Re: Are Out-of-Tree Widgets Second-Class Citizens?

[Tristan Van Berkom <tvb gnome org>]

On Mon, Sep 28, 2009 at 9:36 PM, Brian J. Tarricone <bjt23 cornell edu> wrote:
> On 09/28/2009 06:11 PM, Allin Cottrell wrote:
>
>> I'm not sure if the solution Morten advocates -- namely, applying
>> the GSEAL principle internally -- is the best one, although maybe
>> it is.
>
> It's not even necessary.  Gtk can access struct members directly if
> desired (for performance reasons, or whatever), but the API should
> expose them for subclasser use as well.
>
> I doubt this was done "maliciously"; the hostility shown in this thread
> seems a bit silly to me.  I imagine some of these missing bits just got
> lost in the shuffle in the effort to clean up the public interfaces.
> It's not like the gtk maintainers are unreasonable; I bet a patch to
> expose the needed functionality would be accepted.

I think the best hope is to patch GTK+ case by case on a per
requirement basis, and report/fix breaks in applications that wont
compile with GSEAL sooner than later.

The biggest issue here is that these little issues need attention,
and exposing api in GTK+ is something you want to be careful
about in the long term, so adding api also requires attention and
work.

Cheers,
          -Tristan