[GHC] #10643: GHC cannot import submodules when run from subfolder

[GHC]

#10643: GHC cannot import submodules when run from subfolder
-------------------------------------+-------------------------------------
        Reporter:  FPtje             |                   Owner:
            Type:  bug               |                  Status:  new
        Priority:  normal            |               Milestone:
       Component:  Compiler          |                 Version:  7.10.1
      Resolution:                    |                Keywords:  subfolder
Operating System:  Unknown/Multiple  |  import submodule cd
 Type of failure:  GHC rejects       |            Architecture:
  valid program                      |  Unknown/Multiple
      Blocked By:                    |               Test Case:
 Related Tickets:                    |                Blocking:
                                     |  Differential Revisions:
-------------------------------------+-------------------------------------

Comment (by FPtje):

 > Simply trying to run "ghc -Wall" (plus a few warning/error-related
 flags) is just way to naive to be helpful for any non-toy program
 Nonsense! ''ghc -Wall'' works fine when this issue is fixed/worked around!
 http://i.imgur.com/CxHO2S5.png

 Besides, even if there are ''other'' hurdles that would make ghc -Wall not
 work for single files in bigger projects, that does not mean that this
 issue should not be resolved.

 > [...] other language environments/IDEs

 > [...] haskell-mode for Emacs, it does exactly what you want in a saner
 way

 At the risk of sounding even grumpier, the core of this issue lies with
 ghc not finding modules that it arguably should. Therefore, the
 responsibility lies with ghc to solve it. You can tell me to use emacs and
 you can tell me that IDE's for other languages need more information
 before they can lint code, but that won't make the problem go away for
 ghc.

 The proposed fix isn't that difficult, makes proper sense given how the
 folder structure vs. module name works and should (AFAIK) not break any
 backwards compatibility. I don't see why you would want to throw this away
 at face value before expecting everyone who uses runhaskell, ghci and ghc
 to just work around it while stating "that's how IDEs/haskell linters do
 it".

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10643#comment:8>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler