Re: [PATCH 2/2] Implement a simple delta_base cache

On Sat, 17 Mar 2007, Nicolas Pitre wrote:

Sure.  But at this point the reference to compare GIT performance 
against might be GIT itself.  And while 1 second is really nice in this 
case, there are some repos where it could be (and has already been 
reported to be) much more.

I'd still like to see the KDE repo, that thing went quiet after it was 
supposed to hit sneaker-net..

If it was 30 seconds before to do a "git log" for some individual file, 
after the recent optimizations it should hopefully be down to 10. And I 
agree that I might be more motivated to try to get it down further if I 
could just find a repository where it's that much. 

Right now I can can do a "git log" on any file in the kernel archive in 
under a second (well, when I say "any file", I started with a script, but 
with 22 thousand files I didn't bother to run it for all that long, so I 
ended up testing a few random files in addition to the first few hundred 
files of "git ls-files", and they are all well under a second).

And that's without the "git diff --quiet" thing that is still in "next", 
and that cut down some of the overhead for other reasons (although I 
suspect the effect of that will be less when combined with my patches 
since the stuff it cut down I probably cut down even more).

I really suspect you'll have a hard time beating "normal" git with the 
patches I sent out. I'm sure it's quite possible - don't get me wrong - I 
just suspect it won't be spectacular, and it will be a lot of work.

		Linus