From: Linus Torvalds <torvalds@osdl.org>
Date: Wed, 09 Nov 2005 00:51:11 GMT
Subject: Re: Comments on recursive merge..
Message-ID: <Pine.LNX.4.64.0511081646160.3247@g5.osdl.org>
In-Reply-To: <20051109003236.GA30496@pasky.or.cz>



On Wed, 9 Nov 2005, Petr Baudis wrote:
> 
> BTW, git-show-branch is also by orders of magnitude faster (not that
> this would be any major timesaver). Median 0.006s vs. median 0.124s.

Ouch. That makes me suspicious. One reason git-merge-base is slow is 
because it's being pretty careful about some pathological examples of 
dates being just the wrong way around, and it might just be that the 
reason git-show-branch is faster is because it isn't doing that part 
right.

So yes, git-merge-base does extra work, but it does so because I think it 
needs to.

Junio? You even wrote the comment about the case in git-merge-base, I'm 
wondering whether it's a bug that we use the fast-and-cheap algorithm in 
git-show-branch..

Of course, arguably you can first try the fast-and-cheap thing, and if 
that gives a merge parent that is acceptable, why not? So maybe it's the 
right thing for the "let's see if this is trivial" case, but I think it 
might _think_ some cases are trivial that really shouldn't, because they 
actually have two merge parents.

		Linus