One of the things I noticed about other pi images that I ran that made me create DialogPi was performance issues. Some mame emulators and media centers were slow or did'nt load correctly or fit the screen just right. They said to be happy anything works...
So I started running test with compiled software with the goal to have fewer things but I want those things to load fast and "correctly". What I have been working on is getting one software hand tuned and compiled at a time and if code isnt ready for the pi, I don't try to run it.
It still has a ways to go and I can see were developers run into road block after roadblock. This stuff takes a ton of time and there isnt a whole lot of money in it. Right now Quake3 needs some work but the software is proprietary as are many of the OpenSource projects in linux.
Reguardless I'm looking into way to expand and fix and issues with dialogpi that I can find. Haveing developed software like OmxPlayer available from the repo makes life way easyer. If the pi had full opengles support and optimized software in the repo I could use the packages to fine tune a system.
Thats just it, I have been studying opengl and lots of code needs work being ported over to the pi and then optomized. Right now Qt5 and X still havent been hardware exlerated other then a few hacks but that still leaves large gaps of what will and will not work.
I look on to the future of the raspberry pi software, if and when the code is finally completed the Pi will make for a fast, low power, feature rich system that I'm happy to be a part of.
If you run DialogPi try to think Firefox, Gimp, possibly Blender, and all the possible softwares running like you would expect them to. Fast and clean. It's possible but there is still lots of work to be done. When proprietary the company can decide not to port over to opengles and there are other roadblocks "opensource" faces.
But when I look at the speed and power usage of what does work, I can't help but get excited!