Re: colortext translator
Marcus Brinkmann <Marcus.Brinkmann@ruhr-uni-bochum.de> writes:
> Had you told earlier that you make such a steady progress,
> I probably would not have spent the last days to get the linux
> console working.
I have only been doing this for four days and didn't want to
announce vaporware. Besides, I assumed the Linux console was
important since you were putting your time in it. I was
surprised to read it wouldn't "have a long life expectancy".
> (Are you subscribed to bug-hurd?
Yes. Let's move this thread there.
> You must have noticed my screams for help :)
But I was feverishly (37.4°C) coding and didn't read mail :)
> Have you looked at KGI yet?
I did try it a couple of times on Linux 2.0 but am not very
familiar with its internals.
> Can you take a look at it?
First I want to get input working without it.
> Also, how interacts your translator with the kernel logger?
It doesn't. :) I suppose syslog could forward the messages.
The video card may get confused if the translator and the kernel
try to move the cursor at the same time. I don't think there's
any way to lock against this, apart from not writing to the Mach
console at all.
> How does it interact with the keyboard driver (we need to make
> sure we can get X working with it in some way).
It doesn't read the keyboard yet. It should open the "kbd"
device, set it to event mode and read records. This would make
the "kd" device (to which "console" usually refers) stop
receiving keyboard data.
X would have to open the new console translator, tell it to skip
the keymap, and read keycodes (or scancodes?) from there.
Virtual consoles are more important to me than X, and I'm going
to implement them first. So the X server would allocate one of
those, as in Linux. When the user switches to the VC used by X,
the console translator would send some kind of notification to
the X server so that it can restore graphic mode and redraw its
screen. (In the KGI approach, the console translator would
restore graphic mode but X would still draw its own screen.) I
haven't thought of this much.
Now I'm going to implement keyboard input, and then virtual
consoles (which need hotkeys). After that, the console would
have to get support for graphics-mode programs, possibly with