Xref: utzoo comp.sys.att:5699 comp.mail.uucp:2803
Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!unmvax!ncar!boulder!sunybcs!bingvaxu!leah!itsgw!steinmetz!uunet!ssbell!mcmi!denny
From: denny@mcmi.UUCP (Dennis Page)
Newsgroups: comp.sys.att,comp.mail.uucp
Subject: Re: HoneyDanBer UUCP uucico problem (bug?)
Keywords: hdb uucico 3b2 att
Message-ID: <803@mcmi.UUCP>
Date: 27 Feb 89 23:05:34 GMT
References: <133@aucis.UUCP> <810@sigmast.UUCP>
Reply-To: denny@mcmi.UUCP (Dennis Page)
Distribution: usa
Organization: MCMI, Omaha, NE
Lines: 30

In article <810@sigmast.UUCP> dgy@sigmast.UUCP (Dave Yearke) writes:
>Devices File:
>ACU tty26,M - 2400 modema
>ACU tty26,M - 1200 modema

>Systems File:
>alpha ACU 2400 555-1212 (rest of stuff omitted)
>alpha ACU 1200 555-1213 (rest of stuff omitted)

>uucico will try the first number at 2400 using "modema".  If it is busy, it
>will try the second number in the list using "modema", BUT IT IS STILL TALKING
>TO THE MODEM AT 2400!!

The problem is actually caused in the Dialers file with the use of
the "\M" and "\m" commands.  It seems that the ioctl structure used by
these commands are initialized only at the beginning of the dialing
sequence.  Somebody trying to save an ioctl or two.


>The AT&T hotline people, who are usually on top of things, have never heard of
>this happening.

AT&T (New Jersey) has acknowledged the problem (and tested a fix for
it), and will be sending a change/fix request to whomever they send
those things to within AT&T.  A fix should appear in future versions
of HDB.

What an annoying little bug.
-- 
When the journey of life ends, have you reached a destination?