Subject: Re: Keyboard support for DEC3000/300
To: Jason Thorpe <email@example.com>
From: Jonathan Stone <jonathan@DSG.Stanford.EDU>
Date: 03/21/1998 18:54:23
>As long as it's wscons, not at all.
>(I'm hoping Matthias Drochner will get the New wscons code into the
>tree reasonably soon - it's much better than the old stuff.)
Yeah. I had reworked the pmax rcons code to get rid of the gross hacks
that loop back through the tty driver (there's an open PR on that) but
I didnt see much point debugging it when the new wscons was going to
be done Sometime SOon.
There's a simple bug in the Pmax version of the scc.c. chagnes. The
"TK_NOTYET" needs to changed to be "HAVE_RCONS". When I fix that,
the keyboard sort of works again. Without that, the keyboard input
I get is garbage. (I am guessing the same problem was affecting
Tohru-san's earlier changes, since I saw the same symptom there.)
But the only difference I'm seeing on a 5000/150, with rcons (on an
cfb) after applying this patch and fixing TK_NOTYET< is that keyboard
input is silently lost, rather than generating overflow messages.
Doing an "ls -l" of about 20 files on the console, and touch-typing
very fast, I lose about one char in three.
How was this code tested on a 5000/25? Was it used with PPP or SLIP,
or something else?