(In reply to comment #0)
> Got the following build errors in jhbuild:
> 
> vnc-display.vala:101.24-101.30: error: Argument 1: Cannot convert from
> `Vnc.Display' to `Gtk.Widget'
>         window.remove (display);
>                        ^^^^^^^
> vnc-display.vala:103.9-103.23: error: Return: Cannot convert from `Vnc.Display'
> to `Gtk.Widget'
>         return display;
>         ^^^^^^^^^^^^^^^
> vnc-display.vala:114.16-114.33: error: The name `get_pixbuf' does not exist in
> the context of `Vnc.Display'
>         return display.get_pixbuf ();
>                ^^^^^^^^^^^^^^^^^^
> vnc-display.vala:29.21-29.27: error: Argument 1: Cannot convert from
> `Vnc.Display' to `Gtk.Widget'
>         window.add (display);
>                     ^^^^^^^
> vnc-display.vala:30.9-30.23: error: The name `realize' does not exist in the
> context of `Vnc.Display'
>         display.realize ();
>         ^^^^^^^^^^^^^^^
> 
> Probably something due to changes in vala. A few other packages were also
> broken.

Changes in vala bindings for gtk-vnc more like but yeah.

FYI, a few more:

/home/mcatanzaro/jhbuild/src/gnome-boxes/src/box-config.vala: In function ‘boxes_box_config_load_property’:
/home/mcatanzaro/jhbuild/src/gnome-boxes/src/box-config.vala:132:10: error: incompatible types when assigning to type ‘GValue’ from type ‘gchar *’
                 value = keyfile.get_string (group, property_name);
          ^
/home/mcatanzaro/jhbuild/src/gnome-boxes/src/box-config.vala:134:10: error: incompatible types when assigning to type ‘GValue’ from type ‘guint64’
                 value = keyfile.get_uint64 (group, property_name);
          ^
/home/mcatanzaro/jhbuild/src/gnome-boxes/src/box-config.vala:136:10: error: incompatible types when assigning to type ‘GValue’ from type ‘gint64’
                 value = keyfile.get_int64 (group, property_name);
          ^
/home/mcatanzaro/jhbuild/src/gnome-boxes/src/box-config.vala:138:10: error: incompatible types when assigning to type ‘GValue’ from type ‘gboolean’
                 value = keyfile.get_boolean (group, property_name);

<zeenix> Lethalman: shall i just re-assign that to you?
<zeenix> you -> vala
<Lethalman> zeenix, yes please

Created attachment 267406 [details] [review]
fix vala method call that throws error

This should fix the second issue, but I'm not sure about the reporter issues. Are you able to cook a simple test case?

(In reply to comment #4)
> Created an attachment (id=267406) [details] [review]
> fix vala method call that throws error
> 
> This should fix the second issue,

It does.

> but I'm not sure about the reporter issues.
> Are you able to cook a simple test case?

jhbuild build gnome-boxes? :)

Ah gtk-vnc provides its own vapi so that is not vala's fault then.

I'm not seeing why this is reported as a gtk-vnc error. The code in question is auto-generated by the vapigen from the introspection data, so I'd say the flaw is the code generator

(In reply to comment #7)
> I'm not seeing why this is reported as a gtk-vnc error. The code in question is
> auto-generated by the vapigen from the introspection data, so I'd say the flaw
> is the code generator

Since the binding is generated and shipped by gtk-vnc, first thing would be for gtk-vnc maintainer to check where exact the issue is. :)

(In reply to comment #8)
> (In reply to comment #7)
> > I'm not seeing why this is reported as a gtk-vnc error. The code in question is
> > auto-generated by the vapigen from the introspection data, so I'd say the flaw
> > is the code generator
> 
> Since the binding is generated and shipped by gtk-vnc, first thing would be for
> gtk-vnc maintainer to check where exact the issue is. :)

And it turns out that its the GIR that is wrongly generated:

   <class name="Display"
           c:symbol-prefix="display"
           c:type="VncDisplay"
           parent="GObject.InitiallyUnowned"

Maybe some missing gir annotations? or deps?

commit 423495e6e1cbf5936639e6b68c6cefe4fdadd368
Author: Daniel P. Berrange <berrange@redhat.com>
Date:   Tue Jan 28 15:37:16 2014 +0000

    Add missing dep on Gtk introspection GIR
    
    https://bugzilla.gnome.org/show_bug.cgi?id=723009
    
    Signed-off-by: Daniel P. Berrange <berrange@redhat.com>

(In reply to comment #10)
> commit 423495e6e1cbf5936639e6b68c6cefe4fdadd368
> Author: Daniel P. Berrange <berrange@redhat.com>
> Date:   Tue Jan 28 15:37:16 2014 +0000
> 
>     Add missing dep on Gtk introspection GIR
> 
>     https://bugzilla.gnome.org/show_bug.cgi?id=723009
> 
>     Signed-off-by: Daniel P. Berrange <berrange@redhat.com>

Nope the gir still shows wrong parent.

commit db462e6f919da758fbb2f4e7b3cbbf2897ec852b
Author: Luca Bruno <lucabru@src.gnome.org>
Date:   Tue Jan 28 20:34:47 2014 +0100

    Fix regression in method calls that throw errors with Value target type
    
    Related to bug 723009

Aaron and Michael, could you guys see if Daniel's patch fixed the build for you? It didn't work for me for some reason.

Still getting the same error.

I'm also getting a couple thousand lines like this:

/home/aaron/Documents/progs/gnome/gnome-boxes/libgd/<stdin>:1: syntax error, unexpected LONG in '# 1 "<stdin>"' at 'long'

I'm not sure if this was happening the other day or not. I might not have noticed, because it doesn't halt the build.

It works for me

(In reply to comment #14)
> /home/aaron/Documents/progs/gnome/gnome-boxes/libgd/<stdin>:1: syntax error,
> unexpected LONG in '# 1 "<stdin>"' at 'long'

Stuff like that has been happening for every module that uses gobject-introspection for roughly a month now.

Huh. Build Succeeded today for me.

I'm going to go ahead and close this. Gnome 3.12 is out now, so obviously they've got the build working.