I'd like to ask the following: does there exist a draft of a Haskell-2
Manifesto, that lists the design goals of this new language?
I've seen many wish list items that are "bottom-up": they advocate
and/or denounce some specific set of language features. This was
appropriate for the resolution of Standard Haskell. Yet I do not
believe that this is a good starting place for Haskell-2.
Consider how each of these questions should be answered two, three,
and five years in the future:
1) Who uses applications written in Haskell-2?
2) Who is programming in Haskell-2?
3) What kind of problems do these people solve with Haskell-2?
4) What kind of problems are _not_ solved with Haskell-2?
5) How is Haskell-2 {similar to,different than} other languages that
have evolved over the years?
There are other, more specific, questions that I haven't heard
answered either (but maybe asked, if indrectly):
5) What software design and implementation models should Haskell-2
support? How well?
6) How do other trends in PL design affect Haskell-2's design
goals? (e.g. AOP? Lessons from the OO front-lines?)
7a) What is the learning curve to write a correct Haskell program?
7b) How does this compare to learning to write an efficient Haskell program?
^^^^^^^^^
8) What good idioms (i.e. day-to-day abstractions) should Haskell
support more strongly via {syntactic sugar, libraries, language
modifications, language extensions}? (After all, where would we be
now if we'd stopped at FORTRAN IV and no for loops? 8-)
This list is hardly complete, and I don't expect to see answers
quickly or quietly.. ;-) But I believe they must be in order for
Haskell-2 to have substantial positive impact on the state of
computing.
-- John Whitley
