14 replies [Last post]
josep's picture
User offline. Last seen 12 years 19 weeks ago. Offline
Joined: 05/10/2008
Groups: None

Is there a fundamental reason for not being able to compile the Mac version to a PowerPC processor?

(e.g., is there any dependency on the particulars of intel processors?),

If not, I suggest a Mac version be made available for PowerPC´s (a universal binary is easily obtained using Apple´s tools).

Joined: 05/11/2008
Groups: None

It would be nice if there were one. I would love to play but am unable because of my powerpc mac. I would like to know if it's because the powerpc macs are just too slow to handle the game. I don't think there's anything inherent in the architecture.

Joined: 07/20/2008
Groups: Powerfolders
Power Pc

It may well be partly because of speed.  On a mac Pro, when it's doing lots of things, even it runs a tad slow.  Try to run it on a much slower power Pc...  But still, it would be awsome if there was a Power PC Fold it version.

Joined: 10/03/2008
Groups: None
ppc interest

I would also like to try a PPC version if one could be made.

Joined: 09/26/2008
Groups: None
I would too. I imagine,

I would too. I imagine, however, the lack of PPC version is more a function of extra development time required.

the4js's picture
User offline. Last seen 50 years 39 weeks ago. Offline
Joined: 10/12/2008
Groups: None
unable to use

yea no use for the 4 ppc's in our houe, my 5 year old is set to win the nobel prize! No luck for her only one machine running intel and its not for the kids. Bumer

Joined: 10/24/2008
Groups: None
im keen and dont mind the speed hit

my home PPC Mac hardly does any work - I would not mind doing this out of sheer curiocity

Corin's picture
User offline. Last seen 50 years 39 weeks ago. Offline
Joined: 11/18/2008
Groups: None
Me too

I'd like to see PPC version too.

Corin's picture
User offline. Last seen 50 years 39 weeks ago. Offline
Joined: 11/18/2008
Groups: None
admin info

I sent a message to admin, there response was to be expected (at least by bluskale)...

The response said that there's no plans at the moment and it's a question of development reasources.

It's a pity, I would have imagined that it's not so difficult to build a universal binary using Apple's tools...

axcho's picture
User offline. Last seen 8 years 20 weeks ago. Offline
Joined: 12/18/2007
Groups: None
I'm on a PPC right now

I'm on the dev team, and I'm using a PPC Mac to develop for Foldit! :p It seems to crash more frequently though.

I asked about it, and basically the reason we don't support PPC is because we're lazy, and we don't care enough to respect the desires of the minority. Wheeee! :p

Anyway, I really don't know. I'm not the person who makes those kinds of decisions. At least, I don't know enough about how the build and distribution process works to do it myself.

Yeah, I'm not sure. I guess the question is, can you present a persuasive-enough argument to the powers that be to convince them that it's worth the extra time and energy to build and debug and distribute for PPC?

Joined: 12/18/2008
Groups: None
I don't even own an intel

I find it rather silly that they don't just make a Universal Binary for people, and just alert the users that perhaps the PPC version isn't as stable? Or better yet, just put it on the Download For Mac link that it's Intel only, not just tiger only. I hate getting burned by companies and people like that, don't even bother to mention that they don't support ppc

Joined: 12/17/2008
Groups: None
Actually, several BOINC

Actually, several BOINC projects have similar isues. PPC architecture does not quite follow the same floating point procedures, and so projects were finding that x86 and PPC results were different when compared out several decimal places. Also compiling is a bit different. Unfortunately there is no utility that turns a PPC or Intel Mac binary into a universal binary. Both must be written and compiled, and the utilities fuse them with a wrapper that handles selecting the executable code. (Back in the old days when earlier Mac OS'es ran on different platforms, there were utiliies to "un-universalize" binaries, stripping out code that's not for your specific processor.) There is also the issue that no high quality and rigorously tested compilers for PPC architecture are free. Several projects have complained that the GNU compilers are good, but the code is bulky, sometimes inefficient, and occasionally buggy. Further, designing the program is different, as anything for a G4/G5 that is not designed for SIMD vector processing and AltiVec boost will be slower. PPC processors are designed to do lots of short steps in parallel, while x86 does less parallel work, but does more per thread in a given number of clock cycles. That's why the streaming processors in video cards have similarities to PPC architecture, and the Cell processor (PlayStation 3) uses a PPC-based core.

ClimatePrediction.net permanently dropped PPC development (partially because PPC and x86 results didn't quite match, more because of issues getting a good fortran compiler), and Docking@Home's PPC client is on hiatus until they determine if the work required to update the CHARMM engine on PPC is worth the return.


LizPf's picture
User offline. Last seen 12 years 20 weeks ago. Offline
Joined: 05/09/2008
Groups: None
Add Another Vote for PPC Version

I would love to be able to play/help, but I need to be on my PowerBook G4 for several reasons. If there is a way to create a PPC version, please do.

ivanm's picture
User offline. Last seen 12 years 19 weeks ago. Offline
Joined: 05/11/2008
Groups: None
me too

What about releasing an unsupported PPC version? I just want a chance to play with the damn thing.

And why not make it open-source? That would seem like a perfect match for such an academic project.

admin's picture
User offline. Last seen 1 week 1 day ago. Offline
Joined: 11/10/2007
Groups: vi users
Since this thread has a

Since this thread has a duplicate in the feedback tracker, I'm closing it.


Developed by: UW Center for Game Science, UW Institute for Protein Design, Northeastern University, Vanderbilt University Meiler Lab, UC Davis
Supported by: DARPA, NSF, NIH, HHMI, Amazon, Microsoft, Adobe, RosettaCommons