On Tue, 2016-11-15 at 12:06 -0500, Jeff King wrote:
>  - when an object write is optimized out because we already have the
>    object, git will update the mtime on the file (loose object or
>    packfile) to freshen it

FWIW, I am not seeing this happen when I do "git read-tree --prefix"
followed by "git write-tree" using the current master (3ab2281).  See
the attached test script.

> If you have long-running data (like, a temporary index file that might
> literally sit around for days or weeks) I think that is a potential
> problem. And the solution is probably to use refs in some way to point
> to your objects.

Agreed.  This is not my current scenario.

> If you're worried about a short-term operation where
> somebody happens to run git-gc concurrently, I agree it's a possible
> problem, but I suspect something you can ignore in practice.
> 
> For the most part, a lot of the client-side git tools assume that one
> operation is happening at a time in the repository. And I think that
> largely holds for a developer working on a single clone, and things just
> work in practice.
> 
> Auto-gc makes that a little sketchier, but historically does not seem to
> have really caused problems in practice.

OK.  I'll write a patch to add a summary of this information to the
git-gc man page.

Matt