From: Jeff King <peff@peff.net>
Date: Wed, 09 May 2012 17:26:47 GMT
Subject: Re: [PATCH] sha1_name: support sha1^{note} to return note sha-1
Message-ID: <20120509172647.GA30487@sigill.intra.peff.net>
In-Reply-To: <CALKQrgdGCoKooM7KayNMQ_+Fg9=-1YnwNE_9NHzpX0jfoRjT_A@mail.gmail.com>

On Wed, May 09, 2012 at 04:09:35PM +0200, Johan Herland wrote:

> > No idea. I have never worked/used notes until yesterday (and these
> > patches were the result). I will look into it.
> 
> IIRC, the initialization loads the top-level notes tree object into
> memory. Subtrees (if any) are loaded on demand. FTR, if you have less
> then ~256 notes in the notes tree, there will be no subtrees. As the
> number of notes grows, the number of subtree levels grow roughly
> logarithmically with the total number of notes (see determine_fanout()
> for more details).

Ah, right. I was thinking back to my original crappy implementation that
didn't do fanout. So I don't think an init is that bad. It does look up
the ref each time, but it will only load the top-level tree object
(which you would need to do a lookup anyway).

So it is probably OK to just do an init/lookup/free each time if it
makes the code simpler (and I think it does). We probably won't be
looking up sha1^{note} in a tight loop, anyway, since sha1 expressions
like that generally come from the command line.

-Peff