bug#25851: 25.2; GTK warning when starting Emacs when desktop file has more than one frame

[Eli Zaretskii <eliz@gnu.org>]

> Date: Thu, 02 Mar 2017 12:00:22 +0100
> From: martin rudalics <rudalics@gmx.at>
> CC: nljlistbox2@gmail.com, 25851@debbugs.gnu.org
> 
>  > It's legitimate, but what we do as legitimate as well, right?  And
>  > emitting a warning in such use cases will only annoy users, right?
> 
> The doc-string of gtk_window_parse_geometry says
> 
>   * Note that for gtk_window_parse_geometry() to work as expected, it has
>   * to be called when the window has its “final” size, i.e. after calling
>   * gtk_widget_show_all() on the contents and gtk_window_set_geometry_hints()
>   * on the window.
> 
> and also
> 
>   * Deprecated: 3.20: Geometry handling in GTK is deprecated.
> 
> If this were our doc-string, would you say that calling that function as
> we do now is legitimate?
> 
> IIUC we call gtk_window_parse_geometry here because we cannot call
> gtk_window_set_geometry_hints at this early stage because of
> 
>    /* Don't set size hints during initialization; that apparently leads
>       to a race condition.  See the thread at
>       http://lists.gnu.org/archive/html/emacs-devel/2008-10/msg00033.html  */
>    if (NILP (Vafter_init_time)
>        || !FRAME_GTK_OUTER_WIDGET (f)
>        || FRAME_PARENT_FRAME (f))
>      return;
> 
> Aren't we desperately trying to shoot ourselves in the foot?

I'm sorry, I'm confused.  Earlier you explained that we do that
because users want the ability of placing frames outside of the
visible area, so I concluded that the fact we allow that is because we
want to cater to such users.  Now you seem to be saying that we
shouldn't cater to them?  How to reconcile these two?  Or am I missing
something?