Re: That improved git-gui blame viewer..

From: Johannes Schindelin <johannes.schindelin@gmx.de>
Date: Jun 13, 2007, 14:58 UTC
Message-ID: <Pine.LNX.4.64.0706131553390.4059@racer.site>
In-Reply-To: <e5bfff550706130708k601514c9o1ba2747582027699@mail.gmail.com>

Hi,

On Wed, 13 Jun 2007, Marco Costalba wrote:

Show 16 quoted lines

> On 6/13/07, Johannes Schindelin <Johannes.Schindelin@gmx.de> wrote:
> > >
> > > then running: git log --pretty=raw -z --parents --boundary HEAD > tmp.txt
> > 
> > Isn't that a bit unfair? You should have passed "--pretty=raw" to
> > rev-list, too.
> > 
> > Since you already have that tree, could you try rev-list _with_
> > "--pretty=raw", just for comparison?
> > 
> 
> Of course, here we are
> 
> running: git rev-list --pretty=raw --parents --boundary HEAD > tmp.txt
> 
> times are: 1,16s - 1,11s - 1,13s

I just tested on cygwin. The funny thing is, I never get anything like 8 seconds (this is on git.git itself).

For me it is

~1.0s rev-list _without_ --pretty=raw ~1.6s rev-list _with_ --pretty=raw ~1.4s log _with_ --pretty=raw ~3.5s log _with_ --pretty=raw _and_ -z (!)

So, your delay could stem from -z doing weird things.

For fun, I also did it with the MinGW port:

~0.7s rev-list _without_ --pretty=raw ~1.0s rev-list _with_ --pretty=raw ~0.8s log _with_ --pretty=raw ~1.9s log _with_ --pretty=raw _and_ -z (!)

Draw your conclusions, gentlemen.

Ciao, Dscho