Sorry for the lack of updates to sdlmame. Besides being in lab almost every waking hour i have gotten sucked into playing uqm in the little time i have for fun.

Dev bloggish stuff:
Part of the problem is deciding where to direct my efforts on this project. With my main test ROM, Ghosts 'n Goblins, the drawing routine was taking up a good fraction of the CPU cycles. I profiled gauntlet today and it keeps the emulator a lot busier, making the drawing a smaller fraction of the load (it autoskips about 50% of the frames, but is surprisingly good, considering).

That's out of total usage of about 66% for the sdlmame binary, with pulseaudio hogging up its usual 10-12% and the kernel and such using the remaining 1/5 of cycles.

So, it's not so obvious with that that GL ES is worth the time. GL ES is something i have poked at a bit but it is still not all that well supported in SDL in Maemo from what i gather. I also apparently managed to annoy the main SDLMAME dev with my newbishness as i worked on getting the initial maemo build ready, and he seems to ignore my posts now, so i'm not sure if he has started on GL ES himself. I can't blame him as i was floundering around with pretty basic stuff that was new to me. Anyway, it's not very easy to coordinate arm porting efforts with no communication and frankly not much of a motivation to start on any serious changes to the code that i probably don't have time for right now, anyway.

There are lots of UI things to be fixed and a new upstream version with a lot of fixes to save states, which is a key feature for a mobile phone emulator, so there are more immediate tasks that can be done, but i wanted to ask for input on priorities. I appreciate the efforts of testers to report their experiences with different ROMs. My question to you now is what would your top request be for improvements? I'll try to handle the easy and most desired fixes first.