Playing nice with others
Gert Doering (gert@greenie.muc.de)
Sun, 13 Dec 1998 13:32:58 +0100
Hi,
On Sat, Dec 12, 1998 at 03:30:37PM -0600, Brian wrote:
> I believe it would be very helpful if the mgetty docs/FAQs made some
> mention of possible interactions between mgetty and other serial
> applications trying to share the same line.
There is lots of stuff about this in the "Linux" section of the mgetty
manual (http://alpha.greenie.net/mgetty/, or "mgetty.info").
> In my case, the pppd
> configuration works just fine when mgetty has been diabled, but refuses
> to work when mgetty is brought up. Even though this might be a pppd
> problem, it clearly is mgetty which is the catalyst for the problem.
> Clearly, mgetty is introducing some sort of behavior which alters access
> to the serial line in otherwise well-behaved applications.
Sure - mgetty will trigger bugs in other applications *faster*. If an
application doesn't do the serial line locking right, it will run into
problems eventually. If you use mgetty, the problems are seen
*immediatly*, which is a Good Thing.
> My apologies if this issue has already been hashed out and decided
> upon. I just found the attitude in the original thread similar to that
> of a big o/s vendor who will remain nameless but can commonly be heard
> quoting the same refrain: "It's not *our* problem, it must be a problem
> with your application."
Actually, mgetty's behaviour is very well defined, and all the locking
issues *are* discussed to death in the list archives, the FAQ, and the
manual.
On the other hand: if you are willing to contribute some additional
chapters to the manual about correct setup of kermit, pppd, cu, uucp,
minicom, seyon on Linux, SCO Unix, FreeBSD, SunOS, Solaris, ... (see the
problem?), you're more than welcome. Until then, documentation on mgetty
has to suffice.
gert
--
USENET is *not* the non-clickable part of WWW!
//www.muc.de/~gert/
Gert Doering - Munich, Germany gert@greenie.muc.de
fax: +49-89-35655025 gert.doering@physik.tu-muenchen.de