Source: fenics-basix Followup-For: Bug #1090766 X-Debbugs-Cc: [email protected] User: [email protected] Usertags: armel
I'm pragmatic about adjusting tests for different architectures. I'd prefer not to do, but sometimes tests are so comprehensive and there's no practical way of figuring out what went wrong on minor architectures, so better to reliably pass a subset of tests than to not pass at all, which means the architecture misses out on the package altogether. If the actual problem on armel can be fixed, then great. But in the meantime I'm happy to apply to patch. In this case it's academic since we've just dropped armel anyway. But doesn't hurt to apply the patch.
