Re: [PATCH] Add a 'generation' number to commits

From: Junio C Hamano <gitster@pobox.com>
Date: Jul 15, 2011, 19:49 UTC
Message-ID: <7v8vrz1g02.fsf@alter.siamese.dyndns.org>
In-Reply-To: <alpine.LFD.2.02.1107141126300.4159@i5.linux-foundation.org>

> Comments? This is pretty simplistic, and yes, it's slow. On the kernel, it 
> now takes a few seconds to generate a new commit when there are no 
> generation numbers - and that's on a fast machine.

I agree this is the way to go if we _were_ to use generation number associated with commit objects in the longer term, and if the SLOP logic in still_interesting() in revision.c:

 (1) can gracefully fall back to the date based heuristics for older
     commits without the header; and

 (2) can take advantage of the generation numbers in more recent commit.

If we cannot do (1), we could augment this with Peff's generation number cache. I suspect (1) is doable and in that case we do not have to have (and we may be better off without) the on-disk cache that could go stale, but nobody so far has shown that yet, so...

As I mentioned in a review comment of the actual patch, I however am not convinced that generation number is a better substitute for the timestamp in the context of "tag --contains" optimization.