Re: 2.3 Proposed Features
- From: Luis Villa <louie ximian com>
- To: Havoc Pennington <hp redhat com>
- Cc: Ettore Perazzoli <ettore ximian com>, Christian Meyer <chrisime uni de>, Glynn Foster <glynn foster sun com>, aes gnome org, jdub perkypants org, desktop-devel-list gnome org
- Subject: Re: 2.3 Proposed Features
- Date: 04 Feb 2003 02:20:10 -0500
On Tue, 2003-02-04 at 02:23, Havoc Pennington wrote:
> On Tue, Feb 04, 2003 at 02:07:59AM -0500, Luis Villa wrote:
> >
> > As far as I can see, the GNOME experience with libegg so far is that
> > unstable API means unstable user-visible experience, because things get
> > imported unevenly and updated unevenly, and that is if/when they are
> > imported at all. So, I'm well aware of the distinction you're
> > describing, but so far the /practical/ distinction between the two
> > problems is small. Or so it seems to a naive but slightly technically
> > skilled user. :)
>
> Well, we do have one other model for doing uncommitted API, which is
> the libgnome-desktop/libwnck/libstartup-notification model. That one
> is also an option.
>
> My feeling on cut-and-paste is that it can never be worse than every
> app implementing its own thing separately. It beats wheel reinvention.
> Sure it kind of sucks, but remember that a shared lib with
> soname-of-the-week kind of sucks too.
<nod> I'm not really arguing specifically for one model or another- I
don't have the experience outside of the past year to really know what
works best there. I just feel very strongly we have to settle on a model
soon, we have to settle on one that guarantees either ~0 or ~100%
adoption for GNOME 2.4, and we need to work out the plan to make that
happen. It seems clear that 'get it in gtk' is not that model, so...
where do we go from there?
Luis
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]