[4251] in testers
Re: larvnet is not monitoring ajax
daemon@ATHENA.MIT.EDU (Mike Whitson)
Thu Jun 17 17:35:22 1999
To: Dan Winship <danw@MIT.EDU>
Cc: testers@MIT.EDU
From: Mike Whitson <mwhitson@MIT.EDU>
Date: 17 Jun 1999 17:35:15 -0400
In-Reply-To: Dan Winship's message of Thu, 17 Jun 1999 15:07:20 -0400 (EDT)
Message-Id: <n914sk64ijw.fsf@dragon.mit.edu>
Dan Winship <danw@MIT.EDU> writes:
> I just noticed larvnetd on dill syslogging confusion about ajax.
> LPRng's lpd gives different output for lpq, and larvnetd can't parse
> it correctly.
> Mike, you should comment on whether or not the lpq formatting is
> going to change again. Note that larvnetd sends an lpq request and
> parses the output by hand, rather than using the lpq program.
RFC 1179 states, for both the short and long lpq requests:
The response is an ASCII stream which describes the printer queue.
The stream continues until the connection closes. Ends of lines are
indicated with ASCII LF control characters. The lines may also
contain ASCII HT control characters.
Alas, that's the only specification there is; all it guarantees is
that any client that tries to parse it is going to lose.
As for our lpd, it may well change in the near future; I'm still
working on making LPRng's lpd respond with something more palatable to
older version of lpq. If at all possible, I think we should move over
to using LPRng's machine-parsable lpq command (which is an extension
to RFC 1179). Unfortunately, I don't know of any good documentation
for what it returns. I'll pester papowell.
-mike
