https://bz.apache.org/SpamAssassin/show_bug.cgi?id=7735
--- Comment #35 from Henrik Krohns <apa...@hege.li> --- Sort is just hiding the bug. It should be enough to run a single do_meta_tests/finish_meta_tests in the end, which should iterate all metas over and over until all are ready. I just can't figure out why rule order would matter, even rewrote things now to work in "deterministic" way, firing metas from rule_ready when deps are ready.. and it actually behaves the same way. There's some logical fallacy here that I just can't grasp yet. -- You are receiving this mail because: You are the assignee for the bug.
