> > So, I tried a quick hack in adbhid_input_keycode
> > (drivers/macintosh/adbhid.c)
> > to translate the events appropriately and it seems to work (at last!).
>
> Don't just tease, please post a patch, and we'll love you forever :)
Caution with the 0xff response byte - I think I've seen that as first byte
for a number of 'sporadic' ADB events as well, such as regular interrupts
from the ADB controller once a second (on other ADB hardware). Plus you
sure noticed the second response byte is also 0xff in most packets (except
when the keyboard has lots of events to report). Please add some debug
code first that triggers each time the first byte of a keyboard packet is
0xff, and check if there are any other instances of this reported.
If it works (on all Macs) we could get rid of atong's hack for good, and
turn the capslock state detection into a regular toggle (which is already
handled by the current code I think).
Michael
