https://bugzilla.redhat.com/show_bug.cgi?id=1406946 --- Comment #6 from Adam Williamson <awilliam@xxxxxxxxxx> --- "Why are you using the GitHub commit to do this?" The guidelines used to require this, on the basis that tags are mutable; commit IDs are not. Specifying a commit ID guarantees we will *always* get the same tarball, even if upstream does something sneaky like ninja-ing the version tag. The guidelines now allow the use of tags, by the looks of it, but I still kinda prefer using the commit ID explicitly, I was persuaded by the old logic here. "Why the LANG? Could you put a comment explaining why that was necessary?" Sure, I'll add a comment. setup.py has a fairly common bug where it reads content in from a file without explicitly specifying the codec, and the file contains non-ASCII characters; in the Koji build environment, we don't default to a UTF-8 locale, we get C (I think), so Python 3 uses 'ascii' codec and blows up. I don't recall why this doesn't affect Python 2, but oh well. It's easier to just specify a LANG for the commands than patch setup.py to specify a codec... -- You are receiving this mail because: You are on the CC list for the bug. You are always notified about changes to this product and component _______________________________________________ package-review mailing list -- package-review@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe send an email to package-review-leave@xxxxxxxxxxxxxxxxxxxxxxx