[comp.sys.3b1] tty000 troubles w/modem

ruck@reef.cis.ufl.edu (John Ruckstuhl) (04/28/91)

I've freshly installed 3.51m and HDB on a 3b1 with a US Robotics 
Password 1200 baud modem on tty000.  The Password uses a subset of the
Hayes AT command set, and powers up in a no-echo mode (the command ATE1
will start the command echoing, and then it looks much like a Hayes).

The same hardware ran 3.51 and stock uucp without complaint (using a
slightly different modem definition in modemcap) for a couple of years.

Now, I am *most* frustrated.  
Uucico will not reliably talk to the modem.
It's as if tty000 gets clogged up, because sometimes it seems I can
clear the trouble (for exactly one test) by cycling power on the modem.

I almost always get 
    TIMEOUT (generic open)
    CONN FAILED (CAN'T ACCESS DEVICE)
in the logfile.  The exception is only sometimes, when the modem is
freshly powered up.

It sort of looks like the chat session defined in Dialers is failing,
because -x9 shows that I don't reach the point where I'm trying to chat
per script in Systems.  So then, I try to make the connection to the
modem a "direct" one and embed all the dialing and chatting in the
Systems chat script, so that it will show during uucico -x9, but it's
still failing earlier than that.

I've tried various inittab entries, but to no avail.  I don't accept
computer calls, so I figure I don't need a getty on tty000 and I'm not
now running one (but I got the same symptoms when I did run one).

Some System Administration guides suggest that one ought to be able to
do things like 
    echo "A" > /dev/tty000
but here, even root can't do that -- you hang, (working), until you
break out of it, then you get the error message:  
    /dev/tty000: cannot create
Or at least, I do.

Please help me understand this.  If you have time to help me, and think
that it would be more efficient to talk with me over the telephone while
I run experimental commands at the console, please email me your
telephone number and I will call you at your convenience.
I repeat, I am very frustrated.

Thank you,
ruck.
PS -- don't mail to sphere -- she's offline now (for obvious reasons).
-- 
John R Ruckstuhl, Jr			ruck@alpha.ee.ufl.edu
Dept of Electrical Engineering		ruck@cis.ufl.edu, uflorida!ruck
University of Florida			ruck%sphere@cis.ufl.edu, sphere!ruck

ruck@reef.cis.ufl.edu (John Ruckstuhl) (04/29/91)

I wrote:
>I've freshly installed 3.51m and HDB on a 3b1 with a US Robotics 
>Password 1200 baud modem on tty000.  The Password uses a subset of the
>
>Now, I am *most* frustrated.  
>
>Uucico will not reliably talk to the modem.
>    CONN FAILED (CAN'T ACCESS DEVICE)
>It sort of looks like the chat session defined in Dialers is failing,
>because -x9 shows that I don't reach the point where I'm trying to chat
>per script in Systems.  So then, I try to make the connection to the

Now, I am *most* pleased.
Thank you to
    Floyd Davidson <floyd@ims.alaska.edu>
    ostroff@oswego.Oswego.EDU (Boyd Ostroff)
whose replies I have already received, and to those whose replies are
already on their way to me.

For those who know -- Floyd & Boyd informed me of the ,M and \M & \m,
and I'm back online.
For those who don't know,  I'll wait a few days for other replies, then
compose & post a summary.

Best Regards,
ruck.
-- 
John R Ruckstuhl, Jr			ruck@alpha.ee.ufl.edu
Dept of Electrical Engineering		ruck@cis.ufl.edu, uflorida!ruck
University of Florida			ruck%sphere@cis.ufl.edu, sphere!ruck

dnichols@ceilidh.beartrack.com (DoN Nichols) (04/29/91)

In article <28248@uflorida.cis.ufl.EDU> ruck@reef.cis.ufl.edu (John Ruckstuhl) writes:
>I've freshly installed 3.51m and HDB on a 3b1 with a US Robotics 
>Password 1200 baud modem on tty000.  The Password uses a subset of the
>Hayes AT command set, and powers up in a no-echo mode (the command ATE1
>will start the command echoing, and then it looks much like a Hayes).

	[ ... ]

>Some System Administration guides suggest that one ought to be able to
>do things like 
>    echo "A" > /dev/tty000
>but here, even root can't do that -- you hang, (working), until you
>break out of it, then you get the error message:  
>    /dev/tty000: cannot create
>Or at least, I do.

	Sounds as though your problem is that the older version didn't pay
attention to DTR, and the newer one does.  (Does the robotics machine have a
DTR light?  Is it out when you are unsucessfully attempting a connection?
If so, do you have either a hardware switch or a software switch to either
make it hold DTR true forever, or preferably to drop DTR momentarily, then
restore on loss of carrier.

	Since it works just after power-up, you might try sending the
reset string 'AT Z' during the cleanup after a call, (while you still have
its attention - that is before the other system drops carrier.)

	I seem to remember mention of an alternate minor device number which
allows access while ignoring the DTR signal.  If I remember correctly, you
add 128 to the minor device number of your port, and mknod a new device with
that.  That device ignores DTR.  (I don't remember whether that was in this
group, but for my newly acquired sun-2, it is documented, and is used for
outgoing calls, while the other is used for incoming calls.  When one is
opened, the other cannot be opened, thus preventing an outgoing call attempt
from disrupting an incoming one.  (As I say, I don't know for sure whether
this works on the 7300/3b1.)

	There is an excellent set of help files for HDB on osu-cis --
somewhere.  (I don't know where it landed after the reorg.)  While it was
mostly aimed at Trailblazers, it will probably help you, as well.

	Good Luck
		DoN.
-- 
Donald Nichols (DoN.)		| Voice (Days):	(703) 664-1585
D&D Data			| Voice (Eves):	(703) 938-4564
Disclaimer: from here - None	| Email:     <dnichols@ceilidh.beartrack.com>
	--- Black Holes are where God is dividing by zero ---