|
From: | Phil Holmes |
Subject: | Re: 2.13.40 regtests |
Date: | Wed, 24 Nov 2010 15:47:43 -0000 |
On 21 November 2010 23:16, Graham Percival <address@hidden> wrote:
Huh. So evidently there's some other reason behind the fix for this break not being detected? It might be related to the imagemagick's syntax change: http://code.google.com/p/lilypond/issues/detail?id=908
I don't think this is the problem.
The check relies on a grob existing in the baseline image for comparison with the check image; in this case, there's no BassFigure in the original image, and its appearance in the changed image has no effect on other bounding boxes, so there's no distance change.
Cheers, Neil
So if a completely new bit of graphics appears, the regtest checker wouldn't spot it? Not sure that's too good. What's the checker written in?
-- Phil Holmes Bug Squad
[Prev in Thread] | Current Thread | [Next in Thread] |