2-way calling on voice line

Greg A. Woods woods at ixpierre.uucp
Fri Oct 21 15:55:37 AEST 1988


In article <593 at ssp15.idca.tds.philips.nl> jos at idca.tds.PHILIPS.nl (Jos Vos) writes:
> In article <210 at fantasci.UUCP> jep at fantasci.UUCP (Joseph E Poplawski) writes:
> >Hello.  I am currently running ISC UNIX V.3 on an 80386 with a Hayes Smartmodem
> >1200.  The only phone line I have at this time is my voice line.  What I would
> >like to do is have my system be able to receive calls at certain times/days.
> >But I also need the system to be able to call out when it has to do its polling
> >for news and mail. The modem is connected to tty01.
> 
> Use uugetty (it's in /usr/lib/uucp) for that purpose. An entry in /etc/inittab
> should look like
> 
> 	01:234:repawn:/usr/lib/uucp/uugetty -r -t15 tty01 1200U
> 
> where 1200U is an entry of /etc/gettydefs. Uugetty can handle both
> incoming and outcoming calls.

I have a number of problems when I try this.

The device tty01, as delivered, isn't a modem control port, and
therefore doesn't work right for dial in.

Uugetty dies as the modem answers the call, causing the modem to
immediately drop the line again (when using the modem-control port).

If a process opens either the modem control device, or the no wait
device while someone is dialed in through the modem control device,
the line is dropped.  (Try to determine the baud rate of the incoming
call using 'stty < /dev/tty01'.)

NO, I cannot re-wire, nor re-configure my modem.  The modem is an
internal modem, so can't be re-wired.  The current configuration works
perfectly when the modem is used for either incoming or outgoing calls
seperately.

If only we were using Microport's version, we could dispense with
uugetty and one-way modems.  They use a scheme in which an open on the
modem control device by getty (which is still blocked because there is
no DCD signal) is prevented from completing, by the device driver,
while the no wait device is open, and vice-versa (except that when
the modem control port is in use, no wait device opens fail, not
block).  [ This is how it was described to me, subject to my
interpretation as to how the ideal situation should be. ]
-- 
						Greg Woods.

UUCP: utgpu!woods, utgpu!ontmoh!woods, {ontmoh,tmsoft,telly}!ixpierre!woods
VOICE: (416) 443-1743 [h]		LOCATION: Toronto, Ontario, Canada



More information about the Comp.unix.questions mailing list