Gabor Greif
Gabor Greif
see also http://hackage.haskell.org/trac/ghc/ticket/8039
Yes, I am following this, but not sure whether Simon did the .so unloading, or "just" .o unloading... We are in need of the former.
Added [a new comment to the GHC linker ticket](http://ghc.haskell.org/trac/ghc/ticket/8039#comment:15). Maybe someone picks this up, I have no time a.t.m.
@schell you might get inspirations from my slides http://bobkonf.de/2015/slides/greif.pdf Anyway, extending patch to effectful operation is on my agenda (especially when deletions must be reified effectfully, which I haven't so...
@schell yea, sure, keep asking!
@schell regarding `IO` you only ever `diff` purely packed values, where `unsafePerformIO` is safe :-). Then `IO` is "just like" a polymorphic `Id` functor. `patch` is a different issue, altogether....
@schell I have written https://github.com/eelco/gdiff/wiki/PolymorphicContainers. For now it explains `[a]`, but `Maybe` should be the same, unless you want the optimization `Just' :: Fam a ts -> Fam (Maybe a)...
@schell I think what is tripping you over are the `List ListFam ts` constraints that live in `Ins` and co. I suggest writing a slightly more complex function that also...
@rvl What about setting up a small concrete example of what you want to accomplish (as a gist with MIT license) and give me perms? Then I can give it...
Cool. I doubt you'll get far with the "parse . show" approach. I'll try the method sketched in the presentation I linked in an earlier comment. But now I have...