OSNews

Such things can only happen if one "side" yields to the other side's technology, or they come up with a common technology they can both agree on and abandon their old technology.

This sort of collaboration or surrender doesn't happen often for various reasons, but when it does, both sides benefit. GNOME wouldn't accept DCOP, but they would accept DBUS which is based on many of the ideas of DCOP, except that it's binary. KDE seems to have some commitment to DBUS too (at least through a DBUS-DCOP bridge), so DBUS has become a common standard. The reverse story happened with GStreamer versus aRTs. Other freedesktop standards like the window manager spec, the menu spec, and the mime spec, went through similar consensus decisions or surrenders.

A while back, there was some talk about GNOME and KDE sharing the same main loop. I don't know if it went anywhere, but if it did, Gtk+ and Qt+ apps could be in the same process space and share resources making each feel less bulky in the other environment.

One exciting aspect of this is that it would allow GNOME to build the equivalent of a KParts (called GParts). According to http://gparts.blogspot.com/, some work was started, but judging from the date of the last entry, the project seems to have been stalled. It's a shame. If GParts were finished, KDE and GNOME would have a common plugin spec that would allows KParts to be embedded in GNOME and GParts (Beagle, Evolution Outlook Integration, etc) to be embedded in KDE or KParts be exposed as a Bonobo CORBA object. Both environments could then more easily leverage off each other's work.

Does anyone have any idea if GParts is dead or not or if the common main loop ever got committed to GNOME and KDE?