Post any defects you find in the released or beta versions of the ImageMagick software here. Include the ImageMagick version, OS, and any command-line required to reproduce the problem. Got a patch for a bug? Post it here.
I think it's critical error. The update was stopped with error message.
It's successful updated on FreeBSD 7.0-RELEASE-p2 amd64 (one of my another servers)
This is perplexing since all the regression tests pass on the hosts we have access to, Fedora, Centos, Mac OS X, Windows, and Solaris. Its curious that the test that fails uses doubles. There can be some fuzziness because of precision round-off but CMY is a one-to-one mapping to and from RGB so precision should not be an issue.
We used to test ImageMagick under FreeBSD on the HP TestDrive systems but they have discontinued this program. Can you provide a system we can use to investigate this problem report?
magick wrote:
We used to test ImageMagick under FreeBSD on the HP TestDrive systems but they have discontinued this program. Can you provide a system we can use to investigate this problem report?
Unfortunately I cannot. The system is leased to customer.
I'll address to FreeBSD maintainer of ImageMagick.
Probably I've encountered with known OS-dependent problem.
magick wrote:We found a FreeBSD 7.0 system and built the latest ImageMagick release 6.4.7-7. All the regression tests passed as expected.
So am I. I've reported (see my previous post).
Seems, the problem appears on FreeBSD 6.3 only.
Now, I've encountered with same problem on another server on FreeBSD 6.3-RELEASE-p5 i386, installed two week ago. Error message is:
Constitute check failed: 6615/0.0587497/0.843137
FAIL: tests/constitute_double_cmy.sh
Perhaps there is a bug in the compiler under FreeBSD 6.3. Hopefully we will be able to get access to a FreeBSD 6.3 system in the near future to resolve this problem. Thanks.
Thanks to knigma we have a fix to this problem. The problem was within the regression test itself (i.e. tests/constitute.c) rather than in the ImageMagick API.