> If the package *requires* IPv6 to build, it should not simply be deleted;
> rather, some mechanism should enforce the IPv6 capability, and fail if the

but that means we need a different indicator to denote whether a package
requires IPv6 support (e.g. IPv6 compatible header files) to compile.
USE_INET6 definitely wasn't meant to do that (although it seems to have
been abused for that purpose in a few instances).

-- 
Georg Schwarz    http://home.pages.de/~schwarz/
 georg.schwarz@freenet.de  +49 178 8545053