Thomas Dickey wrote: > On Sun, 4 Oct 2009, Julien Cristau wrote: > >> On Sun, Oct 4, 2009 at 08:50:24 -0400, Thomas Dickey wrote: >> >>> On Sun, 4 Oct 2009, Julien Cristau wrote: >>> >>>> That changes the behaviour, it forces libmd if you don't explicitly tell >>>> configure to use something else. With my patch, the first >>>> implementation found is used (and I think that's better). >>> A frequent cause of problems in *BSD ports is just picking up the "first >>> implementation found", since different users would have different ports >>> installed. >>> >> Your point being? > > other than user surprise? > > Some people might split that into separate options, particularly if the > licensing differs. > > Of course, if there are no technical or other differences, then there's > no reason to not lump things together.
As long as there is a possibility for force an option in such cases, I always explicitely use this option, to avoid surprises. BTW, OpenBSD has sha1 in libc and it has a local patch to configure.ac to check this and use it. I'll look at merging it at some point... -- Matthieu Herrb _______________________________________________ xorg-devel mailing list [email protected] http://lists.x.org/mailman/listinfo/xorg-devel
