OpenGL, X11, KDE & GNOME

Is it possible to create a Super Project for Open X, that will combine and direct the development of OpenGL, X11, KDE & GNOME where we will not see duplicate effort for the same thing or if there any duplicate effor exists, will be merged to get the best of the best. Open X will provide an X-Server(lean and mean) Open GL based gaming and Application platform, and the best of both KDE and GNOME. We just need to syncronize our effort to avoid splitting dependancy problem of applications. Every two years will identify which sub projects needs more attentions and which is already invented. Thus Open X will invite new development ideas and submit them for developer attentions.
Syed Mamun Raihan
CTO, Great Lakes Software Research Institute.

The nice thing about keeping the windowmanager(KDE,GNOME and many others) different from the windowing system(X11) (besides the fact that they do different jobs and therefore it isn’t logical to put them together) is that you can have many different windowmanagers according to your needs, taste, etc. whereas with, say, Windows you’re stuck with one. So it’s not so much a matter of wasted duplicate effort as much as effort spent in different directions. KDE and GNOME for example are different and even more different are other windowanagers like enlightenment, blackbox, windowmaker, etc. I would really not like beeing stuck with just one, the best(best according to whom anyway?). And I don’t understand all this fuss about X lately. Sure X has accumulated some bloat over the years and I’d like a cleaner and smaller windowing system too. Yet everyone seems to be talking about speed and how many cycles X robs from his/her applications. Take a look at this . As you can see you shouldn’t expect any dramatic performance gains. In fact you probably wouldn’t notice a difference when using a cleaner, smaller X. Now programming the thing(xlib) is a different matter. Not that I have a big problem with it but some parts (like say the keyboard manager xkb) have become quite complicated(for example in xkb you have scancodes, keycodes and keysyms with different mappings between each other). So having a cleaner design would be generally nice but don’t expect any dramatic results.

It’s ok I guess that Gnome and KDE are setup differently but why can’t they just use one API for their apps. Why should apps work ONLY for one or the other?

The different API’s is the strength of the X11 project. You can program for whatever API you feel comfortable with. In time, this leads to more apps, as more persons feel comfortable with programming for the X11 server.

This has nothing to do with the window managers themselves. You can use, say, a Gtk-based app on KDE and should have no problem (I suppose, apart from the fact that the themes might not match). So your problem is more in the availability of many widget libraries. As lulf said this is no problem really (in fact I’d like to have another simpler, smaller and cleaner library wich maybe you could count on, on being installed on every system).

This topic was automatically closed 183 days after the last reply. New replies are no longer allowed.