On Wed, Oct 11, 2017 at 12:49:20PM -0400, Gary Allen Vollink wrote:
> IMO, key behavior has everything to do with terminal behavior, and
> nothing to do with X.
"Nothing to do with X" is a pretty strong statement for code that
uses KeySyms, Mod*Mask, and the hardcoded keysym values 0xFD00..0xFFFF,
and which is called exclusively from the handler for the X KeyPress
event. :)
I could see not wanting to move the *entire* function... say, leave
DECPAM/DECCKM tests in the terminal module, and map KeySyms to byte
strings in the X module. What I don't see is a way of doing that which
is equally simple and still removes st.c's dependence on X types. (If
you have an idea, fire away!)
> I think it makes supporting something other than X (at some point)
> harder instead of easier.
My intuition says the opposite, that excising X-specific types and
values from st.c should make it easier to reuse st.c for things which
are not X. Of course, we could always re-refactor if that turns out not
to be the case. :)
*dp
--
<><
Received on Thu Oct 12 2017 - 05:08:15 CEST