Paul F. wrote: > lyndon wrote: > > The "nulls file" test is failing, due to the NUL characters not > > being base64 encoded. This doesn't show up in the buildbot > > regression step output as the raw NULs make diff choke. I haven't > > had time to investigate further. > > i can take a look at this (finally). when i first saw this mail, i > assumed the problem was an interaction with the test framework due > to tool differences (i.e., diff) on netbsd. > > but now re-reading, i'm not sure i understand. is diff's choking a > secondary effect, or is it the primary failure?
My guess is that the problem is with the behavior of mhbuild, not diff. diff is reporting a difference, which is all it does for binary files. At this point, we don't know whether the difference is benign. Removing the attachment of tiny.jpg from the nulls test might help isolate things, though I'm not optimistic. Attachment of tiny.jpg was already tested, so it doesn't add much of value here. David _______________________________________________ Nmh-workers mailing list [email protected] https://lists.nongnu.org/mailman/listinfo/nmh-workers
