I can not approve this release candidate.
I took care to remove all vestiges of libtiff 4.7.2 and re-installed
libtiff 4.7.1. I built and linked GraphicsMagick against libtiff 4.7.1.
I ran the 'genimages' script and saw all of the checks pass. Then I
installed libtiff 4.7.2 and ran the 'genimages' script again. 680 out
of 5568 test cases failed. A strange thing is that often the only
difference between a test case which passes or fails is the requested
compression algorithm. The corrupt image is typically posterized so the
defect is very obvious and is not just a few bad pixels. It is possible
that libtiff is corrupting memory used by GraphicsMagick.
Bob
On 6/23/26 06:31, Even Rouault via Tiff wrote:
Hi,
I've prepared a second release candidate for libtiff v4.7.2:
- https://download.osgeo.org/libtiff/tiff-4.7.2rc2.tar.gz
- https://download.osgeo.org/libtiff/tiff-4.7.2rc2.tar.xz
- https://download.osgeo.org/libtiff/tiff-4.7.2rc2.zip
Differences since rc1: fixing issue reported yesterday by Christoph
Gohlke, and a few build warnings on MacOSX
Motion: approve libtiff 4.7.2rc2 as final 4.7.2
Starting with my +1
Even
_______________________________________________
Tiff mailing list
[email protected]
https://lists.osgeo.org/mailman/listinfo/tiff
