On 2013-7-31 00:49 , Landon J Fuller wrote: > > On Jul 22, 2013, at 10:58 AM, Vincent Habchi <[email protected] > <mailto:[email protected]>> wrote: > >> Exactly. The idea I had is that the users that want their ports to be >> compiled and optimized for their machine should add the +perf variant >> in variants.conf. Since +perf is never enabled by default, this would >> never result in the fetching of the pre-compiled binaries. What do you >> think?
It should clear out archive_sites to be correct, since we want to build some non-default variants in future, and third party archive sources can provide whatever variants they want. > I'm concerned that the net performance gain here is going to be far > outweighed by the maintenance costs and user complexity. Without > actually measuring the gains, there's no guarantee that these changes > will actually improve performance, and in moving away from the > optimization flags et al selected by the original developers, could very > possibly trigger bugs in code generation output (e.g., compiler bugs, > especially with bleeding-edge clang), or reveal bugs in the project that > aren't apparent at lower optimization levels. Certainly this variant should be restricted to ports that demonstrably *need* maximum performance for some class of users, do in fact gain significant performance with those settings, and still pass their test suites. - Josh _______________________________________________ macports-dev mailing list [email protected] https://lists.macosforge.org/mailman/listinfo/macports-dev
