home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!ulowell!m2c!bu.edu!att!ucbvax!VMSFE.ULCC.AC.UK!CZIWKGA
- From: CZIWKGA@VMSFE.ULCC.AC.UK ("Kevin Ashley, Systems Development, ULCC")
- Newsgroups: comp.os.vms
- Subject: Re: problem editing a file - wrong char set ?
- Message-ID: <9301280105.AA02133@ucbvax.Berkeley.EDU>
- Date: 27 Jan 93 16:35:00 GMT
- Sender: daemon@ucbvax.BERKELEY.EDU
- Organization: The Internet
- Lines: 76
-
- About a week ago (we've had some mail backlogs around here)
- Ulli Horlacher wrote:
-
- > We (VAX 9410, VMS 5.5, PSI 4.3) have users which use a databank on a VM
- > System via x.25 remote login. They do a SET HOST/X29/LOG to the VM.
- >
- > The logfile which is produced looks very wired (or at least I don't
- > know what it means. :-) ):
- >
- > If I TYPE it on a vt100, it looks like plain ASCII, but if I TYPE it on a
- > VT220 or edit it (EDT or LSE) it contains only control-characters.
- >
- > Now my questions: - why can I type (and read) it on a vt100 but not on a
- > vt220?
-
- Because the IBM insists on sending out characters with parity bits
- on (can't remember if it likes even or odd, but it likes one of them),
- and the /LOG function of PSIPAD records data exactly as sent to the
- terminal. The VT100 doesn't support eight-bit data, and so either
- ignores the parity bit or, at worst, checks it. The VT200 does support
- eight-bit data and tries to interpret the parity bit as part of the
- character code, unless you've set it to 7-bit mark or space parity (which
- causes it to ignore the parity bit) or some other 7-bit setting which
- matches the IBM.
-
- [Note that 7-bit space parity is the only safe one ;
- VMS will still think that your terminal is using 8-bit no parity, and
- won't recognise carriage return (amongst others) if it arrives
- with a parity bit set.]
-
- When you use EVE or EDT, they always indicate the presence of characters
- with the top bit set, even on terminals such as the VT100 that don't
- support displaying them directly.
-
- > - is there a way to prevent this wrong character set,
- > like SET HOST /qualifiers?
-
- Well, this question is a bit broad - it MAY be possible to prevent
- the IBM system sending out characters with parity bits on by configuring
- things within the IBM system. You don't say if it supports X.29 directly
- (with NPSI and VTAM) or if the link is somewhat more indirect. In any event,
- your message suggests you don't have much control over that end of things.
-
- Given that that is the case, you can't stop the logfile containing these
- characters, since PSIPAD always logs what it gets. Doing a SET TERMINAL/NOEIGHT
- doesn't alter this (nor does it stop PSIPAD sending all eight bits to
- the terminal).
-
- What you can do is massage the logfile to make it readable, which
- brings us to your next question...
- >
- >
- > - how can one translate old logfiles of this kind to a
- > printable format (the databank inquiry was very
- > expensive, please don't say: "do it again with other
- > terminal/SET HOST setups")?
- >
-
- I have a program which will take logfiles produced by PSIPAD (SET HOST/X29)
- and turn them into variable-length record files with implied carriage
- control, optionally stripping parity. You may or may not be aware (you
- will be after removing the parity bits) that the logfiles are
- confusing in more ways than one ; they are variable length, with no
- end-of-record carriage control, but with carriage returns and linefeeds
- possibly embedded within the records. This is another feature of
- how PSIPAD produces the logfiles - each received X.29 packet produces
- one record in the logfile. Such logfiles print OK, TYPE OK, but are
- difficult to edit with most editors (TECO being an exception) and
- difficult to post-process if that is what you want to do.
-
- If you're interested in the program, and haven't already received better
- answers, email me privately and I'll send it.
- ------------------------------------------------------------------------------
- Kevin Ashley K.Ashley@Ulcc.ac.uk
- Systems Development Group Manager ...ukc!ncdlab!K.Ashley
- University of London Computer Centre.
-