I wanted to investigate how much effort it would be to be able to use
emacs in all places in KDE where its "text editor kpart" is used.
Therefore, emacs (or an emacs frame) would need to implement this kpart
interface. Currently, I don't know if and how that relates to XEmbed.
Possibly, it is a completely different approach.
When googling for emacs kpart, I found some few threads on
[x]emacs-devel and kde lists, where people were interested in
implementing such a emacs kpart. Unfortunately, except comments like
"seems possible" the projects never seemed to get actually tackled.
I don't have much experiences with C/C++, and also my knowledge of
KDE/Emacs internals is nearby zero. But maybe that would be a good
topic for a summer of code project mentored by an emacs as well as a KDE
dev. I think I remember that Chong's and Stefan's plans for emacs 24
also include destop integration as a goal.