NetNews Usenet Archive 1992 #31/NN_1992_31.iso/spool/comp/sys/att/2742

Internet Message Format | 1992-12-25 | 3.9 KB

Path: sparky!uunet!paladin.american.edu!gatech!destroyer!cs.ubc.ca!mala.bc.ca!oneb!hakatac!rthomas
Newsgroups: comp.sys.att
Subject: Re: Need source for 3B2->DB-25 adapter.
Message-ID: <[e-mail]>
From: [e-mail] (Robert N Thomas)
Date: Sat, 26 Dec 92 00:30:28 GMT
References: <[e-mail]> <[e-mail]> <[e-mail]>
Organization: Sir HackAlot's Castle BBS, Port Alberni, B.C., Canada
Lines: 69

In article <[e-mail]> [e-mail] (Greg A. Woods) writes:
>In article <[e-mail]> [e-mail] (Robert N Thomas) writes:
>>          When the DCD is set inactive, the 3B2 will assume that the user
>> has either turned off the terminal, or dropped carrier.  If this occurs,
>> the user is logged off the system, and background tasks are aborted..
>> If DCD is inactive, be aware that NONE of the standard AT&T provided utilities
>> are capable of sending anything to the device.
>> {cu won't work on the device, and forget about outbound UUCP's from working}
>
>That's not quite true.
>
>It depends upon which version of HDB-UUCP you have.  Specifically, the
>newer versions, generally available starting with UNIX System V
>Release 3.2, modem control is supported by the ",M" flag in the
>Devices file, and the "\m" and "\M" controls in the Dialers file.

    I am running Rel 3.2 and nowhere do I see mention of the M controls.
I do however recall the discussions of M controls a few months ago..  I
wanted to discuss something that would work on ALL HDB's rather then
expanding the discussion to "new verses OLD HDB's".  Clearly, the above
view works on ALL HDB's no matter what version you run.  Maybe we don't
get every machine running like a HOT ROD, but we get something that works
and a more readable/shorter text...

>Also, the effect of loss of DCD depends upon the default termio
>setting of the "clocal" flag, and on the disposition of the SIGHUP
>signal for each process associated with the terminal.

    Agreed..  I was thinking about clocal, and line disiplines and even the
nohup command as well when I wrote that.  I felt that mentioning every single
exception, and every nifty little flag one could enable, and every nifty way
one could run would make my "short little blurb" a bit larger and harder
to read..

    You are quite right about clocal, and nohup command will stop a process
from termination on hang up..  I opted for simpler, easier to read, and
works no matter what version of HDB you run.  I feel that most of the
systems don't run with CLOCAL, or do nohup for each and every process/
command they run.  Hence the "view" I present is generally a valid one.

I do however, feel the following word sequence should be changed from:
>> If DCD is inactive, be aware that NONE of the standard AT&T provided utilities
>> are capable of sending anything to the device.

to:
>> If DCD is inactive, be aware that in some versions of UUCP, NONE of the
>> standard AT&T provided utilities are capable of sending anything to the
>> device.



>Actually, this is more important as a factor in keeping the EPORTS
>driver alive than anything else.  Long periods without DCD seem to
>cause the driver (at least version 1.3) to go south every so often,
>leaving a hung modem port.

    I have ran into something similar as well..  Try "kill -9 " on your
uugetty and getty.. That seemed to help me.  Again, I didn't want to
get into "do this for this version, but if mixed with that one, do this
except in the case of ... then do this".  I wanted something that was
"simplified", "understandable", and "works" even at the expense of it being
"exactly technically correct" for every HDB version, every EPORTS driver
version, and every clocal and nohup'ed process..

Cheers and Merry Christmas....

                        ---- Robb ----