Re: MIDX woes, was Re: [ANNOUNCE] Git v2.54.0-rc2

[Taylor Blau <me@ttaylorr.com>]

On Thu, Apr 16, 2026 at 09:09:10AM -0700, Junio C Hamano wrote:
> > The part that is striking about Johannes' report is that this happens
> > on a fetch, so it is being written automatically by normal maintenance
> > activities in that case. Without downgrading the default version back
> > to a compatible version, deleting the multi-pack-index(es) will not
> > keep the repo in a good state.
>
> I am not sure what you mean by "deleting will not keep the
> repository in a good state".  Isn't multi-pack-index totall
> optional?

I am also having trouble following this.

Is this saying that fetching from a remote and having a local `git
maintenance` (run automatically after `git fetch`) write a v2 MIDX file
which is unable to be read by a separate local Git is causing the
problem?

If so, then I think we are operating in a mixed-version environment
where some newer Git that understands the v2 format is writing it, and
another older Git version that doesn't is trying to read an unknown
version.

Though I find it totally plausible that I am wildly misunderstanding
what happened for Johannes here, so please let me know if I'm not seeing
it.

Thanks,
Taylor