.40533$7Im.33128@news04.bloor.is.net.cable.rogers.com>  D "Fred Kleinsorge" <kleinsorge@star-dot-zko-dot-dec-dot-com> wrote in0 message news:3e8dc4d6$1_2@hpb10302.boi.hp.com... > 7 > "Bill Gunshannon" <bill@cs.uofs.edu> wrote in message 1 > news:b6k23q$63tbf$2@ID-135708.news.dfncis.de... + > > In article <3E8CF609.6F36B30B@fsi.net>, 7 > > "David J. Dachtera" <djesys.nospam@fsi.net> writes:  > > > John Smith wrote: 
 > > >> [snip] > > > >> More of the same ads in today's paper. No VMS mentions. > > >>
 > > >> SS-DD. ! > > >> (same sh*t -different day)  > > > : > > > I thought that was "Single Sided, Double Density"... > > >  > > A > > I noticed an HP ad on the back cover of my latest copy of Dr.  Dobbs.? > > It is pushing Itanium Workstations.  Mentions RedHat Linux,  Windows and  > > Unix, but no VMS.  > >  > > > Uh, because we aren't currently shipping VMS on Itanium yet?    E Ten bucks says that when you are, there still won't be any mention of  VMS.   ------------------------------   Date: 4 Apr 2003 16:57:41 -0800 % From: sellis@legato.com (Shaun Ellis) Y Subject: Re: BOF sessions at CUO/Interex/DECUS European Users Conference & Expo, May 19-2 = Message-ID: <1faa9425.0304041657.308a17a9@posting.google.com>   @ I will also be in Amsterdam, and am responsible for NetWorker on< OpenVMS. Although not produced by HP, it is supported by HP.   Shaun Ellis ( Product manager for NetWorker on OpenVMS Legato, Palo Alto   j Bart.Zorn@xs4all.nl (Bart Zorn) wrote in message news:<a98cd882.0303270801.23a680ce@posting.google.com>...M > At the upcoming conference it will be possible to organize ad-hoc sessions, N > called Birds-Of-a-Feather (BOF) sessions. These sessions will generally haveH > an informal and interactive nature. Also, they will have most likely a > technical nature.  > M > I would like to organize two of such BOF sessions. I am not sure if it is a 0 > coincidence that both have similar titles :-).K > By announcing them here, I might attract the attention of people involved  > with these products. > S > Anyone who will attend the conference and is interested in, or has concerns about > > the topics, is invited to participate in these BOF sessions. >  > The topics are:  > V > TCP/IP services for OpenVMS: Success stories, Solutions, Problems, Gripes, Concerns. > R > There are a lot of sites which use TCP/IP services as their main network serviceB > for OpenVMS. Unfortunately, that is not always without problems. > Issues discussed might be: >  > - Quality  > - Cluster support  > - Functionality  >  > R > ABS/MDMS, Archive and Backup Services, and Media, Device and Management ServicesF > for OpenVMS: Success stories, Solutions, Problems, Gripes, Concerns. > Q > Although there may be other options for Backup on OpenVMS systems (e.g. Legato, P > Tapesys) ABS/MDMS is the solution of choice for many sites, because it is a HPQ > product and support also comes from HP. Unfortunately, ABS/MDMS does not always 3 > work without problems. Issues discussed might be:  >  > - Quality  > - Cluster support  > - Functionality  > ! > I hope to see you in Amsterdam!  > 
 > Regards, >  > Bart Zorn    ------------------------------  % Date: Fri, 04 Apr 2003 23:48:48 +0200 4 From: Didier Morandi <Didier.Morandi.nospam@Free.fr> Subject: Re: COV Sponsors & Message-ID: <3E8DFDC0.4080607@Free.fr>   Keith Parris wrote: o > qwqwqwqw70@hotmail.com (Jon Power) wrote in message news:<71367ac8.0304030450.27195c3e@posting.google.com>...  >  >> * VMS people are expensive 5 >> * VMS people are difficult to find in an emergency  >  > F > You're out of touch here.  This was true up through the Y2K efforts,G > but VMS expertise is readily available today, and with the surplus of 0 > talent, prices have correspondingly come down.  P You do not know what you are talking about as far as Europe is concerned, Keith.   D.   ------------------------------  # Date: Sat, 05 Apr 2003 04:54:50 GMT # From: "John Smith" <a@nonymous.com>  Subject: Re: COV Sponsors G Message-ID: <uktja.40521$7Im.9920@news04.bloor.is.net.cable.rogers.com>   5 "Bill Gunshannon" <bill@cs.uofs.edu> wrote in message / news:b6kip6$62907$1@ID-135708.news.dfncis.de... . > In article <3e8dc401_2@hpb10302.boi.hp.com>,E > "Fred Kleinsorge" <kleinsorge@star-dot-zko-dot-dec-dot-com> writes: C > > Frankly, there isn't anything to argue about.  If you need what  NSK offers, C > > it is the only real source.  But it doesn't solve all problems, 	 just like . > > OpenVMS isn't the solution for everything. > C > Heretic! Heretic!  Somebody get the fire going aroudn the stake!!  :-)    I like my steaks medium. :-)   ------------------------------  $ Date: Fri, 4 Apr 2003 17:56:29 -0500 From: norm.raphael@metso.comD Subject: Re: DCL anomoly - documented, but I think it's really a bug? Message-ID: <OF4EEB2333.C843AC3E-ON85256CFE.005F62CA@metso.com>   K Thanks for your comments.  I have tried again, below, to articulate what it  isB that is troubling me.  Please note that I said up front that these behaviors AFAIK G are all documented.  I just found them not to be consistent with what I 
 would have	 expected.  -Norm   J From:  Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> > on 04/04/2003        10:46 AM   E Please respond to Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com>    To:    Info-VAX@Mvb.Saic.Com cc:   G Subject:    Re: DCL anomoly - documented, but I think it's really a bug     < > > DCL anomoly - documented, but I think it's really a bug. > > . > > I'm just not sure exactly what the bug is. > > : > > This shows that DCL cannot tell the difference betweenA > > a string of digits and an integer, as documented in F$TYPE().   I > I don't think your example shows this, especially since you didn't post  > the output of F$TYPE.    $ x=10000001 $ write sys$output f$type(x) INTEGER % $ write sys$output f$fao("x = !UL",x)  x = 10000001  , [Integer, sees integer, outputs "correctly"]   $ x="10000001" $ write sys$output f$type(x) INTEGER % $ write sys$output f$fao("x = !UL",x)  x = 2147139612  = [String, sees integer (as documented), outputs "incorrectly"]    $ x=f$integer(x) $ write sys$output f$type(x) INTEGER % $ write sys$output f$fao("x = !UL",x)  x = 10000001  @ [String converted to Integer, sees integer, outputs "correctly"]  % $ write sys$output f$fao("x = !AS",x) 3 %SYSTEM-F-ACCVIO, access violation, reason mask=00, B virtual address=0000000000038248, PC=0000000000110001, PS=7FFABF50  K [Integer, attempt to output as String, Access violation results (expected)]    $ x=10000001 $ show symbol x 4   X = 10000001   Hex = 00989681  Octal = 00046113201 $ x=f$integer(x) $ show symbol x 4   X = 10000001   Hex = 00989681  Octal = 00046113201% $ write sys$output f$fao("x = !UL",x)  x = 10000001  B [Integer, converted to Integer, sees integer, outputs "correctly"]C I would have expected this to try to treat the argument as a string F and give a different result.  I am glad it does not, but I still think something is inconsistent.  D Let me say it again.  SHOW SYMBOL knows if the symbol is stored as aE STRING or as an INTEGER, but F$TYPE(x) cannot tell a string-of-digits I is a string and returns INTEGER.  In that case F$FAO("!UL",x) will return C the binary value of the bits in the STRING that F$TYPE(x) thinks is A an INTEGER, but F$INTEGER(x)will return a converted value if x is D really a STRING, and an unconverted value if x is really an INTEGER,D while F$FAO("!UL",x) will return the binary-to-decimal conversion ofA the bits in the variable x, which is another quantity altogether. E (F$FAO("!AS",x) will access-violate if x is not a displayable STRING, = complicating the discussion, but beside the immediate point.)   C I agree that this behavior of F$TYPE is documented, I just find the - logic of the above not to be what I expected.   " Q.  When is a string not a string?  A.  When its a string of digits.  I > What you are seeing are DOCUMENTED implicit conversions in comparisons.  > 3 > There are more of these goodies, also documented:  > 3 > $  write sys$output f$integer("this is a string")  > 1  > A > > Changing a string of digits into an integer give the expected > > > integer result, but you can also change an integer into an7 > > integer and get the original result back unchanged.  > H > What does "unchanged" mean: as seen by F$TYPE or as seen in the result > of a comparison? >   G I hope the above explains my confusion.  The only point I was trying to G add with the comparisons is that they are no help determining string or @ integer as apparently (and as apparently documented) they do the conversions F behind the curtain and at least for the equal case, always give equal.    ? > > Yet if you display a string that is digits with the F$FAO() A > > directive "!UL" the result is the binary value of this string  > > converted to decimal,  > I > I think that this shows that DCL "knows" what is an integer and what is H > a string; the conversions in comparisons are just done temporarily, so > to speak.  >   @ I do not understand this claim.  If x = "10000001" is treated by@ F$INTEGER(x) as 10000001 and by F$FAO("!UL",x) as 2147139612 and1 by F$TYPE(x) as INTEGER, then something is amiss.   @ > > [I found this in autogen.com (reported) where it reported my< > > pagefile size should be 2147139620 instead of 10000001.] > I > If AUTOGEN is telling you something wrong, then it is a bug in AUTOGEN. @ > Perhaps the author wasn't sufficiently familiar with DCL.  :-) >   D Granted.  I have opened that case, they have already sent me one newE version that does not fix the problem, and I have "fixed" the problem H by adding x=f$integer(x) statements - which as shown above, seem to haveK no effect on integers and convert strings-of-digits to integers "correctly" K so that the F$FAO() directive in the GOSUB routine works.  <smiley deleted>   H > > Notice also that both string and integer comparisons result in sting( > > and/or integer values testing equal. > + > Expected due to the implicit conversions.  > > > > I'm not sure what I expected, but this seems inconsistent.< > > The test for string really should detect a string, and a > ; > The test is F$TYPE; you didn't post any examples of that.   B > > string really should not be both .eq. and .eqs. to an integer.  5 > This is due to the documented explicit conversions.    ------------------------------  $ Date: Fri, 4 Apr 2003 17:55:59 -0500 From: norm.raphael@metso.comD Subject: Re: DCL anomoly - documented, but I think it's really a bug? Message-ID: <OF2417A693.6C2192B8-ON85256CFE.007C7AAD@metso.com>   C From:  John Laird <john@laird-towers.org.uk> on 04/04/2003 01:08 PM   7 Please respond to John Laird <john@laird-towers.org.uk>    To:    Info-VAX@Mvb.Saic.Com cc:   G Subject:    Re: DCL anomoly - documented, but I think it's really a bug     @ On Fri, 4 Apr 2003 12:16:24 -0500, norm.raphael@metso.com wrote:  I >Yes, I know.  the problem is that in autogen.com the variable here shows  as >strJ >is supplied as  a generic variable to a gosub-routine which does not know >that < >it is trying to display a string as an integer using "!UL". > H >If you change that to "!AS" and supply the variable when it contains an >integer, thenH >F$FAO() will probably access-violate.  It certainly will not do what is >wanted. > J >If the content of the variable is always one thing or the other, there is >no problem toD >solve; here we could get either and cannot easily test which we are	 >getting.  >  >Is that clearer?   J Sorry if I misunderstood your original post.  You cannot expect F$FAO (andJ the underlying SYS$FAO service) to be smart - if you state "interpret thisH variable as a longword integer" by using !UL, then you have to pass it aK longword integer.  If instead you pass it a string descriptor, then you see  an address in the output.    ->K  To be clear here, I am not passing it "a string descriptor" but the symbol J  that contains a string-of-digits, which should have contained an integer.J  (Of course, I am not passing it anything, autogen is doing that, but that  is not the point.)  <-  J If F$TYPE(str) returns INTEGER, then str2='str' will yield a truly integer str2 symbol.   ->G   I don't think so.  That's one of the "problems."  F$TYPE(str) returns E   INTEGER when str contains a string-of-digits STRING as well as when C   str contains an INTEGER (each as I determine by SHOW SYMBOL str). C   In the first case str2='str' is still a string-of-numbers string. 2   ***  The fix is to do str2=f$integer('str')  *** <-  ; Chuck that at F$FAO instead.  I briefly browsed AUTOGEN.COM * and it seems to do this in several places.  F What you are describing seems to be a bug in AUTOGEN, where one of theH subroutines appears to have slipped through the net.  It is not a bug in DCL H nor F$TYPE nor F$FAO.  Unless I still can't see your problem correctly !   ->B   Yes, the original real bug is in AUTOGEN and is being addressed.G   My problem, and it is mine, I guess is that as I meant to post before    and will after this post: @     If x = "10000001" is treated by F$INTEGER(x) as 10000001 andD                                  by F$FAO("!UL",x) as 2147139612 and>                                  by F$TYPE(x) as INTEGER, then     something is amiss.  <-                John    ------------------------------  $ Date: Fri, 4 Apr 2003 14:39:47 -0500; From: "Brian Tillman" <Brian.Tillman@notnoone.notnohow.com> - Subject: Re: DECwindows on multiple VMS hosts $ Message-ID: <3e8ddf7c$1@news.si.com>  G >I would like to start DECterms on other nodes without the intermediate 	 >decterm.   I It's easy.  On each node where you want a DECterm client to run, create a  command procedure like this:  
 $! launch.com . $ set display/create/node='p1'/transport=tcpip1 $ create/terminal=decterm/detach/window=("''p2'")   4 Then, on your PC, use an RSH or RXEC command to run:  K @yourdir:launch yourip "title=ME,icon=ME,x_pos=somenumber,y_pos=somenumber"   L where "yourip" is your PC's ip address, and "somenumber" is the value of theI X or Y cooordinates of where you'd like to place the upper left corner of 
 your DECterm.  --  I Brian Tillman         Internet: Brian.Tillman at smiths-aerospace dot com 5 Smiths Aerospace  Addresses modified to prevent SPAM. @ 3290 Patterson Ave. SE, MS Replace "at" with "@", "dot" with "." Grand Rapids, MI 49512-1991 8        This opinion doesn't represent that of my company   ------------------------------   Date: 4 Apr 2003 15:07:59 -0800 . From: spamsink2001@yahoo.com (Alan E. Feldman)) Subject: Re: Disappearing CR/LF in editor = Message-ID: <b096a4ee.0304041507.4572b8e7@posting.google.com>   g JF Mezei <jfmezei.spamnot@vl.videotron.ca> wrote in message news:<3E8DB8E4.9FB0E937@vl.videotron.ca>...  > Poiter wrote: J > > File attributes:    Allocation: 805, Extend: 0, Global buffer count: 0( > >                     No version limitP > > Record format:      VFC, 2 byte header, maximum 100 bytes, longest 100 bytes3 > > Record attributes:  Print file carriage control  > M > OK, when TPU saves the file, I suspect it will change the attributes to the 3 > "standard" variable format without print control.  > L > As I recall, the "print file carriage control" dates from the fortran daysM > where the first byte of a record has special meaning. Some characters would O > skip a line, others skip to a new page, some would not skip a line (overwrite  > previous line)  $ That was "Fortran Carriage Control".  C EDT creates a VFC file writing sequence numbers in the fixed-length 7 control fields if you exit a session via EXIT/SEQUENCE.   F There have likely been other uses for fixed-length control fields over
 the years.  O > When you type the "raw" file, RMS automatically executes these characters and  > displays properly.  D Try typing a fixed-length records file that has no record attributesF and see if TYPE types it right. The editor *will* display it properly.  I > When you edit, you see the raw file without any interpretation of those O > control characters (in the IBM days, such control characters were visible, as 4 > I recall 1 was for skip to new page for instance).  B Well, I'd say that the editor displays one record per line (and itF doesn't display information from the fixed-length control fields), butD TYPE doesn't type base its output like that unconditionally. It usesD only the record attributes combined with any appropriate infromation> from the VFC headers, if it's a VFC file. With carriage returnD carriage control, TYPE displays one record per line. With Print file@ carriage control of VFC files it checks the fixed-length controlF fields for carriage control information. With Fortran it either checksE the fixed-width control field, or the first character of each line. I  don't remember which.   A Of course, <CR><LF> combos will also create an "end of line" when  TYPE-ed to the screen.  L > And when you save, you'd be creating a text file where the first character: > would be part of the data instead of interpreted by RMS.  A When the editor converts the file to var. format it just discards F anything in the fixed-width control fields (the 2-byte headers). (WhatE else can it do? You don't want to save "random" ASCII characters in a  readable file, now, would you?)    > N > One way around this after editing would be to set the file'sattributes back. >  > HELP SET FILE/ATTRIB > M > I think SET file chocolate.txt /ATTRIB=(rat:prn) might set it back to where M > the first character is interpreted for printing. However, I am not sure you   > could set RFM:VFC to the file.  A Well, since the information from the 2-byte headers (fixed-length C control fields) is lost, I don't think this would get back the same  carriage control in general.  N > Does anyone know if VFC and VAR have the same internal format ? (eg: can youN > change from one to another with set file/attrib and still have readable file0 > with just how it is handled being different ?)  E The answer is yes and no. Try it yourself! Make use of the DUMP/BLOCK  command and see for yourself.    Disclaimer: JMHO Alan E. Feldman    ------------------------------  $ Date: Fri, 4 Apr 2003 18:41:47 -0500 From: norm.raphael@metso.com) Subject: Re: Disappearing CR/LF in editor ? Message-ID: <OF7B5BA7AC.88FB02E6-ON85256CFE.0081B785@metso.com>   H There is a (basic) program available on DSNLink that reads PRN files andI puts the print control characters into the records as it copies the file. G That means for example, a <FF> character gets added at the front of the  recordI that has the form-feed code in the before-printing part of the FC part of  the record. H Comments in this program, and one of the manuals, completely map out the meaning of these two bytes.   K I can post that part later if it's wanted.  After this is run, you can edit  to your heart's H content, assuming you want to print the results or (in my case) transfer the file to . a platform that does not understand PRN files.      F From:  spamsink2001@yahoo.com (Alan E. Feldman) on 04/04/2003 06:07 PM  : Please respond to spamsink2001@yahoo.com (Alan E. Feldman)   To:    Info-VAX@Mvb.Saic.Com cc:   , Subject:    Re: Disappearing CR/LF in editor    ; JF Mezei <jfmezei.spamnot@vl.videotron.ca> wrote in message + news:<3E8DB8E4.9FB0E937@vl.videotron.ca>...  > Poiter wrote: J > > File attributes:    Allocation: 805, Extend: 0, Global buffer count: 0( > >                     No version limitJ > > Record format:      VFC, 2 byte header, maximum 100 bytes, longest 100 bytes 3 > > Record attributes:  Print file carriage control  > I > OK, when TPU saves the file, I suspect it will change the attributes to  the 3 > "standard" variable format without print control.  > G > As I recall, the "print file carriage control" dates from the fortran  daysG > where the first byte of a record has special meaning. Some characters  would D > skip a line, others skip to a new page, some would not skip a line
 (overwrite > previous line)  $ That was "Fortran Carriage Control".  C EDT creates a VFC file writing sequence numbers in the fixed-length 7 control fields if you exit a session via EXIT/SEQUENCE.   F There have likely been other uses for fixed-length control fields over
 the years.  K > When you type the "raw" file, RMS automatically executes these characters  and  > displays properly.  D Try typing a fixed-length records file that has no record attributesF and see if TYPE types it right. The editor *will* display it properly.  I > When you edit, you see the raw file without any interpretation of those C > control characters (in the IBM days, such control characters were  visible, as 4 > I recall 1 was for skip to new page for instance).  B Well, I'd say that the editor displays one record per line (and itF doesn't display information from the fixed-length control fields), butD TYPE doesn't type base its output like that unconditionally. It usesD only the record attributes combined with any appropriate infromation> from the VFC headers, if it's a VFC file. With carriage returnD carriage control, TYPE displays one record per line. With Print file@ carriage control of VFC files it checks the fixed-length controlF fields for carriage control information. With Fortran it either checksE the fixed-width control field, or the first character of each line. I  don't remember which.   A Of course, <CR><LF> combos will also create an "end of line" when  TYPE-ed to the screen.  B > And when you save, you'd be creating a text file where the first	 character : > would be part of the data instead of interpreted by RMS.  A When the editor converts the file to var. format it just discards F anything in the fixed-width control fields (the 2-byte headers). (WhatE else can it do? You don't want to save "random" ASCII characters in a  readable file, now, would you?)    > H > One way around this after editing would be to set the file'sattributes back.  >  > HELP SET FILE/ATTRIB > G > I think SET file chocolate.txt /ATTRIB=(rat:prn) might set it back to  where I > the first character is interpreted for printing. However, I am not sure  you   > could set RFM:VFC to the file.  A Well, since the information from the 2-byte headers (fixed-length C control fields) is lost, I don't think this would get back the same  carriage control in general.  J > Does anyone know if VFC and VAR have the same internal format ? (eg: can you I > change from one to another with set file/attrib and still have readable  file0 > with just how it is handled being different ?)  E The answer is yes and no. Try it yourself! Make use of the DUMP/BLOCK  command and see for yourself.    Disclaimer: JMHO Alan E. Feldman    ------------------------------  % Date: Fri, 04 Apr 2003 22:52:27 -0000 - From: wspencer@ap.nospam.org (Warren Spencer) 7 Subject: Re: Do OpenVMS Alpha have a Year 2057 problem? 5 Message-ID: <9353B0A49warrenspencer1977@216.168.3.30>   * win@fom.fgan.de (Rudolf Wingert) wrote in / <MDEJJFGEEOPAFONJONBKOEPNCOAA.win@fom.fgan.de>:    >Hello,  > K >today I did show an OpenVMS novize the SET TIME command. Within the HELP I G >did see that a legal year specification for Alpha is 1957 to 2056 (for F >OpenVMS vax from 1858 to 9999). In case of this, my question is, doesI >OpenVMS AXP have a year 2057 problem? Or why do I have this restriction?   H I suspect the C runtime library will have problems - but IIRC that's in  2032.    ws   --     Warren Spencer' Senior Software Engineer (not a writer)  The Associated Press  + ** What's brown and sticky?    A stick.  **    ------------------------------  % Date: Fri, 04 Apr 2003 23:40:19 +0200 4 From: Didier Morandi <Didier.Morandi.nospam@Free.fr># Subject: Re: Fortran Guru requested & Message-ID: <3E8DFBC3.8020404@Free.fr>   Brian Tillman wrote:- >>I've compiled the source Didier posted on :  >  > M > I've looked at all the posts from Didier on my news server and I can't find . > any source, except for a couple of excerpts.   2-apr-2003 08:38  K (excuse me to be a bit short, but I drove from Birmingham (UK) to Toulouse  L because of some Air France strikes and I'm not very motivated to answer all  posts in here now :-)    D.   ------------------------------  $ Date: Fri, 4 Apr 2003 14:54:29 -0500& From: "Bob  Lail" <robert.lail@hp.com>, Subject: Re: HSD50 Firmware - Latest Version* Message-ID: <3e8de55a@usenet01.boi.hp.com>   Gary  G All the notes I can find indicate that 5.4D and V5.7D are valid for all I versions of OpenVMS  from the minimum V6.2-1H3 Alpha/V6.2 VAX  to current J the current releases. You would only need to upgrade if you are having anyH of the problems fixed in V5.7D.  At this point the ONLY way to get V5.7D" appears to be through HP Services.  	 \Bob Lail    -- Robert Lail  Solution Architect  Network & Service Provider Sales Hewlett-Packard Company 0 EMail: SRobert.Lail@hp.com   Phone: 603.315.05563 ( S added for Spammers remove before sending email)   9 "Gary L. Ross" <rossgl@parknicollet.com> wrote in message 2 news:f6cp8vc0b36drg13vqdkv4vcvofbrr75qr@4ax.com...E > On Tue, 1 Apr 2003 15:50:52 -0500, "Bob  Lail" <robert.lail@hp.com>  > wrote: > L > >No, V5.7D is the latest and last.  V5.4D and V5.7D  require OpenVMS Alpha8 > >V6.2-1H3 or OpenVMS VAX V6.2 as the minimum OS level. > >  > >  > >\Bob Lail > E > What version of firmware do I need to have to run VMS 7.2-2??  Will G > 5.4 work (since that is what we have)?  If not where can I locate/buy F > the correct firmware revision to allow us to run VMS 7.2-2.  Thanks. >  > Gary L. Ross > Park Nicollet Health Services  > rossgl@parknicollet.com  >    ------------------------------  $ Date: Fri, 4 Apr 2003 16:43:53 -0500$ From: "rob kas" <rob@netcarrier.net>< Subject: Re: IBM bugs shut down bank!  Proof IBM is garbage!/ Message-ID: <v8rvntt67ggffe@corp.supernews.com>   B         mmmm So RDB has never had any bugs ever...................  K        IBM makes some great gear and the Power 5 Stuff looks interesting to  say the least.    3                                                 Rob K                                    "Who's 32 bit apps are screaming fast on  his Opteron"    5 "Bob Ceculski" <bob@instantwhip.com> wrote in message 6 news:d7791aa1.0304041009.e692b8e@posting.google.com...5 > this is on the inquirer if you want to go there ... . > I said IBM systems and software are junk and1 > this proves it ... and Cerner would want to put 4 > hospitals on DB2 garbage systems over VMS? only if > they want people to die ...  >  > 3 > Big Bank's Big Blue system failed due to DB2 bugs  > $ > Updated Payments, trading affected > , > By Mike Magee: Friday 04 April 2003, 11:04 > , > None of this shit works -- William Shatner9 > But, you can make a Starship Enterprise out of a floppy B > A REMARKABLE STATEMENT on operational problems at the big DanishD > Danske Bank has unravelled a skein of problems with IBM's flagship > relational database, DB2.* > B > In a report released yesterday, the bank claimed that IBM's techG > division was replacing a defective unit in an RVA disk system causing 0 > outage at two of the Bank's operating centres. > @ > But the replacement, said the bank, has unveiled "never before? > identified errors" in DB2 which caused what Danske said was a # > "critical operational situation".  > > > The bank said that these problems were "significant" for itsC > customers, an affected companies including Realkredit Danmark and  > Danica Pension.  > G > After what the bank described as a breakdown, business data stored on C > the RVA disk was inaccessible and securities and currency trading A > couldn't work. After the disk was re-installed, a batch run was - > started but they weren't running correctly.  > F > And, claimed the bank, a software error in DB2 database software hasG > existed in similar installs since 1997, without IBM knowing about it.  > D > Subsequently, the bank discovered three further errors in DB2, theC > second being that recovery process on database tables couldn't be B > started, while another error caused recovery jobs from being run > simultaneously.  > F > Another bug stopped recovery jobs from restoring data to the tables. > B > IBM had no patches for the problems at the time, but now has, it
 > appears. > F > It says it is discussing the problems with IBM, but reckons that theG > problems are not unique to itself. Luckily, Danske Bank had duplicate : > and back up services, meaning it could recover the data. > E > * AND A READER WRITES: The flaw in the IBM software at Danske Bank, H > that lasted almost two weeks, meant that customers in Denmark couldn'tG > use their homebank systems for a week, and many customers, home users C > and business users couldn't pay their bills or trade on the stock G > market. Because it is the biggest bank in Denmark, other banks had to D > cover for it, even the "National Banken" which is the Danish stateE > bank (don't know the word in English). Danske Bank has announced in B > Danish Radio that it will have to spend up to 100 million danishF > kroner (approx 13 million Euro) on the system, so there won't be anyA > problems in the future. It did cause quite a stir in the Danish  > financial market.    ------------------------------  $ Date: Fri, 4 Apr 2003 14:52:36 -0500; From: "Brian Tillman" <Brian.Tillman@notnoone.notnohow.com> 9 Subject: Re: Include an automated date within a printout. $ Message-ID: <3e8de27d$1@news.si.com>   >> How about $ PRINT/HEADER  >> ? >RobE >This will just produce a header page prior to printing of the actual  >file.  ) Huh?  When was the last time you used it?    $ help print/header    PRINT   	   /HEADER            /HEADER          /NOHEADER (default)         Positional qualifier.  B      Controls whether a heading line is printed at the top of each
      page.  L That said, some network symbionts (like the one we use, ScriptServer), don't honor the /HEADER qualifier. --  I Brian Tillman         Internet: Brian.Tillman at smiths-aerospace dot com 5 Smiths Aerospace  Addresses modified to prevent SPAM. @ 3290 Patterson Ave. SE, MS Replace "at" with "@", "dot" with "." Grand Rapids, MI 49512-1991 8        This opinion doesn't represent that of my company   ------------------------------  $ Date: Fri, 4 Apr 2003 15:00:48 -0500& From: "Bob  Lail" <robert.lail@hp.com>G Subject: Re: Inquirer: HP's Alpha RetainTrust programme a complete bust * Message-ID: <3e8de6d3@usenet01.boi.hp.com>   John  J No, INTEL is NOT paying for the OpenVMS port to Itanium. It's HP's dollarsI all the way. I have seen this FUD coming from folks in this newsgroup for  far to long now.  	 \Bob Lail    -- Robert Lail  Solution Architect  Network & Service Provider Sales Hewlett-Packard Company 0 EMail: SRobert.Lail@hp.com   Phone: 603.325.05563 ( S added for Spammers remove before sending email)   . "John Smith" <a@nonymous.com> wrote in messageB news:Bqija.37622$pNv.20975@news02.bloor.is.net.cable.rogers.com... > B > "Bob Koehler" <koehler@eisner.nospam.encompasserve.org> wrote in7 > message news:Cx3RU5DdkrSz@eisner.encompasserve.org...uB > > In article <v8otgtbtat4ued@corp.supernews.com>, "Lee Courtney"2 > <lcourtney@-REMOVETHISFILTER-mvista.com> writes:C > > > I have been lurking in this news group for a couple years andC > observe thatF > > > this sounds like the HP3000 and MPE/iX a few years ago. Where is
 > MPE now? > > >l > >uB > >    I forgot to add, VMS is making enough profit to justify the > comparably> > >    small cost of porting it.  MPE was not.  HP isn't dumb. >r >eD > The cost of porting is being paid by Intel, so it's not HP's money > that's being spent.b >!H > Why isn't HP using 'their' money to advertise and grow the VMS market?E > Is the money not being spent so that carly can show a better bottomn2 > line today so her personal bonus will be bigger? >tE > Bottom line is that for many years VMS has generated profits yet it F > has not had advertising/marketing money spent on it in proportion to > either sales or profits. >dG > Doubling the current small rate of VMS/Alphaserver sales would show ai= > huge impact on cost of production and profits, and could beh > accomplished for a tinytD > fraction of the advertising budget spent on increasing HP's Wintel > sales by 1%. >  >b   ------------------------------  % Date: Fri, 04 Apr 2003 12:04:21 -0800v9 From: "gregc at gregcagle.com" <"gregc at gregcagle.com">-G Subject: Re: Inquirer: HP's Alpha RetainTrust programme a complete bustt/ Message-ID: <v8rpaimavirr8d@corp.supernews.com>t  @ I think people are confusing things with Intel's funding a chunk@ of the Itanium design effort. You're right - Intel is not paying for any HP OS port to Itanium.   - Greg   Bob Lail wrote:p > John > L > No, INTEL is NOT paying for the OpenVMS port to Itanium. It's HP's dollarsK > all the way. I have seen this FUD coming from folks in this newsgroup for0 > far to long now. >  > \Bob Lail  >  > --
 > Robert Lailr > Solution Architect" > Network & Service Provider Sales > Hewlett-Packard Companyy2 > EMail: SRobert.Lail@hp.com   Phone: 603.325.05565 > ( S added for Spammers remove before sending email)u > 0 > "John Smith" <a@nonymous.com> wrote in messageD > news:Bqija.37622$pNv.20975@news02.bloor.is.net.cable.rogers.com... > B >>"Bob Koehler" <koehler@eisner.nospam.encompasserve.org> wrote in7 >>message news:Cx3RU5DdkrSz@eisner.encompasserve.org...  >>A >>>In article <v8otgtbtat4ued@corp.supernews.com>, "Lee Courtney"u >>2 >><lcourtney@-REMOVETHISFILTER-mvista.com> writes: >>A >>>>I have been lurking in this news group for a couple years and: >> >>observe that >>D >>>>this sounds like the HP3000 and MPE/iX a few years ago. Where is >>
 >>MPE now? >>A >>>   I forgot to add, VMS is making enough profit to justify the8 >> >>comparably >>= >>>   small cost of porting it.  MPE was not.  HP isn't dumb.W >> >>D >>The cost of porting is being paid by Intel, so it's not HP's money >>that's being spent.e >>H >>Why isn't HP using 'their' money to advertise and grow the VMS market?E >>Is the money not being spent so that carly can show a better bottomc2 >>line today so her personal bonus will be bigger? >>E >>Bottom line is that for many years VMS has generated profits yet itoF >>has not had advertising/marketing money spent on it in proportion to >>either sales or profits. >>G >>Doubling the current small rate of VMS/Alphaserver sales would show ah= >>huge impact on cost of production and profits, and could bet >>accomplished for a tiny D >>fraction of the advertising budget spent on increasing HP's Wintel >>sales by 1%. >> >> >  >  >    -- e
 Greg Cagle gregc at gregcagle dot com   ------------------------------   Date: 4 Apr 2003 16:28:31 -0800t% From: sellis@legato.com (Shaun Ellis)h7 Subject: Re: Legato vs TSM for VMS backup pros and conss= Message-ID: <1faa9425.0304041628.63b9b22c@posting.google.com>r  D It is kind of Veritas to let me know how they are doing with OpenVMS :-)o  0 I'll answer the points, and make some of my own.  8 > The VERITAS NBU OpenVMS client will support the backup6 > and restore of a bootable OpenVMS system disk in the1 > next v4.5 MP4 (Maintenance Pack Four) release:-e > > >  o  This release will backup and restore a bootable OpenVMS 9 >     system disk.  All alias directories, including the tA >     SYSCOMMON.DIR directory which points to the VMS$COMMON.DIR ): >     directory, are now backed up and restored correctly.9 >     System disk files that are required to be restored t8 >     contiguously are now restored as contiguous files.* >     See Section 12 System Disk Backups.   ; Will do. Ah well Legato does it NOW. Fully supported TODAY.    < >  o  Where possible this release will restore all files as : >     contiguous files. A complete restore of a disk will - >     effectively defragment the entire disk.s  B Where possible? Hmm, of course a complete restore defrag's a disk.B Having said that, most people concerned about this issue now use aB disk defragger. No benefit to Veritas anyway. any restore from any product will do this.V  p; >  o  This release provides full support for OpenVMS ODS-5 n2 >     file systems (this is available now in MP3). > 7 > I will also point out that VERITAS NBU OpenVMS clients> > backups are very fast, especially over a gigabit connection. > = > Before you decide you may want to compare the backup times u? > of VERITAS vs Legato (even when they are backing up to local  	 > tape). -  B Please tell us EXACTLY what times you have seen on a customer siteD with lots of small files all with ACL's. We are getting 120Gb per hr9 across 1Gb network. This is real data at a real customer.   C NetWorker has a CLI and user level Motif GUI for backup and restore2; that uses the OpenVMS file syntax. When reading the VeritashB documentation, for restore they have CLI only, and have to use theF syntax of what ever the Master Server is. (this is taken from examples in their own documentation)f  B NetWorker also includes Archiving capabilities, saveset and volumeD cloning, and we have a Storage Node - the ability to backup directlyC from disk to disk, or tape. Also included is the ability to stage - A backup from disk to disk, and then automatically move it to tape.-  F With the Storage Node capability, it means we can share tape librariesE and tape devices with other OS'. This means you can achieve excellent E throughput, without going across the network, and maximise the ROI onn< your tape devices. Something you cannot do with veritas NBU.  B HP supports and resells NetWorker, including NetWorker on OpenVMS.  E This year you will also see support for AlphaStor, our advanced mediat: management product, and modules for Oracle and Oracle RDB.  B Legato is fully committed to the OpenVMS platform. I finish with a. quote from Mark Gorham, the HP VP for OpenVMS.  A "HP OpenVMS customers require industry-leading storage management-< solution options that support both direct attach and complexC heterogeneous fibre-based SAN environments," said Mark Gorham, viced= president, HP OpenVMS Software Group. "We're pleased that ourx> customers will have a robust solution available to them from aD reliable industry leader such as LEGATO that can help reduce network= traffic and improve overall backup and recovery performance."i  ; If you have any questions, please e-mail openvms@legato.com    Shaun Ellis % NetWorker for OpenVMS Product Manager  Legato, Palo Altof   ------------------------------   Date: 4 Apr 2003 16:38:41 -0800c% From: sellis@legato.com (Shaun Ellis)y7 Subject: Re: Legato vs TSM for VMS backup pros and cons = Message-ID: <1faa9425.0304041638.674d6acd@posting.google.com>t   Alan,u  F this is a good point. There are many notes in usenet conferences whereD I've seen instructions on how to build a bootable CD for OpenVMS. If+ you can't do that, there are other options:c  ? 1) Boot the machine as a remote cluster client into an existingVA cluster and get VMS that way. You can the restore the system disk3  A 2) Have a spare disk that has OpenVMS and the desired backup tool ; built on it. Boot that, and again, restore the system disk.0  E Obviously, with a NetWorker Storage Node, in both scenarios, the disk1C can access local tape, or disk, devices instead of going across the  network.   regards    Shaun Ellisp! NetWorker OpenVMS product manager1 Legato, Palo Altor  \ Alan Frisbie <Abuse@NelsonUSA.com> wrote in message news:<3E8B625D.2010707@NelsonUSA.com>... > Alan Fay wrote:r: > > The VERITAS NBU OpenVMS client will support the backup8 > > and restore of a bootable OpenVMS system disk in the3 > > next v4.5 MP4 (Maintenance Pack Four) release:-  > > @ > >  o  This release will backup and restore a bootable OpenVMS ; > >     system disk.  All alias directories, including the UC > >     SYSCOMMON.DIR directory which points to the VMS$COMMON.DIR $< > >     directory, are now backed up and restored correctly.; > >     System disk files that are required to be restored  : > >     contiguously are now restored as contiguous files. > < > Is there a bootable CD which will allow us to quickly do a= > "bare metal" restore?   That is a key item.   Anything lessi > is not nearly enough.  >  > Alan   ------------------------------  $ Date: Fri, 4 Apr 2003 14:28:00 -0500; From: "Brian Tillman" <Brian.Tillman@notnoone.notnohow.com> 0 Subject: Re: OpenVMS and Security - REALLY !!!!!$ Message-ID: <3e8ddcb9$1@news.si.com>  + >REFRESH the page ! It happened to me too !   L No matter how many times one refreshes a page, if it's written strictly with: javascript and one had disabled javascript, it won't show. -- wI Brian Tillman         Internet: Brian.Tillman at smiths-aerospace dot comr5 Smiths Aerospace  Addresses modified to prevent SPAM.n@ 3290 Patterson Ave. SE, MS Replace "at" with "@", "dot" with "." Grand Rapids, MI 49512-1991l8        This opinion doesn't represent that of my company   ------------------------------  # Date: Sat, 05 Apr 2003 00:03:15 GMTt1 From: Michael Austin <maustin@firstdbasource.com>a5 Subject: Re: Palmer killed Alpha so AMD could get it! 2 Message-ID: <3E8E0FBB.82EDD04C@firstdbasource.com>   Bob Ceculski wrote:c > 6 > it all makes sense now!  this is on the inquirer ... > , > Microsoft's Dave Cutler enthuses about AMD >  > On the track of a Tomato > 0 > By INQUIRER staff: Friday 04 April 2003, 09:21 > G > LEST YOU FORGET that the Microsoft insider Tomato leaked the INQUIRER B > and c't magazine top information about how senior "distinguishedD > engineer" Dave Cutler enthused about AMD's chips a year back, we'd. > like to refer you to the following articles." > AMD-Microsoft Hammer memo leaked6 > AMD's Heye hunts down Tomato to make his pips squeak$ > AMD, Microsoft firm up 64-bit deal > F > And just in case you had any doubt at all how much Dave Cutler lovesF > Dirk Meyer's Alpha Hammer technology, we'd like to refer you to this > AMD page.0 > 1 > and the videotaped Microsoft statement therein.n > C > Keep trying if you can't get it the first time. Dave and Dirk, ofyF > course, worked together at DEC... Bob Palmer was the CEO of DEC. Bob/ > Palmer is now a director of AMD. Go figure...o  B are you sure Palmer wasn't brought into AMD to kill it too?  Every@ company he has touched since the 80's has crashed and burned.... -- t Regards,  6 Michael Austin            OpenVMS User since June 19847 First DBA Source, Inc.    Registered Linux User #261163l   ------------------------------  % Date: Fri, 04 Apr 2003 11:57:00 +0200s6 From: Serge ZANGHERI <szangheri@grenoble.sema.slb.com>' Subject: Re: pathworks administer issues5 Message-ID: <3E8D56EB.7F9D4D1E@grenoble.sema.slb.com>s   Hi,c Finally thanx for this problem.nJ Actually, I found that the DECNET name was different of the content of the SCSNODE.@ The decnet configuration was bad.A mc ncp sh k node didn't work.J There was no error with net error. Net commands worked but I didn't try toA connect with this command since it works with a good decnet name.5   Thanx againm Serger   "serge.zangheri" a crit :   > Hi,o, > I installed Pathworks for the second time.5 > But when I start admin I got the following answer :m- > Cannot continue - the server is not runninge >oH > I can connect pwrk$root:netlogon etc, on my pc, I can do admin/config. > No problem about licence.f > swapfile put to 150000 > pagefile put to 300000
 > 80Mb RAM% > other system paramters semmes good.  >n > Someone could help ? > Thanks > Sergee   ------------------------------  $ Date: Fri, 4 Apr 2003 14:42:38 -0500; From: "Brian Tillman" <Brian.Tillman@notnoone.notnohow.com>v Subject: Re: tcpip Programming$ Message-ID: <3e8de027$1@news.si.com>  I >> I don't know if it is the right newsgroup, so i explain what i want toA do.dL >> I would like to write a programm to download http files from www sites in5 >> the backgroud, without surfing over a web Browser.w >eI >the VMS OSU Web Server comes with a utility called FETCH_HTTP which doesh that.  > 0 >KERMIT also has that ability. And so does Perl.  
 Lynx, too. -- hI Brian Tillman         Internet: Brian.Tillman at smiths-aerospace dot comy5 Smiths Aerospace  Addresses modified to prevent SPAM.n@ 3290 Patterson Ave. SE, MS Replace "at" with "@", "dot" with "." Grand Rapids, MI 49512-1991t8        This opinion doesn't represent that of my company   ------------------------------  # Date: Sat, 05 Apr 2003 00:00:42 GMTd1 From: Michael Austin <maustin@firstdbasource.com>  Subject: Re: tcpip Programming2 Message-ID: <3E8E0F21.6AD7BE1E@firstdbasource.com>   News Account wrote:c >  > Try WGET ... > * > http://www.antinode.org/dec/sw/wget.html >  > Don Woodward > / > "Manser" <nmanser@progis.de> wrote in messager9 > news:2178d61f.0304031627.49ee0915@posting.google.com...  > > Hello folks, > > N > > I don't know if it is the right newsgroup, so i explain what i want to do.M > > I would like to write a programm to download http files from www sites in 6 > > the backgroud, without surfing over a web Browser.; > > as a start point i have the tcpip Programming examples.g > > > > > Have someone already done this kind of tcpip Programming ?# > > I appreciate any help or hints.e > >O > > Thanks in advance.  G a PERL example that logs into a Linksys router and displays the currentn IP and DNS information:e   execute using: r $ @getip.plp& Current IP address is: aaa.bbb.ccc.ddd& DNS 1   IP address is: aaa.bbb.ccc.ddd& DNS 2   IP address is: aaa.bbb.ccc.ddd DNS 3   IP address is: 0.0.0.0   $PERL -w # getip.pl   #!/usr/bin/perl -w #!/usr/local/bin/perl -w  = ## $ENV{'PATH'} = (exists($ENV{PATH}) ? "$ENV{PATH}:" : "") .i "/sbin:/usr/sbin:/bi ## n:/usr/bin:/etc:/usr/lib:";   use Getopt::Long;  use Sys::Hostname; use IO::Socket;$       $0 = "findIP";     my $version  = "3.5.4";      my $programd  = $0;      $programd =~ s%^.*/%%;     my $program   = $programd;     $program  =~ s/d$//;     my $now       = time;r     my $hostname  = hostname();m     my ($peer, $server, $port);n$     my ($sd, $rq, $request, $reply);1     my $url = "HTTP://192.168.1.1:80/Status.htm";d       $url    =~ s%^HTTP://%%i;k     $server = $url;n     $server =~ s%/.*%%;i&     $url    = "/" unless $url =~ m%/%;     $url    =~ s%^[^/]*/%%;   &     ## determine peer and port to use.     $peer   = $server;     $peer   =~ s%/.*%%;p     $port   = $peer;     $port   =~ s%^.*:%%;)     $port   = 80 unless $port =~ /^\d+$/;      $peer   =~ s%:.*$%%;      # username:password if required.0     my $auth = encode_base64("removed:removed");     $request  = "GET ";r" #    $request .= "HTTP://$server";#     $request .= "/$url HTTP/1.0\n"; # #    $request .= "Host: $server\n";k-     $request .= "Authorization: Basic $auth";e6     $request .= "User-Agent: ${program}/${version}\n";%     $request .= "Connection: open\n";A     $request .= "\n";i  F     ## make sure newlines are <cr><lf> for some pedantic proxy servers$     ($rq = $request) =~ s/\n/\r\n/g; #        printf("$rq");o       local $^W = 0;1    $sd = IO::Socket::INET->new(PeerAddr => $peer, 3                                  PeerPort => $port,14                                  Proto    => 'tcp');      my $skip='WAN.*?IP Address';     my $skip1='DNS.*?';e"     my $result = send $sd, $rq, 0;%         if ($result != length($rq)) {i<             printf("cannot send to 192.168.1.1:80 ($!).\n");         } else {              while ($_ = <$sd>) {> #        verbose("RECEIVE:", "%s", define($_, "<undefined>"));+                 $reply .= $_ if defined $_;b            } #       printf(" $reply\n");             if ($reply =~:5 /${skip}.*?(\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3}).*/is)s {e                 $WANip = $1;
             }s             if ($reply =~a6 /${skip1}.*?(\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3}).*/is) {"                 $DNS1ip = $1;t
             }t             if ($reply =~ 7 /${DNS1ip}.*?(\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3}).*/is)l {                  $DNS2ip = $1; 
             }w             if ($reply =~r7 /${DNS2ip}.*?(\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3}).*/is)r {.                 $DNS3ip = $1;i
             }b2         printf("Current IP address is: $WANip\n");3         printf("DNS 1   IP address is: $DNS1ip\n");e3         printf("DNS 2   IP address is: $DNS2ip\n");h3         printf("DNS 3   IP address is: $DNS3ip\n");a             close($sd);a
          }       sub encode_base64 ($;$) {i     my $res = '';e     my $eol = $_[1];$     $eol = "\n" unless defined $eol;B     pos($_[0]) = 0;                          # ensure start at the	 beginninge$     while ($_[0] =~ /(.{1,45})/gs) {)         $res .= substr(pack('u', $1), 1);I         chop($res);      }e@     $res =~ tr|` -_|AA-Za-z0-9+/|;               # `# help emacs       # fix padding at the end.     my $padding = (3 - length($_[0]) % 3) % 3;8     $res =~ s/.{$padding}$/'=' x $padding/e if $padding;  H     # break encoded string into lines of no more than 76 characters each     if (length $eol) {%         $res =~ s/(.{1,76})/$1$eol/g;e     }a	     $res;t }V -- a Regards,  6 Michael Austin            OpenVMS User since June 19847 First DBA Source, Inc.    Registered Linux User #261163e   ------------------------------  % Date: Fri, 04 Apr 2003 23:57:42 +0200g4 From: Didier Morandi <Didier.Morandi.nospam@Free.fr>1 Subject: Re: The Space Invaders minute of the daye& Message-ID: <3E8DFFD6.6050801@Free.fr>   Carl Perkins wrote:F > B > There may also just be a logical name SPCINV that points to some > other version. > 
 > --- Carl     no   ------------------------------  % Date: Fri, 04 Apr 2003 23:59:07 +0200 4 From: Didier Morandi <Didier.Morandi.nospam@Free.fr>1 Subject: Re: The Space Invaders minute of the day & Message-ID: <3E8E002B.1040102@Free.fr>   Joseph Huber wrote:e > ? > One possibility not yet checked: is spcinv maybe INSTALLed  ?e  Q no. I got the source via mail, following a request here. I did a cut/paste to my  / Kea VT emulator on my Alpha and FOR, LINKed it.L   D.   ------------------------------   Date: 4 Apr 2003 12:04:58 -0800i- From: merritt.robert@spsd.sk.ca (rob merritt)r; Subject: Re: Trying something dangerous with cma* libraries = Message-ID: <b6bf97d5.0304041204.7e406d7f@posting.google.com>b  F Ok I Implemented your Ideas ,got ride of the /super and eliminated the " F still this damn image is going for syslib. Strange thing is first time after a reboot It works ?????e   $ Deflog  := define/proc/execn $!( $ Deflog CMA$TIS_SHR LIS_EXE:CMA$TIS_SHR" $ Deflog DECC$SHR LIS_EXE:DECC$SHR' $ Z3950_server:==$LIS_EXE:Z3950_UCX.exe  $! $ show log cma$tis_shr<    "CMA$TIS_SHR" = "LIS_EXE:CMA$TIS_SHR" (LNM$PROCESS_TABLE) $ show log decc$shrc6    "DECC$SHR" = "LIS_EXE:DECC$SHR" (LNM$PROCESS_TABLE) $ Z3950_server -p2103 %DCL-W-ACTIMAGE, error activating image CMA$TIS_SHRc -CLI-E-IMGNAME, image file3 ADENB$DKA0:[SYS0.SYSCOMMON.][SYSLIB]CMA$TIS_SHR.EXEO: -SYSTEM-F-SHRIDMISMAT, ident mismatch with shareable image8   SOBSERV      job terminated at  4-APR-2003 13:13:56.70     Accounting information:tF   Buffered I/O count:              28         Peak working set size:   1856E   Direct I/O count:                 8         Peak page file size:   t 41520rF   Page faults:                    177         Mounted volumes:            0A   Charged CPU time:           0 00:00:00.18   Elapsed time:     0b 00:00:00.28a        v koehler@eisner.nospam.encompasserve.org (Bob Koehler) wrote in message news:<fpvGjma2I6jX@eisner.encompasserve.org>...o > In article <b6bf97d5.0303271511.52f35239@posting.google.com>, merritt.robert@spsd.sk.ca (rob merritt) writes:sJ > > I was trying to implement this hack but can't seem to get it to work i > > define logicals like this: > > ( > > $ Deflog  := define/group/super/exec > 6 >    /super and /exec conflict, I thnk you'll get exec2 > > $ Deflog CMA$TIS_SHR "LIS_EXE:CMA$TIS_SHR.exe" >  >    leave out the quotes: > . > $ Deflog CMA$TIS_SHR LIS_EXE:CMA$TIS_SHR.exe >   >    And how is LIS_EXE defined?   ------------------------------  % Date: Fri, 04 Apr 2003 11:38:00 -0800e From: "JeffK." <uce@ftc.gov>4 Subject: Re: VMS XTERM to Sun Solaris: Keyboard Map?' Message-ID: <3E8DDF18.B2D6BF8A@ftc.gov>t  E Once I figured out it is a lot like edlin, a lot of stuff made sense.r   Edlin!   JeffK.    1 >    No.  vi is a separate program frm the shell.  > I >    But it looks like you've got some vi commands mistook.  The commands 4 >    to exit vi from command mode are the sequences: >       :wqt >    orl
 >       ZZ >    not zz.   ------------------------------  $ Date: Fri, 4 Apr 2003 21:14:24 -0500, From: "David Turner" <dbturner@islandco.com># Subject: We buy NEWER Alpha Systemsc/ Message-ID: <v8sf7k4i52mt17@news.supernews.com>   + We are currently looking for quantities of:    Alpha DS10 Systems Alpha DS20 Systems Alpha DS20e Systemsr( Alpha ES40 Systems (Model 1 and Model 2)  9 If you are ready to sell or even just thinking of sellingiJ we can offer good money (worldwide!) or trade for Intel Boxes, SUN Systems etc.  J Email us at sales@islandco.com or call USA 912 447 6622 with the equipment you have available   ------------------------------   Date: 4 Apr 2003 18:28:31 -0800 . From: spamsink2001@yahoo.com (Alan E. Feldman)/ Subject: Re: who is procedure of backup for vmsi= Message-ID: <b096a4ee.0304041828.566b051e@posting.google.com>i  v koehler@eisner.nospam.encompasserve.org (Bob Koehler) wrote in message news:<3coJ4Pxhr9MQ@eisner.encompasserve.org>...h > In article <B71ja.248$Ob4.154@news.cpqcorp.net>, hammond@not@peek.ssr.hp.com (Charlie Hammond) writes:8 > > In article <qlHmUwVw8cbb@eisner.encompasserve.org>, A > > koehler@eisner.nospam.encompasserve.org (Bob Koehler) writes:d > > ? > >>   And I always get a standalone backup after an OS update.  > > H > > {Ahem} It may well be much MORE important to get a known good backup( > > BEFORE an operating system updgrade. > G >    Gee, I guess I assumed that was obvious.  It's in the installations >    manual! > F >    The reason I get one after the OS update is so that I know I can G >    recover all critical files if an /ignore=interlock didn't pick up  G >    something I didn't know about.  I don't want to have to restore toe> >    the backup from before the update and reapply the update.  A Well, since we're getting picky here, I do recall that the VMS6.2eC upgrade manual, at least, recommends saving your work with a BACKUPeF after the installation. And I'd be surprised if it doesn't say to also& do a BACKUP just prior to the upgrade.  F I'd (and I believe the manual does too) recommend a backup both before
 and after.   Disclaimer: JMHO Alan E. Feldman    ------------------------------   End of INFO-VAX 2003.187 ************************