Re: What's cooking in git.git (Nov 2025, #03; Fri, 7)

[Junio C Hamano <gitster@pobox.com>]

Toon Claes <toon@iotcl.com> writes:

> Junio C Hamano <gitster@pobox.com> writes:
>
>> * tc/last-modified-active-paths-optimization (2025-10-23) 1 commit
>>   (merged to 'next' on 2025-11-03 at 9ab444edfb)
>>  + last-modified: implement faster algorithm
>>
>>  "git last-modified" was optimized by narrowing the set of paths to
>>  follow as it dug deeper in the history.
>>
>>  Will cook in 'next'.
>>  source: <20251023-b4-toon-last-modified-faster-v3-1-40a4ddbbadec@iotcl.com>
>
> Do you plan to include these changes in the upcoming v2.52 release? I
> would like to argue to include them.

I plan to exclude, especially if it was not merged to 'next' before
2025-10-30, which is one calendar week before -rc1, to give the
topic the usual cooking period.

> Maybe it wasn't very clear how I phrased it in the v4 cover letter[1],
> but without this patch, some results are wrong:

Indeed it was unclear.  With respect to the output, the only thing I
recalled was (I just went back to the thread)

    One thing to note though, the results might be outputted in a different
    order than before. This is not considerd to be an issue because nowhere
    is documented the order is guaranteed.

and the patch came with a single test that gives a much later than
before in check_last_modified output, which does not clearly
demonstrate how the old results were wrong and the new results
correct them.

>> As an added benefit, results are more consistent now. For example
>> implementation in 'master' gives:
>> 
>>     $ git log --max-count=1 --format=%H -- pkt-line.h
>>     15df15fe07ef66b51302bb77e393f3c5502629de
>> 
>>     $ git last-modified -- pkt-line.h
>>     15df15fe07ef66b51302bb77e393f3c5502629de	pkt-line.h
>> 
>>     $ git last-modified | grep pkt-line.h
>>     5b49c1af03e600c286f63d9d9c9fb01403230b9f	pkt-line.h
>> 
>> With the changes in this patch the results of git-last-modified(1)
>> always match those of `git log --max-count=1`.
>
> So this patch speeds up git-last-modified(1), but it also fixes some
> buggy results. That's why I consider this a bug fix and worthy to make
> it into the upcoming release. (which is by the way also the first
> release that includes this subcommand)
>
> [1]: https://lore.kernel.org/git/20251103154726.26592-1-toon@iotcl.com/