Re: [gtk-list] Bug fix or feature addition?

[Elliott Hughes <Elliott Hughes genedata com>]

> I've noticed that when using a ctree and adding a signal callback
> for the selection signal, I don't get a signal for the default of
> the first item getting selected.  Since it is easier to keep all
> the code for selection changes in the signal callback, what I'm
> doing now is that after I load the ctree, I then manually select
> row 0 again.

> Does anyone else agree that this should be how it works by default?

as a short-term fix, maybe. in the longer term the solution to this
and similar problems (with row focus, for example, and reducing
the size of the clist widget's source to something more reasonable)
would be to separate out the selection model. it's a shame to have
to duplicate the selection information inside and out of the widget,
or to have to try and use the widget's data structure as your
application's data structure. separate selection models also allow
users to write new selection modes; Swing, stupidly enough, has
no equivalent of GTK's very useful `browse' selection mode, but
anyone can write their own in a few minutes. and use it for widgets
other than just JTables (the Swing equivalent of clist).

-- 
"Intelligence ... has caused our troubles; but it is not
 unintelligence that will cure them." -- Bertrand Russell