Re: GNOME 2.0 Status and Martin's Travelling Schedule



jrb redhat com writes:

> > On 2 May 2001 jrb redhat com wrote:
> > > > Then create your UI, bonobo_control_new() it and call
> > > > gnome_selector_bind_to_control().
> > > 
> > > Woah!  Are you proposing mirroring every widget in libgnomeui as a
> > > bonobo control or something?  That sounds like a huge amount of bloat
> > > for very little gain?  What's wrong with just using widgets here??
> > 
> > 	There is no bloat there that cannot be pruned down, I don't know
> > that I'd personaly do it for every widget, but the promise of trivial,
> > working language bindings for every language with little effort is a nice
> > one.
> 
> My gut instinct is that the CORBA stubs will be the same size as all the
> widgets - there is hardly a lot of code in libgnomeui here.  

I think you're confused.

1 API (GNOME::Selector) + 10 widgets implementing it == 1 time the CORBA stubs.

There's no "bloat" at all, and even performance-wise, the difference between

    entry = gtk_combo_new ();

and

    widget = gtk_combo_new ();
    control = bonobo_control_new (widget);
    entry = bonobo_widget_new_control_from_objref (BONOBO_OBJREF (control));

is almost non-existant if it's in the same process.

> > > Why?!?!??  Why are we bonoboizing stuff in libgnomeui?  I'm confused
> > > here.  I don't think our widgets are complicated enough, and our
> > > component system light-weight enough for us to consider this.
> > 
> > 	There is no need per-se for a widget to be complicated in order
> > for it to be componentized IMHO.
> 
> I'm just talking about granularity.  A component per widget in GTK/gnome
> seems like overkill at this stage -- especially given that we are on a
> tight schedule and the GNOME API is fairly fluid.

I'm not proposing to replace any single widget in GTK/gnome.
I only want to do this where it makes sense:

        Gnome*Entry
        GnomeColorPicker
        GnomeFontPicker

etc.

The big advantage of having a widget like GnomeIconSelector a component is
to replace it with whatever we like at a later point; for instance Nautilus
could provide its own GnomeIconSelector and use it instead of the one in
libgnomeui, a printing application will most likely provide its own font
picker etc.

-- 
Martin Baulig
martin gnome org (private)
baulig suse de (work)




[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]