Well, it's good to see that those who actually knows how Hildon works is doing the work (and know what sins they committed while developing it) - hope something useful comes out of it. And I wonder how GTK+3 fits in all this. My worry is that it doesn't include enough platform support to make most maemo.org extras applications work, which was what turned me off the task at first (I actually had Hildon working on stock GTK+ with wazd's Marina, on MeeGo already) The weight of supporting Maemo API is hard - so there is bound to be need for rewrites in applications, which may just motivate people to rebuild applications in Qt+QML instead.