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

[Elijah Newren <newren gmail com>]

On 5/25/05, Rob Adams <readams readams net> wrote:
<snip>

> My suggestion is that if you think that the notification for new gaim
> messages is insufficiently obvious

<snip>

> But, really, it all works perfectly for gaim as it is without any code
> changes

Actually, it doesn't quite work perfectly, I believe for the reason
you mention (the notification for new gaim messages is totally
non-obvious and gets missed by people for hours).  There's a couple
things we could do to fix this bug (though there are some current
technical difficulties with the "most useful" one).  But one of the
problems is that I'm not convinced I know all the issues.  Naturally,
we occasionally get complaints that the-whole-system-is-broken but
rarely are they accompanied with useful details (the generic complaint
that "the window doesn't get focus when it's supposed to" has been
tracked down to around a few dozen totally different bugs, varying
from case to case).  If I could get detailed explanations of what is
happening that people don't like, I can often find bugs and fix them. 
But anyway, were' now off-topic for this list.  I'd really like for
some gaim developer(s) to file a bug against metacity (even if turns
out it's a problem with other modules as well, like libwnck) and state
the exact behaviors (with steps to reproduce) that they see and don't
like.

<snip>
> The only case where an application needs to worry about any
> of this is when it launches new windows by communication out-of-process.

Actually, there's one other case--if an application launches a window
that was due to a not-so-recent user interaction with that
application, then it also needs to do a little work.  But that's rare
(there's only case I can think of for this--a user trying to open a
web page in one tab of a browser, switching to another tab and doing
stuff, and then getting an error message that the page in the first
tab couldn't be loaded).


Cheers,
Elijah