Quote:I read that the Dolphin Emu was coming to PS3. Can anyone confirm this ?
Yes and no. Someone out there did start working on a port but they didn't get very far and apparently now the googlecode page is gone.
Keep in mind that even if a moderate sized team of good dedicated programmers got together and decided to make this:
1. There is no telling how far they would get. Just because somebody starts working on an emulator doesn't mean it will ever be in a usable state (a lesson that apparently the ps3 communities just can't seem to learn).
2. It would take years to reach a usable state at the minimum and several additional years to reach a good state. Most of dolphins code would have to be completely rewritten from scratch. We would be well into the lifecycle of the PS4 before the emulator reached a good state, and that's if we were extremely lucky and had a good team of programmers working on it almost nonstop.
3. It would run slow, very slow. The cell architecture is very poorly suited to applications like dolphin. Even if it uses a similar ISA the fact is the SPEs in the cell processor would be pretty much totally useless for something like dolphin and the PPE is even slower than a pentium IV at running control heavy code (which dolphin contains a ton of). Not to mention dolphin is not heavily multithreaded to top it all off. I wouldn't expect many games to run well even after it reached a good state.
4. Memory limitations are a serious concern that would require some genius level programmers to get around.
5. The gpu will likely struggle with anything past 2x IR.
etc.
Don't hold your breath waiting for good GC/Wii emulation on the ps3, you might pass out.
Quote:Start here Big Grin -> http://dolphin-emu.org/system-requirements.html
.....those are the minimum requirements needed to run dolphin.
In other news, I recently created a KDevelop project called "WiiUEmu". Guess we should post that on every emulation news site out there since it's about as important and meaningful as that ps3 emu.
Quote:In other news, I recently created a KDevelop project called "WiiUEmu". Guess we should post that on every emulation news site out there since it's about as important and meaningful as that ps3 emu.
Thank you. I couldn't figure out a way to point that out that wouldn't make me sound like an angry lunatic ranting from a basement somewhere. But it did really annoy me when a quick google search on the subject turned up dozens of popular ps3 news sites that wrote articles talking about it as if it was right around the corner and guaranteed to happen.
Getting the interpreter (if Dolphin has one) and the software renderer running under PS3 Linux probably wouldn't be to hard but the speed would be measured in seconds per frame.
Quote:Getting the interpreter (if Dolphin has one)
Obviously it does. What do you think that radio button labeled "interpreter" is for?
Quote: and the software renderer running under PS3 Linux probably wouldn't be to hard but the speed would be measured in seconds per frame.
Good luck getting it to work with 120-160MB of ram. The software renderer could actually run a lot faster on a cell processor than modern x86 cpus if properly optimized. The interpreter on the other hand would be even slower thanks to the PPEs wonderful in-order execution system.
But yes both of those would be fairly easy to port so I guess that would be a start.
Porting is much easier and faster than coding from scratch. An iOS game takes more than a year to complete, but it takes a team of 5 novice programmers 2 months to have an alpha working Android version. But I'm not sure whether it is the same in case of emulation programming.
The PS3 CPU hardware should be capable, but keep in mind that even Sony failed to emulate their own PS2 on the PS3. AFAIK it is due to the Cell architecture, which is no good for such thing. So I do not expect it much, although playing a game from a console on another console sounds fun.
(05-01-2012, 04:51 PM)Livy Wrote: [ -> ]Porting is much easier and faster than coding from scratch.
It's not really any easier if the emulator is so platform specific that you basically end up having to write all of the core pieces from scratch. And yes, that is the case for Dolphin in about every aspect.
(05-01-2012, 04:51 PM)Livy Wrote: [ -> ]An iOS game takes more than a year to complete, but it takes a team of 5 novice programmers 2 months to have an alpha working Android version. But I'm not sure whether it is the same in case of emulation programming.
Yo, games usually don't contain
too much platform-specific code. When porting a game you usually end up hiding all system calls (e.g. Direct3D calls) behind an abstraction interface (unless that's already the case), implement a new backend for that and bascically be done. Most games don't depend on stuff like CPU endianness, for example, which can be really awkward to port if there's next to no documentation which places of the code rely on which CPU endianness (which pretty much is the case for Dolphin).
tl;dr: yes, it's easy to port a game to another platform in a reasonable amount of time, porting an emulator boils down to rewriting most of it from scratch, though.
Fwiw, I'm kinda biased in this regard since I only know the Dolphin code base, other emulators might have a less HLE-ish approach and a cleaner codebase and thus might be easier to port.
As we've said, as the CPU is unsuitable, it would be pretty slow, and would take forever to code a new jit recompiler. I've read that a PS3 or XBox could barely run frostbite 2, so there's a severely cut back variant on the consoles compared to the pc version of battlefield. If it struggles with native code, it's unlikely to do well with emulated code.