I've been tearing my hair out trying to get a Supra 56k modem working 
on a IIcx with the subject 1.5 snapshot.  That's generic #46.

Without going into details, I started with a working configuration 
from a 1.4.1 installation on a PC, and went from there including 
using cdtrcts.  Doing a stty -ef /dev/tty00 while pppd had control of 
the port showed the correct change of baud rate, but some other 
handshake configuration than dtrcts.

I switched to /dev/tty01, the printer port, and everything works 
perfectly.  Doing a stty shows the correct port configuration.

I swear that changing the device and moving the cable is the only 
change.  /etc/ttys has getty off on both lines.  I'd blame hardware, 
but the stty shows the software-commanded configuration, right?  How 
can that be different?

Does this problem deserve a PR, or is July ancient history?


Possibly relevant hardware notes:

The Supra is a cute little snot that doesn't have room for a 25-pin 
connector so they ran a cable out directly to a 8-pin mini-DIN on the 
Mac version.  From the dates and the fact that the modem works under 
MacOS I have little doubt they wired the signals properly for 
hardware handshaking.  They don't document what they did though.

The Mac II machines use a z8530 which IIRC does not run all the 
handshaking signals out of the chip on the second port.  I don't 
remember if they run all the ones that the Mac uses, but maybe not. 
That, and the in-chip priority is why the modem port is labeled for a 
specific chip port.

The Mac IIcx has an under-specified component on the motherboard 
associated with detecting whether the printer port has a serial 
device or LocalTalk.  It will fail if you have a serial device 
connected for a long period of time, and I switched my DeskWriter 
from serial to LocalTalk a long time ago because of sporadic problems 
that could have been related to this.  I don't know how the interface 
detects LocalTalk so I don't know if NetBSD would be immune to the 
failure or if some handshaking line would go dead.

Signature held pending an ISO 9000 compliant
signature design and approval process.
h.b.hotz@jpl.nasa.gov, or hbhotz@oxy.edu