Hayes modem problem with Suns

[Greg Earle]

I have come up with a rather puzzling situation.  This machine's dialup
line formerly had Call Waiting (TM?) enabled, with the result that humans
logged on could get nuked by another computer, or by an outgoing uucico.
Needless to say, this was pretty annoying, so it was taken off.  Shortly
thereafter, I found a pattern emerging in my nightly cron-based uucico's:
A system would be called (whichever was first), and the transaction would
proceed normally (that is, either the system would be reached successfully
or not, it didn't matter) until it completed; either by successful completion
or the login would fail.  From then on, however, every succeeding attempt
to reach all the remaining sites with work for them are failing, with
the following entry in my uucp spool LOGFILE:

uucp machine (4/1-3:16-9607) LOST LINE (LOGIN)
uucp machine (4/1-3:16-9607) LOST LINE (LOGIN)
uucp machine (4/1-3:16-9607) FAILED (LOGIN)
uucp machine (4/1-3:16-9607) FAILED (call to machine )

Thinking this odd, I put the following line into my nightly script:

/usr/lib/uucp/uucico -r1 -x7 2>&1 | mail usenet

and what I see is: both attempts to get an 'OK' back from the modem (1st via
'AT', 2nd via 'ATZ') get nothing, and the debug output shows
 (deleted) ... lost line -- kr -1
or some such.  I don't have source, so I can't see what is happening.
If (as root) I execute a 'uucico' when more than one system has work, it
works ok (aka I can't seem to replicate the problem manually) which would
make me think it is some sort of permission problem; except that I changed
*nothing* after the Call Waiting was removed from the line, and it worked
fine before that!!  Arrrgghhh ...

Anybody seen any similar symptoms anytime??  This is *not* an April Fool's
posting!! :-)

	Greg Earle
	Jet Propulsion Laboratory
	sdcrdcf!smeagol!earle (UUCP)
	ia-sun2!smeagol!earle at csvax.caltech.edu (ARPA)