/ INFO-VAX	Sun, 04 Jan 2004	Volume 2004 : Issue 7       Contents: Re: Bugcheck - PCAPVCM?  end-search next sentence. RE: How to make exe return value to dcl symbol. Re: How to make exe return value to dcl symbol. Re: How to make exe return value to dcl symbol Peeking in to the serial buffer # Re: Peeking in to the serial buffer # Re: Peeking in to the serial buffer # Re: Peeking in to the serial buffer & Re: Pictures from the OpenVMS bootcamp& Re: Pictures from the OpenVMS bootcamp& Re: Pictures from the OpenVMS bootcamp  F ----------------------------------------------------------------------  % Date: Sat, 03 Jan 2004 15:53:11 -0700 4 From: Norman Lastovica <norman.lastovica@oracle.com>  Subject: Re: Bugcheck - PCAPVCM?* Message-ID: <3FF747D7.FF1D262D@oracle.com>  ? We frequently build/develop on current VMS versions even though ; we ship/support older ones.  Mostly it means keeping a copy < of SYS$LOADABLE_IMAGES, SYS$LIBRARY and SYS$SYSTEM from the ? "target" version and using them when you build.  Takes a little > setup, but then you'd be free of having to stick with antiques: and, of course, you wouldn't force anyone else to upgrade.  < Oracle Rdb is supported back to VAX on 5.5-2 and (I believe)9 6.1 on Alpha.  All development is done on V7.3 and later.    Martin Vorlaender wrote: > 7 > Norman Lastovica (norman.lastovica@oracle.com) wrote:  > > Martin Vorlaender wrote:N > > > I just had my otherwise reliable AlphaStation 200 (VMS 7.2-2 with UPDATEK > > > 1.0, ACRTL 3.0, DRIVER 3.0, and a few others; DECnet Phase IV; TCP/IP N > > > Services 5.1 ECO5) bugcheck with "PGFIPLHI, Pagefault with IPL too high"H > > > in SYS$LAN_CSMACD+08A38. As the last thing I did was test the PCAPI > > > (downloaded from simh.trailingedge.com) I have a strong feeling the K > > > guilty part is the PCAPVCM execlet. Does it really need VMS 7.3-1 (as , > > > SIMH's 0readme_ethernet.txt suggests)? > > K > >       One would have to suspect that the people that write the software K > > would be relatively familier with its limitations.  it might make sense # > > to follow the instructions, eh?  > E > It's just that I found it working before reading that it shouldn't.  > D > >  why not upgrade to vms v7.3-2 and see if it works better there? > J > Because this is the system I build distributions on, and I wouldn't want@ > to force all people running ht://Dig under VMS 7.2 to upgrade.   ------------------------------  # Date: Sun, 04 Jan 2004 01:25:18 GMT 1 From: tutor_removespam_@cfl.rr.com (_removespam_) ! Subject: end-search next sentence 8 Message-ID: <npqevvsqttp1vib1q2mirmqfbhd6d5qqjt@4ax.com>   vax allowed 	    search 
       when         next sentence 
       when
    end-search    dec / alpha will not allow it.   what is the work around?< email   tutor_nospam_@cfl.rr.com (remove the _nospam_ first)   ------------------------------   Date: 3 Jan 2004 13:08:09 -0600 - From: Kilgallen@SpamCop.net (Larry Kilgallen) 7 Subject: RE: How to make exe return value to dcl symbol 3 Message-ID: <ZpIpDOM9hJLh@eisner.encompasserve.org>   _ In article <NDEMLKKEBOIFBMJLCECIGEIACJAA.tom@kednos.com>, "Tom Linden" <tom@kednos.com> writes:  >  >  > < -----Original Message-----1 > < From: jlsue [mailto:jefflsxxxz@sbcglobal.net] , > < Sent: Saturday, January 03, 2004 8:30 AM > < To: Info-VAX@Mvb.Saic.Com ; > < Subject: Re: How to make exe return value to dcl symbol  > <  > < L > < On Fri, 26 Dec 2003 06:48:54 -0800, "Tom Linden" <tom@kednos.com> wrote: > < 5 > < >I am aware of that, but I would like a function.  > < >  > < F > < But functions in DCL are lexical functions.  If you want to learn 
 > < internals F > < and develop your own lexical, I think it's safe to say that you'll; > < definitely learn some interesting things about OpenVMS.  > L > Got a ptr to a cookbook? No reason why this should be difficult, is there?  G For starters, there is no documentation other than the source listings.    ------------------------------  $ Date: Sat, 3 Jan 2004 12:29:31 -0500) From: "Neil Rieck" <n.rieck@sympatico.ca> 7 Subject: Re: How to make exe return value to dcl symbol ; Message-ID: <%ZCJb.20264$Vl6.3754694@news20.bellglobal.com>   7 "JF Mezei" <jfmezei.spamnot@istop.com> wrote in message # news:3FE9DA07.CD3738E6@istop.com...  > Tom Linden wrote:  > > G > > Using lib$get_foreign one can pass arguments to an executable.  But J > > this behaves like a subroutine.  How would you make it a function such1 > > that it would return a value to a DCL symbol?  > N > You catch Guy Peleg and give him a beer/cookie jar/chocolate/cognac/whateverM > and then kindly ask him to implement your program as a lexical function :-)  >  > There are 2 ways about it: > J > Your program simply returns a value (for instance, in C:  exit(myvalue); and K > then right after you can executed the program, the value is stored in the ; > symbol $STATUS which you can then copy to another symbol.  > N > Another way to do it is to have the program explicitely set a symbol's valueM > with LIB$SET_SYMBOL. You could supply the symbol name as an argument to the  program.  N Most high level languages will write some value to symbol $STATUS upon exit soK trying to call LIB$SET_SYMBOL to change symbol $STATUS just before you exit M will not work. the final "END" statement will overwrite it). In Compaq-BASIC, M the statement "END" will always supply an integer value of "1" (success) as a K default. However, an alternative to "END" is "END PROGRAM  I%" (where I% in N this case is an integer variable containing the desired exit code to pass back  to DCL via $STATUS) can be used.  I Intermediate level languages like "C" operate a little differently. IIRC, J "exit" used without a parameter properly defaults to returning the integerH value of "0" (warning) to $STATUS. As stated above, "exit(myvalue)" will% return the desired exit value to DCL.   N One final point, in "C" on UNIX the integer value of "0" usually means successH while some non-zero value represents the failure message. In OpenVMS theF lowest 3 bits are reserved for severity while higher bits are used forL identification. I leave the identification bits blank when working with DCL.       +---------0-error, 1=ok  000=0=w 
 001=1=success  010=2=error  011=3=informational  100=4=fatal 	 101=5=??? 	 110=6=??? 	 111=7=???   
 Neil Rieck Kitchener/Waterloo/Cambridge,  Ontario, Canada.! http://www3.sympatico.ca/n.rieck/ J http://www3.sympatico.ca/n.rieck/links/cool_openvms.html (free Programming stubs)   ------------------------------  $ Date: Sat, 3 Jan 2004 12:32:44 -0500) From: "Neil Rieck" <n.rieck@sympatico.ca> 7 Subject: Re: How to make exe return value to dcl symbol ; Message-ID: <11DJb.20272$Vl6.3756135@news20.bellglobal.com>   7 "JF Mezei" <jfmezei.spamnot@istop.com> wrote in message # news:3FE9DA07.CD3738E6@istop.com...  > Tom Linden wrote:  > > G > > Using lib$get_foreign one can pass arguments to an executable.  But J > > this behaves like a subroutine.  How would you make it a function such1 > > that it would return a value to a DCL symbol?  > N > You catch Guy Peleg and give him a beer/cookie jar/chocolate/cognac/whateverM > and then kindly ask him to implement your program as a lexical function :-)  >  > There are 2 ways about it: > J > Your program simply returns a value (for instance, in C:  exit(myvalue); and K > then right after you can executed the program, the value is stored in the ; > symbol $STATUS which you can then copy to another symbol.  > N > Another way to do it is to have the program explicitely set a symbol's valueM > with LIB$SET_SYMBOL. You could supply the symbol name as an argument to the  program.  N Most high level languages will write some value to symbol $STATUS upon exit soM trying to call LIB$SET_SYMBOL to change symbol $STATUS will not work (because L the final "END" statement will overwrite it). In Compaq-BASIC, the statementH "END" will always supply an integer value of "1" (success) as a default.L However, an alternative to "END" is "END PROGRAM  I%" (where I% in this caseK is an integer variable containing the desired exit code to pass back to DCL  via $STATUS) can be used.   I Intermediate level languages like "C" operate a little differently. IIRC, J "exit" used without a parameter properly defaults to returning the integerH value of "0" (warning) to $STATUS. As stated above, "exit(myvalue)" will% return the desired exit value to DCL.   N One final point, in "C" on UNIX the integer value of "0" usually means successH while some non-zero value represents the failure message. In OpenVMS theF lowest 3 bits are reserver for severity while higher bits are used forM indetification. I leave the indentification bits blank when working with DCL.        +---------0-error, 1=ok  000=0=w 
 001=1=success  010=2=error  011=3=informational  100=4=fatal 	 101=5=??? 	 110=6=??? 	 111=7=???   
 Neil Rieck Kitchener/Waterloo/Cambridge,  Ontario, Canada.! http://www3.sympatico.ca/n.rieck/ N http://www3.sympatico.ca/n.rieck/links/cool_openvms.html (free OpenVMS program stubs and examples)    ------------------------------   Date: 3 Jan 2004 16:42:51 -0800 / From: stuie_norris@yahoo.com.au (Stuart Norris) ( Subject: Peeking in to the serial buffer= Message-ID: <51262235.0401031642.16063124@posting.google.com>   	 Hi Group,   A I am about to write a program that reads unsolicted data from the  serial port on my Alpha.  F I am use to write programs with QIO and the UCX/TCP QIO interface.  InE the UCX/TCP QIO interface you can peek at the messages in the UCX/TCP F buffer before reading them.  In the programs that I wrote this allowedF me to peek at the message length in my message header prior to reading! the data from the UCX/TCP buffer.   D Is there any way to peek into the terminal buffer on the serial portE using the QIO interface?  I wish to do this since the messages that I > will receive have the message length in the second byte of theE message.  I wish to peek at the message so I can determine the length * and then read the entire message in one go   Thanks   Stuart   ------------------------------  # Date: Sun, 04 Jan 2004 00:49:10 GMT " From:   VAXman-  @SendSpamHere.ORG, Subject: Re: Peeking in to the serial buffer0 Message-ID: <00A2B5BB.2988864F@SendSpamHere.ORG>  o In article <51262235.0401031642.16063124@posting.google.com>, stuie_norris@yahoo.com.au (Stuart Norris) writes: 
 >Hi Group, > B >I am about to write a program that reads unsolicted data from the >serial port on my Alpha.  > G >I am use to write programs with QIO and the UCX/TCP QIO interface.  In F >the UCX/TCP QIO interface you can peek at the messages in the UCX/TCPG >buffer before reading them.  In the programs that I wrote this allowed G >me to peek at the message length in my message header prior to reading " >the data from the UCX/TCP buffer. > E >Is there any way to peek into the terminal buffer on the serial port F >using the QIO interface?  I wish to do this since the messages that I? >will receive have the message length in the second byte of the F >message.  I wish to peek at the message so I can determine the length+ >and then read the entire message in one go  >  >Thanks  >  >Stuart   H Since you'll be doing two $QIOs anyway, why not read your message headerF (two bytes) in to a buffer (P1=buffer).  Then issue a second $QIO thatG has P2 equal to the size in the second byte (massage the value if it is G the length of the message plus the two bytes just read) into the buffer  +2 butes (P1=buffer+2)     --B http://www.legacy-2000.com  for the *best* OpenVMS system securityC                             solutions that others only claim to be.  --  K VAXman- A Bored Certified VMS Kernel Mode Hacker   VAXman(at)TMESIS(dot)COM              5   "Well my son, life is like a beanstalk, isn't it?"     ------------------------------  % Date: Sat, 03 Jan 2004 20:28:14 -0500 * From: JF Mezei <jfmezei.spamnot@istop.com>, Subject: Re: Peeking in to the serial buffer) Message-ID: <3FF76C1F.A717A172@istop.com>    Stuart Norris wrote:F > Is there any way to peek into the terminal buffer on the serial portG > using the QIO interface?  I wish to do this since the messages that I @ > will receive have the message length in the second byte of theG > message.  I wish to peek at the message so I can determine the length , > and then read the entire message in one go  N Much easier to just issue a QIO for 2 bytes, then, when that IO completes, youK get your length and issue a second QIO to read a now known number of bytes.   @ Peeking into the buffer requires that you also have some form of7 synchronisation to know when you've received something.   C Don't forget that the terminal driver has something very useful for J programming: a built-in timeout feature. You can specify the IO$M_TIMED (II think) flag, and provide a timeout value as one of the parameters. If you L expect 76 character to come, issue a read for 76 characters, but only 75 getK delivered, then after a while, the IO will complete anyways (with a timeout I status code), this makes it much easier to avoid a program getting stuck.    ------------------------------  % Date: Sun, 04 Jan 2004 01:35:41 -0500 3 From: "Richard B. Gilbert" <rgilbert88@comcast.net> , Subject: Re: Peeking in to the serial buffer0 Message-ID: <R6OdnaJ7JLWjKWqiRVn-gg@comcast.com>  H Yes, there is!   It's been about fifteen years since I last used it and A I don't recall the details but it's documented.  ISTR a function  D modifier for the $QIO call.   Instead of reading one character at a F time, I was able to see  how many characters were available and  read I that many.   The busier the system was, the more efficiently my code ran.   5 If anybody still has a copy of HOST32, it's in there.    Stuart Norris wrote:  
 >Hi Group, > B >I am about to write a program that reads unsolicted data from the >serial port on my Alpha.  > G >I am use to write programs with QIO and the UCX/TCP QIO interface.  In F >the UCX/TCP QIO interface you can peek at the messages in the UCX/TCPG >buffer before reading them.  In the programs that I wrote this allowed G >me to peek at the message length in my message header prior to reading " >the data from the UCX/TCP buffer. > E >Is there any way to peek into the terminal buffer on the serial port F >using the QIO interface?  I wish to do this since the messages that I? >will receive have the message length in the second byte of the F >message.  I wish to peek at the message so I can determine the length+ >and then read the entire message in one go  >  >Thanks  >  >Stuart  >    >    ------------------------------   Date: 3 Jan 2004 12:31:02 -0800 . From: spamsink2001@yahoo.com (Alan E. Feldman)/ Subject: Re: Pictures from the OpenVMS bootcamp < Message-ID: <b096a4ee.0401031231.dbc94cc@posting.google.com>  X VAXman-  @SendSpamHere.ORG wrote in message news:<00A2B558.A0F22FA0@SendSpamHere.ORG>...p > In article <b096a4ee.0401022330.42dbb03c@posting.google.com>, spamsink2001@yahoo.com (Alan E. Feldman) writes: > {...spin...}G > >No. I don't know the details of your situation, and I am NOT asking, E > >but I realize that such 4 years can produce much anxiety which can F > >cause serious health problems. But I cannot comment further without' > >more details, which I am NOT asking.  > J > No you don't know but you keep trivializing it as if it is somehow less I > horrific that another calamity you haven't witnessed first hand either.   A I was only six short blocks from WTC when it happened. I spoke to F co-workers and former co-workers who were part of and/or witnessed the horror of that day.   - Nonsense. I am not trivializing your plight.    H > >how devastated I'd be if one of them were killed. If you think that's? > >comparable to 4 yrs. of harrassment by lawyers, fine. I feel  > >differently.  > C > You're simply naive of the evil powers law(lie)yers possess.  The D > snarling, money-worshiping, process abusing law(lie)yer learned ofD > my brother's death.  He swore out a warrant for my arrest if I didE > not appear in his office to return a worthless product quick refer- D > ence card (claimed to be secret proprietary company information --E > so secret in fact that it can now be downloaded from a web site) on G > the day of my brother's funeral.  So, instead of being at my mother's F > side on that day, I was looking at the face of a vile lying theiving > schmuck.    E And you keep missing my point. Blaming an entire group because of the ! actions of a subset is not right.   E > >And what about the juries? Do they not bear any responsibility for < > >this state of affairs? OK, some trials don't have juries. > % > Yeah, right... blame the juries...    F I didn't say it was all their fault. I said they are in part to blame.  F > Juries are tampered with everyday... by the law(lie)yers involved in  ( That doesn't excuse the jurors' actions.  D > the case.  A jury of one's peers is bullshit.  In my case, my juryG > would have been the likes of people such as those here -- my peers -- G > that know computers, VMS, VAX and Alpha, and what it would take to do F > a port.  Not to be.  The law(lie)yers want to populate the jury withH > a group of half wits and morons that understand technical issues about2 > as well as the cross dresser behind the bench.    D I've been on jury duty. I've seen how lawyers do their initial juror@ rejections. They want easily malleable (or whatever the word is)B minds. They don't want independent thinkers. They don't want looseD cannons. They want to win their cases. Yes, I am well aware of this.  E > >I'll give you that some lawyers are despicable, but many aren't. I  >  > ...all law(lie)yers.    D Advertisers lie, too. Magicians lie. Kids lie. So do many others. So what?   0 > As long as the supreme tenet of their religionG > is that greed is good, they're all tainted.  Money is the root of all ( > evil so says the old biblical proverb.  D That's the love of money. Big difference. Pink Floyd makes this sameF mistake in their song "Money" on Dark Side of the Moon: Money, so theyC say/Is the root of all evil today/But if you ask for a rise it's no D surprise that they're giving none away. (Still a great album for itsE music, but that particular quoted part of the lyrics contains bizarre E illogic.) And what does the bible have to do with any of this anyway?   B I once lived in a group house in which one of the housemates was aC lawyer. She works for her state legislature and she is not anything E like what you say all lawyers are. She is overall a very good person. B You keep telling me things that I know are not true because I haveD seen them with my own eyes. And I tell you that and you still repeatE your mantra. And you'll repeat it again with the same silly spelling. D I guarantee it!!! And I guarantee that it won't sway my opinion even( an epsilon. But you'll repeat it anyway.  I > >I hope you don't continue to think badly of all New Yorkers. There are E > >good and bad NYers. And I guarantee you'd be surprised how many of  > G > Not the ones that moved into *MY* neighborhood.  Last year, whilst on H > my yearly pilgrimage to see Marillion, one claimed that my van clippedG > their car.  Wonderful to return home to three tickets, a missed court G > date, $1500 in more (il)legal fees and... in the end, a $100 fine for H > not being there.  A tough nut to crack when your in England but the NY2 > neighbor swears that you were driving the van.    B Well, maybe other NYers were fed up with them and kicked them out!  E Man, you have more bad luck than most. I sympathize with your plight, E but I disagree with your conclusions. Just because I say one thing is F worse, that doesn't trivialize the lesser. E.g.: 900 > 800, but 800 isE still a lot. Not that I apply those numbers to anything in particular  in this subthread!  E And you still haven't expalined the beanstalk bit. I liked the Calvin  and Hobbes quote, though.   D I'm sorry you have endured all this injustice, and I hope somehow itB is made right one day. I have family members who have endured even worse. Much worse, actually.  / But I still disagree with some of what you say.    Alan E. Feldman   C Everything you know is wrong. --Firesign Theater: Did you know that  everything you know is wrong?    ------------------------------  # Date: Sat, 03 Jan 2004 21:03:10 GMT " From:   VAXman-  @SendSpamHere.ORG/ Subject: Re: Pictures from the OpenVMS bootcamp 0 Message-ID: <00A2B59B.975362E3@SendSpamHere.ORG>  m In article <b096a4ee.0401031231.dbc94cc@posting.google.com>, spamsink2001@yahoo.com (Alan E. Feldman) writes:  {...snip...}E >Advertisers lie, too. Magicians lie. Kids lie. So do many others. So  >what?  L There not in a court of law there their are (supposedly) penalties for lyingL -- I believe it's considered the crime of perjury.  Oh, I suppose that thereM is some defense that the law(lie)yer is only regurgitating what hisher client K has told them.  I've seen this too.  Don't tell me that the law(lie)yers is L not aware of the truth.  I know fully well that the shyster that went for myK throat and the throats of serveral other good and well intention people and M corporation knew the suit was a farse and a lie.  It was anything for a buck!     E >That's the love of money. Big difference. Pink Floyd makes this same G >mistake in their song "Money" on Dark Side of the Moon: Money, so they D >say/Is the root of all evil today/But if you ask for a rise it's noE >surprise that they're giving none away. (Still a great album for its F >music, but that particular quoted part of the lyrics contains bizarreF >illogic.) And what does the bible have to do with any of this anyway?  ; There's not illogic on DSotM.  Brilliant commentary is all.     C >Well, maybe other NYers were fed up with them and kicked them out!   * Yeah... NJ gets its reputation for it too.    F >And you still haven't expalined the beanstalk bit. I liked the Calvin >and Hobbes quote, though.    & http://www.procolharum.com/w/w0208.htm    E >I'm sorry you have endured all this injustice, and I hope somehow it C >is made right one day. I have family members who have endured even  >worse. Much worse, actually.    Think?    0 >But I still disagree with some of what you say.   As I do you.     --B http://www.legacy-2000.com  for the *best* OpenVMS system securityC                             solutions that others only claim to be.  --  K VAXman- A Bored Certified VMS Kernel Mode Hacker   VAXman(at)TMESIS(dot)COM              5   "Well my son, life is like a beanstalk, isn't it?"     ------------------------------   Date: 3 Jan 2004 22:04:18 -0800 . From: spamsink2001@yahoo.com (Alan E. Feldman)/ Subject: Re: Pictures from the OpenVMS bootcamp = Message-ID: <b096a4ee.0401032204.490079a5@posting.google.com>   X VAXman-  @SendSpamHere.ORG wrote in message news:<00A2B59B.975362E3@SendSpamHere.ORG>...o > In article <b096a4ee.0401031231.dbc94cc@posting.google.com>, spamsink2001@yahoo.com (Alan E. Feldman) writes:  > {...snip...}G > >Advertisers lie, too. Magicians lie. Kids lie. So do many others. So  > >what? > N > There not in a court of law there their are (supposedly) penalties for lying  E *You* are the one who is putting all the emphasis on lying. I've been F trying to point that out for several posts now. So please stop putting words in my mouth.  N > -- I believe it's considered the crime of perjury.  Oh, I suppose that thereO > is some defense that the law(lie)yer is only regurgitating what hisher client M > has told them.  I've seen this too.  Don't tell me that the law(lie)yers is  > not aware of the truth.    I didn't, and I won't.    5 > I know fully well that the shyster that went for my M > throat and the throats of serveral other good and well intention people and O > corporation knew the suit was a farse and a lie.  It was anything for a buck!   D I never refuted that. I still think Al Qaeda is worse. When Al QaedaF sets off multiple nukes, will you still think the trial lawyers are asA bad? The lawyers will haul you into court and take your money. Al @ Qaeda will shoot your wife for not covering herself up, and thenC probably kill you and the rest of your family. Which do you prefer?   C You have suffered more from lawyers than from Al Qaeda. Others have E suffered more from Al Qaeda. I know of no one except you who puts the F two on the same level. Have you made your case to someone who has lostD loved ones from 9/11? I think they'll find it kind of hard to see it your way, don't you think?  G > >That's the love of money. Big difference. Pink Floyd makes this same I > >mistake in their song "Money" on Dark Side of the Moon: Money, so they F > >say/Is the root of all evil today/But if you ask for a rise it's noG > >surprise that they're giving none away. (Still a great album for its H > >music, but that particular quoted part of the lyrics contains bizarreH > >illogic.) And what does the bible have to do with any of this anyway? > = > There's not illogic on DSotM.  Brilliant commentary is all.   0 They got the quote wrong. How is that brilliant?  E > >Well, maybe other NYers were fed up with them and kicked them out!  > , > Yeah... NJ gets its reputation for it too. >  > H > >And you still haven't expalined the beanstalk bit. I liked the Calvin > >and Hobbes quote, though. >  > ( > http://www.procolharum.com/w/w0208.htm   Finally an answer to this one.  G > >I'm sorry you have endured all this injustice, and I hope somehow it E > >is made right one day. I have family members who have endured even  > >worse. Much worse, actually.  >  > Think? >  > 2 > >But I still disagree with some of what you say. >  > As I do you.     Agreed. :-)    ------------------------------   End of INFO-VAX 2004.007 ************************