https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=239682
--- Comment #33 from Brooks Davis <bro...@freebsd.org> --- (In reply to Jan Beich from comment #31) > > You are correct that some bugs won't be found until LLVM_DEFAULT is bumped, > > but > > doing it without coordination with me (the PR does not count) and making the > > switch the I was unavailable to respond to the reports is unacceptable. > > "(the PR does not count)" bit is offensive to me. In ports/ the primary way > to cooperate with each other is either via bugzilla. Other ways are too > easily lost in the noise. For one, portmgr@ encourages every ports/ > contributor to file a bug even for stuff submitted on phabricator. A PR is useful and necessary for a something like this, but a group PR with noisy history is no substitute for making individual contact with the most effected parties. If you only want to communicate via PRs then I can make a habit of setting maintainer-feedback to "-" immediately, but that doesn't feel like a constructive way to proceed. One additional point on LLVM updates, many downstream projects don't do their updates to new versions until after release. This isn't great, but it's reality. In the past I've found we had better luck waiting a few weeks or even a month so more things had been updated to support modern LLVM versions to reduce the amount of version smear in a typical pkg set. Making the LLVM_DEFAULT bump lift as many ports as possible at once is helpful in that it reduces the number of stragglers that need to be chased down later as we try to keep the number of versions under control. -- You are receiving this mail because: You are on the CC list for the bug. _______________________________________________ freebsd-gecko@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-gecko To unsubscribe, send any mail to "freebsd-gecko-unsubscr...@freebsd.org"
