home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!ogicse!das-news.harvard.edu!spdcc!jti.com!richb
- From: richb@jti.com (Richard Braun)
- Newsgroups: comp.os.linux
- Subject: Re: UUGETTY & Lock files...
- Message-ID: <C0EFGJ.706@jti.com>
- Date: 5 Jan 93 20:55:30 GMT
- Article-I.D.: jti.C0EFGJ.706
- References: <1993Jan3.061828.27179@exucom.com>
- Sender: news@jti.com (News Admin)
- Organization: Jupiter Technology Inc. / Waltham, MA
- Lines: 37
- Nntp-Posting-Host: bart.jti.com
-
- kjv@exucom.com (Karl J. Vesterling) writes:
- > Where do I obtain a version of uugetty which functions, or at
- >least has the documentation? Here are the problems I am experiencing:
- >
- > o Lock-files (They exist, BUT they all are invalid/stale)
- > Maybe the pid in them is not the pid of the users shell, and
- > is/was the pid of uugetty itself? IE, uugetty creates the
- > lock file, puts it's PID in the file, then exits, thus
- > causing the lock file to be stale?
-
- It took a bit of source-code reading, but I finally figured out how
- to get uugetty working properly. First off, there's an issue with
- the contents of the lock files. Apparently some software uses ASCII-
- format PIDs, and other programs use binary-format. (You can tell
- by seeing if the file is 4 bytes long or about 10: the 4-byte version
- is binary.) My version of kermit uses binary PIDs so I had to change
- and recompile uugetty to do likewise. (I've not checked the format
- of the files created by uucp.) It'd be nice if all this were
- consistent in Linux 1.0.
-
- The next trick is this: you have to tell uugetty to wait before
- creating a lock file. This can be done by putting WAITCHAR=YES in
- the /etc/default/uugetty.ttyxx file.
-
- Problems with uugetty can sometimes be solved by putting DEBUG=0377 in
- the uugetty.ttyxx config file; it'll create voluminous output in
- the /tmp directory, which can be helpful especially if you compare
- it with the source code.
-
- I'm happy to say that once I went through all the code walk-through,
- the program does work as advertised and makes my system far more
- accessible. (Especially with the nice new modem!) As is typical of
- Linux, the source code is wonderful and the doc leaves something to
- be desired. (Any tech writers out there, wanting to make a meaningful
- contribution?)
-
- -rich
-