Re: _NET_WM_[GET_|TAKE_|REQUEST_]FOCUS & urgency

[Elijah Newren <newren gmail com>]

On 5/25/05, Luke Schierer <lschiere users sf net> wrote:
> On Wed, May 25, 2005 at 05:42:22PM -0600, Elijah Newren wrote:

> > Yes.  And we needed a concrete way of doing that across toolkits and
> > applications (i.e. some implementation details about how to achieve
> > this), so Lubos came up with _NET_WM_USER_TIME (and related stuff,
> > like DEMANDS_ATTENTION) in the spec.  ;-)
> 
> Right. and Etan and Ethan found _NET_WM_USER_TIME to be insufficient,
> hence "* While it may be valuable to specify or further specify how
> this should be done, _NET_WM_USER_TIME in it's current overloaded state
> does not seem to us to be a viable solution for this."  I anticipate
> that you will respond that this statement is vague.  It does
> continue, giving an idea for an alternate method.

Pardon me for being rude, but why exactly should I care that two
people whom I don't know and have never heard of have found
_NET_WM_USER_TIME to be insufficient?  Not a single example has been
provided where it isn't enough to provide some kind of expected
behavior, nor have any examples been shown where its use would
preclude the ability to provide correct or desirable functionality. 
Off-handed/unsubstantied comments like this aren't helpful.

Now, it may well be that Etan and Ethan are X11 experts and geniuses;
they are likely smarter than me.  But without some kind of
substantiation it is impossible to differentiate their claims from the
standard user who sees any old bug in a program they are using and
shouts "This program has a crappy design; it is clear that we need to
rewrite this from scratch" rather than providing information on
tracking down and actually fixing bugs.  Saying it's
insufficent/crappy/dumb or whatever is fine, as long as concrete
reasons for its shortcomings are provided.

Also, as you noted, merely suggesting "something like a
_NET_WM_[GET_|TAKE_|REQUEST_]FOCUS property" is pretty vague, but the
problem is not so much that all the details aren't spelled out as that
it doesn't even provide enough details for me to see how it could work
as an alternate method.  It isn't at all clear to me that this
proposal is sufficient (e.g. I don't see how it possibly could handle
application launch cases), or even innocuous (it appears that it might
be placing policy in apps instead of in the WM--something we've run
into a big hurdle with in TAKE_ACTIVTY/MOUSE_ACTION).  At least a
little more detail really would be necessary, but first you'd have to
convince us that the current spec has some kind of problem or
shortcoming or there's no way we'd want to dump it for something else
that hopefully-does-the-same-thing.

> I absolutely understand not implementing something without
> understanding a need for it.  We rather came to this
> (DEMANDS_ATTENTION) from the *other* side having needed, looked for,
> and found urgency before DEMANDS_ATTENTION was written.

I'm really still confused why you think you need to do anything with
DEMANDS_ATTENTION.  But, considering the confusion, I think we
*really* need to clear it up in the spec.  Maybe Lubos can comment
here since he's the one that introduced the new state...


Thanks for your time and thought on this issue.

Cheers,
Elijah