On 17 June 2011 16:47, Simon Peyton-Jones
So: Under Plan A, some Hackage packages will become un-compilable, and will require source code changes to fix them. I do not have any idea how many Hackage packages would fail in this way.
Of the 372 direct reverse dependencies of haskell98: http://bifunctor.homelinux.net/~roel/cgi-bin/hackage-scripts/revdeps/haskell... there are 344 which also depend on base (See http://hpaste.org/47933 for calculating the intersection). Am I correct that a package which depends on both base and haskell98 will always fail to build when the Prelude is also exported from haskell98? (Unless of course the package uses the PackageImports extension) I don't know how many of these 344 packages use PackageImports or have upper bounds on their haskell98 dependency (I guess not many). So I guess many of these 344 will break. Still I'm in favour of plan A since it's simple and discourages mixing haskell98 with more modern modules. Regards, Bas