Re: gtk_tree_view_get_visible_range

[Christopher Backhouse <cjbackhouse gmail com>]

> What do you exactly expect from gtk_tree_view_get_visible_range()
> when the tree view is not realized, has no size allocated,
> etc. (beside a more helpful error message)?  What is
> *visible* then?

I probably expect FALSE again but I suppose the current behaviour is OK

> The errors disappear if you do for example
>
>     GtkWidget *window;
>     window = gtk_window_new(GTK_WINDOW_TOPLEVEL);
>     gtk_container_add(GTK_CONTAINER(window), GTK_WIDGET(view));
>     gtk_widget_show_all(window);
>
> first.

Not for me they don't.
But further research shows that also adding
while(gtk_events_pending()) gtk_main_iteration();
just above the gtk_tree_view_get_visible_range does work.

Oddly the original full program is also fixed by doing this - presumably 
the problem function somehow gets called from a place where the list has 
just been modified so random important bookkeeping of some sort hasn't 
happened - still odd, and a more helpful error message, like "Your 
treeview is an inconsistent state, please don't call me from here" would 
have been nice.

Thanks for providing 4/5 of the code that made the testcase work.