[Getting caught up with old email, and can't remember if I responded to
this before.]

On Tue, 3 Nov 1998, Reinhold Huber wrote:

> Since then I encountered another problem: At least when I interrupt some
> process which consumes some system load, e.g. a make installing emacs,
> the ctrl-key inverts its function: I have to press C-c instead of c, c
> instead of C-c and so on.
> 
> Is it possible that this change has caused this? What info should I supply
> to help finding out what's up?

  I think I've seen this once in a while, but haven't tracked it down.  I'd
guess that the key-up transition of the control key got lost.  The dtop.c
interrupt handler still misses data once in a while [there's a printf() 
that's enabled with DEBUG or DIAGNOSTIC that should print a message when
the bus gets an 'overrun'].  I thought that the way the LK501 keyboard
data is passed should prevent that, but I don't remember the details at
the moment.  [I seem to recall that each packet of data from the keyboard
is supposed to indicate which keys are currently down.]  I also don't 
recall the details of how that data is processed and passed to the X 
event queue, which could also lose the control key up transition.

Michael