On Fri, Jul 13, 2012 at 3:54 PM, John Wiegley <johnw@newartisans.com> wrote:

>>>>> Jan Djärv <jan.h.d@swipnet.se> writes:

> This might be true in some sense, but it is not practical.  GNUStep does not
> get much attention in Emacs, but it is more likely that a developer that has
> looked at the OSX code takes a stab at it if they are similar and use the
> same API:s.  If we bring in the Mac port, I think we must drop GNUStep due
> to lack of developer time.

Isn't the lack of developer time being spent on GNUstep a fact that there
aren't many developers interested in maintaining it?

By not splitting these two, you are losing out on the consistent efforts of
Yamamoto Mitsuharu, who has done a superlative job at providing an excellent
experience for Mac users.  If we use his code, we also gain him as an active
developer for a very active platform.  By sticking with GNUstep, however much
the FSF may want that, we are restricting ourselves to a developer pool
interested in GNUstep -- which is not going to include many people from the
Mac development camp.

If there were another alternative to the ns tree that would produce an OS X Emacs binary, I would build it in a heartbeat and then choose between the two which one I will put time into supporting.  Seems crazy to me not to give Yamamoto's tree a chance to grow.  Is there a git-able source tree somewhere or is it strictly a patch on top of the ns tree?