[GHC] #11437: Don't put static (version-based) feature gates in compilerInfo
#11437: Don't put static (version-based) feature gates in compilerInfo -------------------------------------+------------------------------------- Reporter: ezyang | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.11 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- It started when someone added `("Support parallel --make", "YES"),`, and then ballooned to a bit more when I added a few more feature gates. I think it's probably wrong for GHC to put feature gates in this structure which are always on or off; Cabal can just version test GHC for appropriate information. (The one benefit of using a feature gate is when you're running a development version of GHC and Cabal, and there is a breakpoint without a version change when a feature is supported). If people agree, we should remove these, and change Cabal to version test to determine if these features are available. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/11437 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler
#11437: Don't put static (version-based) feature gates in compilerInfo -------------------------------------+------------------------------------- Reporter: ezyang | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.11 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: | Unknown/Multiple Type of failure: None/Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): Wiki Page: | -------------------------------------+------------------------------------- Comment (by thomie):
I think it's probably wrong for GHC to put feature gates in this structure which are always on or off
Why? Why do you want to remove them? And which fields are you talking about specifically? -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/11437#comment:1 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler
#11437: Don't put static (version-based) feature gates in compilerInfo -------------------------------------+------------------------------------- Reporter: ezyang | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.11 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: | Unknown/Multiple Type of failure: None/Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): Wiki Page: | -------------------------------------+------------------------------------- Comment (by ezyang): The feature gates which I want to remove are `Support parallel --make`, `Support reexported-modules`, `Support thinning and renaming package flags`, `Requires unified installed package IDs` and `Uses package keys`. I want to remove these because, if this is followed to its logical extreme, you'll end up putting lots of feature gates in `compilerInfo`, ballooning its size, slowing down Cabal (since it always has to parse this structure when it is configuring). What should you put in this structure? If we want to continue feature-testing, we should do it another way, like giving a list of supported flags, etc. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/11437#comment:2 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler
#11437: Don't put static (version-based) feature gates in compilerInfo -------------------------------------+------------------------------------- Reporter: ezyang | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.11 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: | Unknown/Multiple Type of failure: None/Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): Wiki Page: | -------------------------------------+------------------------------------- Comment (by thomie): Ah, for some reason I was thinking about the `settings` file. Sure, if everything keeps working, remove 'em. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/11437#comment:3 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler
participants (1)
-
GHC