There are however two things that I personally feel are keeping PF from really competing with the other libraries out there. Yes, I know other libraries perform better, have better support and are generally accepted as better, but I think PF is still certainly sufficient for many projects, and in some ways better (source code availability and simplicity chief among them).
The two things I refer to are a custom image loader and a switchover to DirectX on the desktop.
The first is actually a pretty major problem. As I understand it from reading the GapiDraw documentation, imgdecmp.dll is very inefficient in its memory usage. I totally believe this is the case based on the fact that my Invasion: Trivia game requires about 14M to run, and that's about twice as much as I calculate it should!
GapiDraw uses a custom image loader, and I believe PF has to do the same to continue to be useful. The problem is, I'm not sure I can do this (firstly because I don't have the time to put into research, and secondly because I may just not be capable of it).
Secondly, the fact that PF uses GDI was an excellent design decision on Thierry's part since it allowed it to work on a wide variety of devices, but now with DirectX being ubiquitous on the desktop, I think it has to switch over to that. GDI performance is obviously not good enough for anything but relatively simple games, so PF remains relegated to a strictly PocketPC usage, which it really doesn't have to.
I personally am now considering using GapiDraw for my next project, in fact I've started down that path already. For a variety of reasons though I'd prefer to stick with PF. Unfortunately, these two things are the main reasons I probably won't.
The point to all this? Simply put, is there anyone out there that has the time, skill and desire to make these two things happen? Even if it was just the image loader thing I think that would be a major improvement. I know there are open-source image loaders out there that might be useable, but I think there are others more qualified than me to tackle that job. Any takers?
If not, I would expect one more PF revision at some point in the future from me to address the relatively minor things left that I want to address, and then that may wind up being it, aside from bug fixes. Don't take this to mean I'm dropping PF support right now, that's not the case, but I frankly think the writing is on the wall if I myself am starting to use GapiDraw
