
#11370: Redundant superclass warnings being included in -Wall destroys the "3 Release Policy" -------------------------------------+------------------------------------- Reporter: ekmett | Owner: Type: bug | Status: new Priority: highest | Milestone: 8.0.1 Component: Compiler (Type | Version: 7.10.3 checker) | Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: | Unknown/Multiple Type of failure: None/Unknown | Test Case: Blocked By: | Blocking: Related Tickets: #11369, #11429 | Differential Rev(s): Wiki Page: | -------------------------------------+------------------------------------- Comment (by ekmett):
Doesn't that mean that 8.2 will be in precisely the same situation as 8.0, and we'll just have the identical conversation then?
The main difference is that in 8.2 we'd have at least 2 releases with #11429 resolved, so if users used `-Wno-whatever` and the compiler didn't understand it, it'd not be a hard error. That makes increasing the length of the options list that the 3 release policy applies to a fair bit more palatable. By 8.4 users wouldn't even have to use cabal flags, so in many ways this is the same conversation, but it gets easier to resolve the further out we push.
Plus there is some protocol that a new warning always starts in -Weverything; may move in the next release to `-Wall`; and may move in the release after that to the default set. Both moves are a judgement call.
That sounds like a good rubric to me. All of the warnings we'd need for the items on the current roadmap (https://prime.haskell.org/wiki/Libraries/Proposals) seem to fit these guidelines. As for `-Wlint` it was just a suggestion as to what one of the extended warning sets might be, if it existed at all, it'd be included in `-Weverything`. Herbert has given a lot more thought to what groupings of flags make sense than I have, and I'm happy to defer to him there. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/11370#comment:11 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler