http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41495
--- Comment #9 from Jonathan Wakely <redi at gcc dot gnu.org> 2011-05-23 16:22:18 UTC --- (In reply to comment #8) > Would a financial incentive help? My company may be interested in funding this > to speed up the work. I only work on libstdc++ in my spare time and I lack the time, not the motivation :) There are some others who expressed an interest in improving the locale code for Darwin and they might be interested. See the libstdc++ mailing list archives for December 2010, January, February and May 2011 for various threads on the subject. > > I think a better long term plan is to write a new locale model based on the > > POSIX 2008 API, as that could be common to more platforms and would provide > > thread-safety and the other advantages of the GNU/Linux code. > > Would this be a significant volume of work? I don't *think* so, the GNU locale model already uses those APIs as they've been in glibc for ages, so a good starting point would be copying the GNU code and fixing it up a bit to replace anything GNU-specific with the POSIX.1-2008 equivalent. Someone already started doing that for Darwin earlier this month.
