Re: ds/multi-pack-index (was Re: What's cooking in git.git (Jul 2018, #02; Wed, 18))

From: Junio C Hamano <gitster@pobox.com>
Date: Jul 20, 2018, 16:09 UTC
Message-ID: <xmqqzhylc2jd.fsf@gitster-ct.c.googlers.com>
In-Reply-To: <c1a697ed-a060-1901-073f-7c8d5d5d0f10@gmail.com>

Derrick Stolee <stolee@gmail.com> writes:

Show 8 quoted lines

>>   What's the doneness of this one?  I vaguely recall that there was
>>   an objection against the concept as a whole (i.e. there is a way
>>   with less damage to gain the same object-abbrev performance); has
>>   it (and if anything else, they) been resolved in satisfactory
>>   fashion?
>
> I believe you're talking about Ævar's patch series [1] on
> unconditional abbreviation lengths.

Yes, this is a total tangent, but what happened to that one? I did not queue because I was led to expect v2 to follow soonish [*1*].

> Lookup speeds improve in a multi-pack environment.

True. I recall that years ago there was a discussion, but nobody came up with patches, to do the consolidated .idx for exactly that reason (not the "abbrev" reason).

> That's the best I can do to sell the feature as it stands now (plus
> the 'fsck' integration that would follow after this series is
> accepted).

Heh, 'fsck' intergration is not a 'feature' to sell anything, I would think. Nobody wants to run fsck for the sake of running it---it is just having one extra file that must not go corrupt _requires_ one to have a way to check its integrity and fsck is the logical place to do so X-<.

In any case, we've had this for about a week in 'pu' after 4 iterations, and review comments seem to have quieted down [*2*], so let's consider merging it down to 'next'. I think at least I need to "commit --amend" (or something like that) 16/23.

[Footnotes]

*1* <87a7s4471y.fsf@evledraar.gmail.com>

*2* That does not indicate either of these two:

    - nobody is interested in the topic
    - the topic is now without any flaw

    It only means that keeping it in 'pu' as a dormant topic would
    not do anybody any good.