Re: [PATCH 13/16] prune: keep objects reachable from recent objects

From: Junio C Hamano <gitster@pobox.com>
Date: Oct 4, 2014, 03:04 UTC
Message-ID: <xmqqfvf4k02e.fsf@gitster.dls.corp.google.com>
In-Reply-To: <20141004003036.GE17063@peff.net>

Jeff King <peff@peff.net> writes:

Show 19 quoted lines

> On Fri, Oct 03, 2014 at 02:47:57PM -0700, Junio C Hamano wrote:
>
>> With this patch applied, the system will not prune unreachable old
>> objects that are reachable from a recent object (the recent object
>> itself may or may not be reachable but that does not make any
>> difference).  And that is sufficient to ensure the integrity of the
>> repository even if you allow new objects to be created reusing any
>> of these unreachable objects that are left behind by prune, because
>> the reachability check done during prune (with this patch applied)
>> makes sure any object left in the repository can safely be used as a
>> starting point of connectivity traversal.
>
> Your use of "safely" in the last sentence here made me think.
>
> In a repository that has had this patch from the beginning, it should be
> safe to traverse the unreachable but unpruned objects, because the
> property of the repository we are trying to guarantee means that we will
> have all of the referents.
> ...

Another case that may be of interest is to start a dumb HTTP commit walker to fetch new history and kill it before it manages to complete the graph and update the refs. It has the same property as running hash-objects to create an unconnected cruft object after you got your repository into a "safe" state, but it would be a lot less easier to blame the user for the resulting "breakage".

Perhaps the dumb commit walkers outlived their usefulness and we should start talking about their deprecation and eventual removal at a later version of Git?