Hi all, colord creates an extra-profiles subpackage that splits out some extra, lesser user profiles. These are no-arch, but it appears that on different architectures the results are very, very slightly different. As we're dealing with 4Gb+ worth of floating point calculations, I'm sure it's something subtle to do with precision of floating point on different architectures. The failure is: FAILED: BuildError: mismatch when analyzing colord-extra-profiles-1.2.3-2.fc21.noarch.rpm, rpmdiff output was: added /usr/share/color/icc/colord/FOGRA27L_coated.icc ... The build is here, if anyone can work it out: http://koji.fedoraproject.org/koji/taskinfo?taskID=7572550 Thanks for any ideas -- either to ignore the rpmdiff output, or just if making the subpackage non-noarch is the right thing to do. Richard -- devel mailing list devel@xxxxxxxxxxxxxxxxxxxxxxx https://admin.fedoraproject.org/mailman/listinfo/devel Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
