[GHC] #12618: Add saturated constructor applications to Core

[GHC]

#12618: Add saturated constructor applications to Core
-------------------------------------+-------------------------------------
        Reporter:  simonpj           |                Owner:
            Type:  feature request   |               Status:  new
        Priority:  normal            |            Milestone:
       Component:  Compiler          |              Version:  8.0.1
      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 carter):

 I'd like to chime in and agree that it would be a good design to reflect
 saturated function application and related info into a corresponding multi
 arg saturated app application form. I've some woke in progress type theory
 developments critically use having a built in motion of simultaneous
 arguments and results (a la types are calling conventions or sequent
 core), where the logical strength of what can be described needs a built
 in motion of simultaneous arguments.


 I presume / assume this multi arg app form is essentially a function
 application against an unboxed tuple? Right?


 One thing I'd like to point out is that a knock on effect of this change
 you may want to consider is having Unboxed tuples in arg and return
 positions act like pi and sigma telescopes respectively.

 If what I'm sketching out needs more clarity, I'm happy to do a clearer
 exposition on the wiki or the like

 If the meat of the ideas discussed / articulated are consistent with /
 facilitated by the details I'm hopefully articulating clearly, this would
 be a set of changes I would strongly support. In. Fact I could probably
 get support for putting work time into helping out on this change if need
 be. Partly because then certain ideas I would like to experimentally add
 to ghc would then be much much easier to add :)
 (The simultaneous arguments stuff makes embedding / supporting linear
 logical stuff in a clean way much nicer than previous efforts p)

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12618#comment:12>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler