Re: [PATCH] Bisect: fix calculation of the number of suspicious revisions

From: Uwe Kleine-König <ukleinek@informatik.uni-freiburg.de>
Date: Mar 21, 2007, 21:39 UTC
Message-ID: <20070321213956.GC17632@informatik.uni-freiburg.de>
In-Reply-To: <7vodmmz3wy.fsf@assigned-by-dhcp.cox.net>

Hello Junio,

Junio C Hamano wrote:

Show 12 quoted lines

> Uwe Kleine-König  <ukleinek@informatik.uni-freiburg.de> writes:
> 
> > Up to now the number printed was calculated assuming that the
> > current revision to test is bad.  Given that it's not possible
> > that this always matches the number of suspicious revs if the
> > current one is good, the maximum of both is taken now.
> >
> > Moreover I think the number printed was always one to high,
> > this is fixed, too.
> 
> I know you mean well, but is it really worth an extra rev-list
> for this off-by-one, I wonder?

It may be more than one.  E.g. 
   b
  / \
 /   \
a--c--e
 \   /
  \ /
   d

Given a is bad, e is bad, b is rev to test.

I didn't verify it, but I think the current code gives one (as it assumes b is bad and counts one to much). If b is good there are two suspicious left. I'm sure you can construct an example where it differs still more.

Best regards Uwe

-- 
Uwe Kleine-König

http://www.google.com/search?q=half+a+cup+in+teaspoons