Jaguarandine said:
Sorry to disappoint everyone. I tried to do my part in getting a gp2x to the FreeDO developer, but whatever he has in store for the gp2x is in his hands. This has been a really important project for me for a long time and I hope to see it finished. Anyone who's seen Admin's latest comment in the FreeDO forums should know his intentions at least.
I don't have any new information so that's all I can say for now.
Admin (Felix L) is really mysterious to me. I don't think I've ever encountered an emu author less clear cut than he is.
On the one hand, let's take the good things:
- He was a lead developer in FreeDO itself and that obviously works.
- Posts on freedo.org show that he really knows what he's talking about when it comes to 3DO hardware and other consoles, and emulation in general.
And balance with the bad things:
- Several promised release dates have come and gone, for many ports. Excuses are really weird, like being funded by a commercial entity which is not ready to release it, or the controls aren't good enough. Never have there been screenshots, videos, anything, even though a video of the PocketPC version was promised.
- His explanations can be overly optimistic or simplistic compared to real world situations (like "just virtualizing" things such as ARM on ARM or SH2 on SH4 for Dreamcast, or 1:2 average recompilation ratio for ARM6 to SH4): some of these things might be doable but bear a lot of technical explanation to really be convincing.. and others might just be exaggerated.
- Some other things are kinda hard to believe, like getting a "Virtual Console license" to do 3DO games for it. I don't think it really works like that, and even if it did, 3DO is one of the least practical platforms to offer on Virtual Console since it wasn't very popular and the CD-ROM based games will eat up space heavily.
- Word is that FreeDO for PC is quite slow. The CPU is a 12MHz ARM, and while I know that takes quite a bit to emulate with an interpreter that shouldn't be bringing down all of the CPU time in a fast PC. Meaning that the emulation of the other peripherals must be taking a major toll, so how could Admin be so confident that this will run so well on a 200MHz ARM9 CPU? Writing optimized ARM code helps, but probably not by more than a factor of 2, and compilers don't do so badly with x86 code anyway. If he has such fast 3DO chip emulation then what has been holding FreeDO back all these years?
- Some of the technical things put forth have been really weird, like claiming that Linux uses up 40% CPU time on GP2X.
With most cases of emu dev it's usually pretty black and white, either the person is completely honest about it or they're flat out lying about everything. Sometimes it's somewhere inbetween, especially if the person has already done an emulator before. There was a case recently where word got out that someone had done GBA virtualization on Nintendo DS (using the MMU) and that it was much faster. I called BS on this, but after talking to some DS devs it did seem plausible. In the end the author admitted that only a couple games worked, with some glitches, and without sound, but it was running them beyond real time speed. The original claim was vague, but seemed to suggest more. We could call this an exaggeration.
I think that this might be the case here... actually I really don't know what to think. But I'm sure this must be very frustrating for a lot of people, something weird is definitely up.