On 11/20/2012 12:35 PM, Ingo Weinhold wrote:
On 11/20/2012 12:44 AM, Landon Fuller wrote:On Nov 19, 2012, at 6:17 PM, Ingo Weinhold wrote:PS: What I'd really like to see is support for LLVM bytecode binaries. Why fat when it can be portable?Me too, but one of these things I'm capable of doing in a reasonable amount of time (end of Thanksgiving weekend, if things keep going smoothly?). The other, not so much :)Fair enough. If you want to implement FatELF support, just do it. Really. Even if we don't adopt it now, having it around as an option would certainly be nice (be sure to keep it in a public repository). If the runtime loader changes will be as small as I think they will be, we can probably incorporate them anyway. For the rest it might be a good idea to see how package management turns out first.
I think that package management and FatELF support are two diametral mechanisms. For everything that uses our package manager, FatELF doesn't really make much sense IMO.
I would consider FatELF support an optional feature that software distributors may choose if they want, or if users demand it enough. Who knows. I'd assume the necessary changes will be rather small, so it shouldn't hurt to support it. No one has to use it, anyway.