On Jan 11, 2012, at 11:02, Stephen Langer wrote: > > On Jan 11, 2012, at 11:09 AM, Ryan Schmidt wrote: > >> If enabling threads in tcl will not cause any problems, then we should >> remove the variant, enable threads all the time, and revbump tcl and all >> ports that use it to force a rebuild. >> >> But if there is still a valid reason why we'd sometimes want threads and >> sometimes not, then really we should have a tcl-threads subport within the >> tcl port, work should be done so that both tcl and tcl-threads can be >> installed simultaneously and not conflict with one another, and all ports >> that use tcl should decide whether they want threads support or not and >> depend on either tcl or tcl-threads. > > The right thing to do is for all libraries to allow threads to be enabled or > disabled at *run* time, not compile time. The same thing goes for libraries > using OpenMP. Since OpenMP is incompatible with pthreads on some systems > (including OS X), the decision to use pthreads or OpenMP should be made by > the calling program. > > I realize that this will be hard to do and it's completely unrealistic to > expect it, but it would solve a lot of problems.
Ok, so what is a realistic solution to the problem, then? _______________________________________________ macports-users mailing list macports-users@lists.macosforge.org http://lists.macosforge.org/mailman/listinfo.cgi/macports-users
