But now that the situation is what it is I stand by what I say
above: it
doesn't interfere with anything in core emacs (as it is / should be
additive with line-spacing), users benefit from it, and others can
use it
to try and see whether it's something that would be worth having
on X and
W32. There's no benefit to removing it.
This would encourage the use of MacOSX to try the feature. As long as
MacOSX is not Free Software, this goes against our goals. Also I
think
it's pretty clear that the `line-spacing' feature can be expanded to
cover the featureset of `ns-expand-space' and that it would be good
since it would remove a redundant setting, and bring that feature to
all architectures.
So we should remove ns-expand-space right now. If we can come up with
a clean enough patch to extend line-spacing's semantics, we might be
able to consider it for inclusion in Emacs-23.1, otherwise it'll
have to
wait for Emacs-23.2.