On Saturday February 07 2015 16:48:46 Jean-Baptiste Kempf wrote: > Yes, it is at fault. Hacking our buildsystem, and then attacking us, > because of that, is wrong.
MacPorts' overall approach is not wrong, that's what I meant. It may seem surprising to some that people might want to run FOSS software on a Mac (imagine an aureola over that word) without making it adapt perfectly to OS X guidelines but that doesn't make it wrong. It wouldn't still be around otherwise, with Apple's support at that. We'd love not to have to "hack" your build system (which doesn't strike me as the least complicated I've ever seen, but don't read a judgement in that), and be able to invoke it so that the build result corresponds to what we're looking for. > > Sedding, as in using sed to change things? That is indeed correct in > > certain well defined use cases to adapt a project to the MacPorts build > > system. > > This is not true. Clearly not. It is incorrect in all cases. > Where you INSTALL, post building, can be patched, sure, not sedding the > list of modules. Why do you insist on reading things behind my words? Adapting the list of modules shouldn't have to be done by editing the list of modules, I agree. But you know as well as I do that some install locations get built into the compiled code, so those have to be patched before building. > We have a very strict policy against insults. And there is a clear line Sigh. Do you also have a very strict and above all clear policy how insults are quantified? One that doesn't rely on everyone speaking his or her native language, with a neutral robot translator that cannot be accused of insulting when a translation goes awry or a way of saying something happens to be insulting in one language or culture and a compliment in another? > there. And you cannot claim someone insulted you if he did not. I'd say "You shouldn't claim someone insulted you unless you're damn sure he just did". There is a clear line somewhere, yes, but there's a gradient that leads to it from the "no insult" side. > Who's right and wrong is different. But comparably vague. > https://trac.macports.org/browser/trunk/dports/multimedia/VLC/files/patch-ffmpeg-2.4.diff Hmmm, that's a patch for VLC 2.1.5 which is debatable indeed, and which should never have been submitted here without proof that the newer versions were in fact supported after all (or without companion patches implementing that support). I don't know the case, I just didn't notice any issues with VLC 2.1.5 myself that could be attributed to running a too new libavcoded (I have 56.13.100). That's all I say about this. Oh, and that the patch is no longer included with port:VLC-devel . Has the limitation been removed? R. _______________________________________________ macports-dev mailing list [email protected] https://lists.macosforge.org/mailman/listinfo/macports-dev
