On Nov 19,  9:53pm, maximum entropy wrote:
> >From: Jonathan Stone <jonathan@DSG.Stanford.EDU>
> >
> >It might be cleaner to move the check into the LK-201 [sic]
> >keycode-to-ASCII state machine and be done with it, as maximum entropy
> >suggests.  I think that will have to wait for after 1.3.
> 
> Is it too drastic a change for 1.3, or you won't have time to do it
> yourself?
> 
> Since I had already planned to do this, and it looks like doing it
> would fix Toru's problem as well, I'd be willing to make the change
> and test it by Saturday, if you're willing but just unable.

  I've got one working using lk201.c.  I'm using left-shift, left-compose,
and DO.  It's a little tricker doing it in lk201.c, since the code has to
track all the up and down transitions.  What I'm testing right now is
using a static variable to track the left-shift, left-compose, and
DO state.  When all three are down, I set a flag indicating that all
three keys were down at the same time.  When all three keys have been
released, the flag will cause Debugger() to be called.  I also had to
specifically clear the "shiftDown" variable (otherwise my shift was
reversed after exiting DDB).

  If no one has any objections to left-shift/left-compose/DO, I'll go ahead
and commit it after I test it some more [I want to test it on the 3100
as well].

Michael

-- 
Michael L. Hitch			mhitch@montana.edu
Computer Consultant
Information Technology Center
Montana State University	Bozeman, MT	USA