Re: [PATCH 2/2] log: do not shorten decoration names too early

[Jeff King <peff@peff.net>]

On Fri, May 22, 2015 at 02:21:16PM -0700, Junio C Hamano wrote:

> I ended up doing it as a variant of the latter, "free unless we have
> v->buffer pointing at it".

Thanks, this version looks good to me minus one micro-nit below.

> Sorry for a long delay.

No problem. I'm sometimes amazed you find time to write any patches at
all. :)

> -- >8 --
> Subject: [PATCH] commit-slab: introduce slabname##_peek() function
> 
> There is no API to ask "Does this commit have associated data in
> slab?".  If an application wants to (1) parse just a few commits at
> the beginning of a process, (2) store data for only these commits,
> and then (3) start processing many commits, taking into account the
> data stored (for a few of them) in the slab, the application would
> use slabname##_at() to allocate a space to store data in (2), but
> there is no API other than slabname##_at() to use in step (3).  This
> allocates and wasts new space for these commits the caller is only
> interested in checking if they have data stored in step (2).

s/wasts/wastes/

-Peff