Re: vc-state-heuristic is gone

[Karl Fogel <kfogel@red-bean.com>]

Richard Stallman <rms@gnu.org> writes:
>What is the caching you plan to eliminate?
>
>Is it caching of data in Emacs to represent the state of local disk
>files?
>
>Is it caching of data in local disk files to represent the state of
>a remote repository?
>
>The former, I have no opinion about.  The latter would be a disaster
>for me.  The operations that now work locally in a few seconds will
>take HOURS (until I have a network connection).
>
>Have you tested the speed of the operations when your connection to
>Savannah is going through Tor?

Without even looking at the VC code, I'm pretty confident that VC
doesn't cache anything significant about remote repository state beyond
what the local working copy already caches (for systems that do not
simply clone the entire repository locally, that is).  After all, it
would be very surprising for Emacs VC to find a need to cache something
that the underlying version control system doesn't already cache for its
own purposes... though I suppose stranger things have happened.

I could be wrong, and if I am then it'll be interesting to be corrected.
If I'm right, at least I've saved ESR the time of typing the above :-).

Richard, is there a particular operation you're worried about?

-K