> Can you explain what could go wrong? The predicate would
> be passed in
> the KeyRelease event (as an XPointer), and return True if it finds
> a KeyPress with matching time and keycode.
Sorry, there is no problem. I didn't quite grasp the arbitrary predicate
part before.
I am skeptical about whether ignoring repeats is worth it.
All it does is fumble your combo, and that's only if you hold the keys for
more than half a second. Not really a big deal. If it was just a matter of
testing some ev.isrepeat flag (which it should be), I would fix it,
but it's a matter of a new function and global variable.
I don't think it's worth it. Maybe I'll do it as an academic excercise.
As for general nonblocking peek, a predicate that is always true would do it.
I've figured out hg diff now, so I've attached the patch.
Heres a summary of it:
I found multitagging by toggle to be a bit tedious, so this patch changes the
interface so that if you mash multiple tag keys like a key combo, it tags or
views with those tags. Lifting a key ends the combo. Once you apply it,
replace tag with combotag and view with comboview in config.h.
It used to have trouble if the current tag was in the combo, but works now.
It also didn't used to behave properly with the tagset toggle (MOD-Tab).
It behaves properly now; the new functions (combotag and comboview) are
completely backwards-compatible to the default tag and view functions, so you
can replace all instances with no loss. If you hold the combo too long, it will
fumble it and just select the last tag you hit.
Have fun!
This archive was generated by hypermail 2.2.0 : Thu Sep 23 2010 - 04:36:02 CEST