Re: [RFC PATCH] multi-pack-index: allow operating without pack files

From: Johannes Berg <johannes@sipsolutions.net>
Date: Aug 23, 2021, 08:23 UTC
Message-ID: <3505a827a42096938746691acf4b1a2f5bf9d04f.camel@sipsolutions.net>
In-Reply-To: <YSMenndGYr14okwv@nand.local>

On Mon, 2021-08-23 at 00:05 -0400, Taylor Blau wrote:

Show 5 quoted lines

> 
> We're going to run into problems much earlier than that, though: the
> MIDX needs to load information about objects from packs in order to
> break ties when multiple copies of the same object exist in multiple
> packs (according to that pack's mtime).

Huh, I guess I never ran into a need - we make sure in bup that each object only exists once, that's kind of the point :)

I guess we could break ties by "lower hash of the pack" instead of "mtime"? It doesn't really matter how they're broken, as long as it's consistent?

Arguably, mtime is not a good measure anyway, since the same repo elsewhere would have different mtimes.

> So I'm not sure how we would even write a MIDX without physical packs on
> disk that we can open and stat, let along how we would teach Git to
> handle a situation where packs that did exist when writing a MIDX went
> away when we tried to read from the same MIDX later on.

It handles that just fine on the read side.

johannes