The musepack-tools are not available for ppc. Anyhow the mppdec utility compiles and runs fine on my Macmini. mppenc and replaygain are using x86 asm-code.
Created attachment 81192 [details] musepack-tools ebuild for ppc (mppdec only)
Created attachment 81193 [details, diff] Makefile patch
Thanks for the patches -- I'll see what I can do about getting the other tools to build. According to the musepack forum, the encoder can be built on OS X as well (it just doesn't have the optimized assembly, as you noticed) so it must also be big-endian safe http://www.musepack.net/forum/viewtopic.php?t=180
What you basically want to set is #define ENDIAN HAVE_BIG_ENDIAN at least, that's what i get from the source code (and what's defined w/ __APPLE__, other big endian arches are screwed, too) I'll have a shoot at it once i return home. :) Oh and musepack-tools were repackaged to mppenc, Diego made a new package for that. We should be getting our fixes there.
Yeah, compiles fine without the -D__APPLE__ thing, but produces crap output (mplayer can't play it). With that set, mplayer recognizes and plays the file just fine. Sound: can you take a look at my patch and include it in a -r bump that ppc can mark ~arch?
Created attachment 109714 [details, diff] Patch for mppenc-1.16.ebuild That's all that seems to be required to get it working on my machine.
I don't see why a new revision, as you didn't have the keyword there, please don't get people to rebuild a package just because. Anyway please note that musepack-tools would probably be deprecated as soon as upstream decides to release the sources for mppenc.
Not bumping is fine with me. However, we need the patch in mppenc first... I don't get your second paragraph, though. I thought mppenc sources were what we're talking about here?
Fix in CVS and keyworded ~ppc.
