Amiga.org
Amiga News and Community Announcements => Amiga News and Community Announcements => Topic started by: Pyromania on February 19, 2011, 05:13:35 AM
-
News via amigaworld.net from Troels
Update on the bounty for a JIT enabled version of E-UAE for AmigaOS4.1.
The bounty has been assigned to Álmos Rajnai. Álmos is the author of the JIT 68K emulator in AmigaOS4 (Petunia)and undoubtedly highly qualified for the job.
Our current version of E-UAE on AmigaOS4.1 (and also MorphOS) supports only interpretive emulation and is quite slow compared to f.ex. winUAE.
Gaining JIT 68K code emulation (like Petunia) will speed things up. Especially AmigaOS4.1 users with slower machines should gain a lot from a JIT version of E-UAE. The JIT version will automatically become open source and other PPC OS will hopefully be able to benefit from the work as well.
As usual donations can be made via the Amigabounty.net website: JIT enabled E-UAE bounty
-
Rachy Rulez 4ever!!!
-
Great news for those using ppc machines :)
This could potentially be fairly big news to the emulation world in general, not just amiga. Even though I'm mostly an x86 AROS and 68k classic guy, it'd be nice to see an amiga developer get a tidy little sum for some of their work. That is of course assuming the general ppc population chips in a few donations to the bounty here and there :) This could even be a chance for OS4.x to get a little free publicity, posting genuinely relevant information to various emulation sites. Xbox360, ppc linux, Wii, PS3(?), Mac, and whatever other devices that can run uae using a ppc architecture could be potential targets for this news, which by default attracts attention to OS4.x. It's hardly billboard and mass commercial marketing, but it surely couldnt hurt :)
-
I sure hope he has learned since Petunia and is not going to write it in assembler... You get more benefits in actually optimizing the high level algorithms rather than writing everything in asm, as is proven by the fact that Trance is significantly faster than Petunia. Trance is written in pure C.
But welcome news, regardless.
As for this being any kind of advertisement for any platform: forget it. x86 has had their JIT for ages, they're not going to bother with some PPC platform just because it 10 years later finally gets JIT. Since the code will be GPL it will not be locked to OS4, and thus there is no "lock in" factor either. Why would you obtain a 1000+ USD machine just to run this code when you can run it on your PS3/Wii/Mac just as well, and even better?
Now that I think about it, if I were to do it I'd actually write the JIT on Linux or Mac OS X and then backport it to amigoid systems. Lack of memory protection and some badass debugging tools (valgrind) would be just pain it the arse.
-
Might not convert x86 lovers to PPC, but this is great news for PPC loving crowd.
Better late than never as they say... ;)
-
Wow!
This is fantastic news! I've been waiting for this.
-
I sure hope he has learned since Petunia and is not going to write it in assembler... You get more benefits in actually optimizing the high level algorithms rather than writing everything in asm, as is proven by the fact that Trance is significantly faster than Petunia. Trance is written in pure C.
But welcome news, regardless.
As for this being any kind of advertisement for any platform: forget it. x86 has had their JIT for ages, they're not going to bother with some PPC platform just because it 10 years later finally gets JIT. Since the code will be GPL it will not be locked to OS4, and thus there is no "lock in" factor either. Why would you obtain a 1000+ USD machine just to run this code when you can run it on your PS3/Wii/Mac just as well, and even better?
Now that I think about it, if I were to do it I'd actually write the JIT on Linux or Mac OS X and then backport it to amigoid systems. Lack of memory protection and some badass debugging tools (valgrind) would be just pain it the arse.
Im not suggesting it'll convert anyone from x86 to ppc, or from anything else to OS4, but the simple fact is to the ppc emulation world this is potentially pretty big news, ergo it'd be relevant news to post to quite a few sites, ergo a little extra public awareness of OS4, even if just in reference to when the GPL 68k emulation core is coming from. It's hardly going to change the world, but mass exposure does often result in a few extra curious onlookers.
Thinking about it though, and Im not just saying it cos of who I directed the post at, perhaps a little ironically it could even benefit MOS a little too. People interested enough to be intereted in amiga emulation on a ppc mac, may find out about MOS while researching the "new" amiga, which in turn may find them trying it out if they have a supported model.
Personally it doesnt really have much influence on me (being a x86 AROS and 680x0 amiga guy), but I think this is something people could jump on to give a little exposure to the amiga outside the usual places.
-
This is great news! I was holding off because it looked as if no one would ever do it, but just donated 12 euros wish it could be more.
Plus I agree the effort will not just befit AmigaOS4 if done well, but every PPC platform, so you MorpthOS guys get donating! :)
-
A bit off topic, but people who are using Uae on PS3 and wondering if they can run executable files on it, the answer is yes. Just rename your file to end with .adf and it will load just fine.
Back to topic ;)
-
If this means that the code will then, without a doubt, be carried on to MorphOS, I will make a contribution, otherwise to hell with it. Not because I have anything against OS 4.x. After all it can run OctaMED S.S., it's just that I have no desire to fork out the kind of cash it costs to own an X1000 or any other sort of high end computer to run the OS on.
I'd happily run UAE on AROS, but it seems that 30% of the time when I exit UAE, AROS crashes along with it (well that and a number of other unAmiga like issues), thus taking away all the fun. When AROS becomes half as stable as MorphOS, I'm on it!!!
-
wow! great news indeed:pint: