> On Mon, Jun 23, 2003 at 06:53:28AM -0700, Jeff Garland wrote: > > [snip] > > Ok will do. I'll add some protections or rounding for when durations > > support higher than microsecond resolution. I'll let you know when > > this gets added. > > It seems that since 1) timeval is non-(C++) standard, and 2) it is > used only by select, that this conversion is more appropriate in a > Boost socket library, and even there only as an internal component.
I think that it is broader than sockets. The thread library, for example, uses something that basically amounts to timeval in some cases. Finally, if someone wanted to use a different socket library (especially since boost.socket isn't in boost) why make them rewrite this code? Jeff _______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
