Re: [GHC] #15887: Ideas for a better GHCi editing experience

11 Nov 2018

      #15887: Ideas for a better GHCi editing experience
-------------------------------------+-------------------------------------
        Reporter:  Anchpop           |                Owner:  (none)
            Type:  feature request   |               Status:  closed
        Priority:  high              |            Milestone:  Research
                                     |  needed
       Component:  Compiler          |              Version:  8.6.2
      Resolution:  invalid           |             Keywords:  QoL
Operating System:  Unknown/Multiple  |         Architecture:
                                     |  Unknown/Multiple
 Type of failure:  None/Unknown      |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:                    |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------
Changes (by goldfire):

 * status:  new => closed
 * resolution:   => invalid

Comment:

 Thanks, Anchpop, for a detailed description of how you see this feature
 working. While one can always debate minutiae, I agree that a design such
 as you suggest would be nice to use and would aid beginners.

 However, I agree with comment:2 that these features are better suited to
 tools outside of GHCi. Furthermore, implementing these would be quite an
 engineering challenge, as it fundamentally changes the way that GHCi
 interacts with the user. In the end, the GHC API is expressive enough that
 you (or anyone) could write an alternative to GHCi, with all the features
 GHCi has, but with this (or other) interface.

 If you're keen to pursue this, the best avenue is to make an official
 proposal at https://github.com/ghc-proposals/ghc-proposals (indeed, a
 little tweaking of your original message could form such a proposal), as
 that's the place where any changes to the way GHC works are considered. My
 guess is that you'll receive a negative reaction there (not to the
 interaction design itself, but to the fact that it would be in GHCi), but
 I could well be wrong.

 In the meantime, I will close this ticket, as this really would need to go
 through the proposals process before we take action on it. If a proposal
 is accepted, perhaps this ticket can be reopened.

-- 
Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/15887#comment:4
GHC http://www.haskell.org/ghc/
The Glasgow Haskell Compiler