and yes, teaching is important, and Haskell is not only about teaching, so teaching needs need to be addressed like the needs of any other important Haskell application area: by optional, but widely supported, domain-specific libraries/packages/flags/syntax/..
oh, and it might be useful to look at the interests of Haskell textbook authors in a different way: if I was an author of a Haskell textbook, and I had the foresight to rely not on "the" prelude or "the" standard libs (because they will both change, among other things, before the first edition of that book gets out of circulation..), what help could Haskell' or Haskell'' could give me? here are a few examples to get the discussion going - imagine things like: -package Thompson fromInt = fromInteger -- easy error message += explanation -- hmm? -package Hudak module SOE -- ouch, years of hard work by several hackers, -- and still mostly trouble and pain libaries += portable graphics and stuff -package Bird syntax pattern += (n+k) -- hmm -package Helium -- is there a book to go with this? language -= type classes -- not really easy? (error messages -= abstract messages) += specific messages -package NewKidOnTheBlock everything = haskell' and state-of-2006 -- to be adapted as things change as we can see from this exhaustive survey.. okay, I'm just kidding!-) but it seems that there are quite a few things that *are* important to Haskell teaching, and have disturbed Haskell textbook authors, that ought to be investigated as language and library design and maintainance problems - though it is a bit late in the day for Haskell', perhaps the committee and current authors could think of a roadmap, so that authors and implementers have something to aim for? and of course, so that these issues are adressed properly, rather than returning here as limitations on the design process? of course, that's no bikeshed, but still, perhaps people have an opinion?-) claus