Re: [Deskbar] Code freeze break for deskbar-applet
- From: Raphaël Slinckx <raphael slinckx net>
- To: Vincent Untz <vuntz gnome org>
- Cc: release-team gnome org, Deskbar Applet List <deskbar-applet-list gnome org>
- Subject: Re: [Deskbar] Code freeze break for deskbar-applet
- Date: Mon, 04 Sep 2006 15:03:28 +0200
Hey !
Thanks for your replies, since I didn't know if I would have time before
the release deadline to apply the patches and make the release (cause
today i'm a bit busy), i just went ahead and comitted plus released
deskbar 2.16.
That said, I know i should have waited, and i'm sorry for what I did. I
did it in good faith :)
Maybe i can release a 2.16.0.1 that follows your suggestions ?
Here are my comments:
> > Bug 353605
> > I don't see how this is release critical, though it's possible I'm
> > misunderstanding the bug. It just looks like a minor cosmetic
> > problem, and as such something that we should not break freeze for,
> > especially so late. While the patch is very simple, I'd still prefer
> > it were reverted unless this actually causes a severe usability issue.
>
> Agree. This is something you can commit after 2.16.0 and so you'll get
> some testing for 2.16.1.
>
Agreed, i can revert this and publish the new .0.1 tarball without it
> > Bug 354014
> > Does this cause any severe issues? I know the patch is simple, but
> > being so late into code freeze I really don't like seeing changes like
> > this. Can you explain more about the user impact before releasing
> > with this patch?
This is really an important fix that will cause an error to be raised
whenever the code is reached. Basically we have a hidden gconf key that
will clear up the deskbar entry for privacy reasons after each search.
The code without the patch overwrites a instance method called
self.clear_entry with a boolean, any invocation of self.clear_entry()
will result in an error cause it isn't a method anymore.
Another problem in the logic of the code this time, the "if
self.clear_entry" check will always return True because self.clear_entry
is a method, and thus != None. so the if is equivalent to "if True:"
which is not what we want by default (clearing the entry).
I hope I was clear enough.
Now if the unknown patch is accepted, maybe we could just let th e
cosmetic one and avoid doing a re-release ?
Thanks !
Raf
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]