My freebsd-current flamewar is dying out, so time for something new ...
Why are parallel mkfiles in the /go tree considered eeevil? They seemed to be very low overhead. Surely accomodations are being made for Windows.
Regardless, if we want native go, is it that unreasonable to maintain a port ala spin? My fiddling with the go tree makes me think it isn't that hard to set up an overlay that adopts the live go tree to what Plan9 needs for a build. bind(1) is a wonderful thing -- we should use it!
The most dynamic part of the go source tree these days seems to be pkg/Makefile (as far as a native build is concerned). It can't be that hard to parse out DIR= from there and turn it into a native mkfile.
