Re: [dev] [dwm] tagging interface

On Thu, Oct 07, 2010 at 05:44:50PM +0100, Ethan Grammatikidis wrote:
>
> On 7 Oct 2010, at 1:44 am, Wolf Tivy wrote:
>
> >Ok, I'm not sure I understand you on this one. What you are talking
> >about (killing apps) is not affected by my patch. If you mean that we
> >should filter repeats globally in dwm, not just in this case, then
> >I don't think it should be done. That feels like abstracting and
> >re-interfacing the OS, which does not belong in dwm. The default
> >behavior in X is key repeats, so users should expect repeats when
> >they hold keys. Period.
>
> What should we do for programs which need controls which don't repeat?
> Make a new kind of input device which looks just like a keyboard but
> which isn't because keyboard keys repeat, by definition?

Just a technical note. As we discussed in this thread originally, there
(appears to be) a reliable test to detect individual repeated keys.
This test would not affect the repeat of other keys handled by dwm, nor
would it affect the repeat behavior of managed windows.

The tradeoff here is: adding the change would increase lines of code,
but would save cpu cycles. As things are now, repeats arrive as
KeyRelease/KeyPress pairs. In some situations, these have absolutely no
effect (for instance, activating a tag once it is already activated).
This test would prevent spending undo processing on them.
Received on Thu Oct 07 2010 - 23:46:50 CEST