Dmitry Gutov <dgutov@yandex.ru> writes:
The code should be straightforward, but I'd like to see some
performance measurements: both for the local case, and for the remote
one (Tramp).
The difference can be small, though, given that we already try a
number of other backends first (Git is near the end of
vc-handled-backends; we might want to change that, BTW).
As you point out, Git is already towards the end, so the typical case
would just be two more tests, and both of the ones after Git (Hg and
Mtn) look pretty trivial, too.
So it looks rather unlikely that there'd be a noticeable performance
impact, I think?