Re: gok broken with new libwnck
- From: Elijah Newren <newren gmail com>
- To: Fernando Herrera <fernando herrera tecsidel es>
- Cc: Kjartan Maraas <kmaraas broadpark no>, desktop-devel-list gnome org
- Subject: Re: gok broken with new libwnck
- Date: Tue, 22 Feb 2005 09:09:54 -0700
On Tue, 22 Feb 2005 13:12:30 +0100, Fernando Herrera
<fernando herrera tecsidel es> wrote:
> Elijah: would this timestamp be ok here?
Actually, I don't know (I haven't looked close enough at the code).
In general, "current" time is dead wrong. See bug 166379, and how
gtk_get_current_event_time() caused problems for at least gedit,
galeon, and epiphany (and likely many others we're unware of). I did
use gtk_get_current_event_time() in a number of places in libwnck and
also in gnome-panel (the only other gnome module that uses libwnck),
but in each and every case I documented why it was okay to use it.
What is needed is the timestamp of the user interaction that caused
the wnck call--if there is no user interaction that caused it, a
timestamp of 0 should be used.
> PS: I know that #error "libwnck should only be used if you understand
> that it's subject to frequent change, and is not supported as a fixed
> API/ABI or as part of the platform" but these internal API changes 6
> days before hard code freeze are painful! :)
So are inherent bugs caused by incorrect code. ;-)
] [Thread Prev