[Haskell-cafe] Seeking feedback for a tutorial paper draft about GADTs

[Alexander Batischev]

On Tue, Jan 08, 2013 at 11:22:39PM +0400, Anton Dergunov wrote:
> I have written a draft of an introductory-level tutorial paper about
> GADTs in Haskell (for submittion to proceedings of the recent LASER
> summer school) and I would like to seek initial feedback about its
> content: what information is probably missing? are there any subtle
> mistakes?

Not exactly the feedback you asked for, but I hope it still can be of
some use.

As a person with no prior knowledge about GADTs, I found your paper a
good introduction. Following proof of correctness of red-black tree
insertions turned out to be a little bit of a challenge as type
annotations quickly become tangled (made me wondering how one should
prove correctness of the proof).

I particularly liked how you handle things that are not central to
tutorial, like phantom and existential types: you give single-sentence
explanation that provides good enough intuition to follow you further.

Talking of quick explanations, I would love to see kinds and singleton
types explained in the same manner. You tried to explain singleton types
at page 13, but I find explanation provided by GHC documentation[1] to
be much more clear. As for kinds, it just puzzles me why you use the
term but don't explain it.

  1. http://hackage.haskell.org/trac/ghc/wiki/TypeNats/Basics

The same goes for type families - while I was able to quickly look up
definition of singleton types, I failed to comprehend basics of type
families by reading Haskell Wiki. I ended up pretending that type
families are just type-level functions.

I would love to see Yampa optimizations section expanded with more
interesting examples. Are there any?

Last but not least, a few minor typos and errors I spotted:

* at page 8, "Than we need to declare type instances..." should be "Then
  we need...";

* probably due to excessive editing, parameters to repeatElem at page 13
  are in the different order than before;

* at page 14, you state:

  > As in all binary search trees, for a particular node N c l x r
  > values greater than x are stored in left sub-tree (in l) and values
  > less than x are stored in right sub-tree (in r).

  But later on, your code contradict that statement by recursing to left
  sub-tree when x we look for is less than value in the node, and to
  right sub-tree when x is greater than value in the node.

Thank you once again for a nice introduction to GADTs!

-- 
Regards,
Alexander Batischev

PGP key 356961A20C8BFD03
Fingerprint: CE6C 4307 9348 58E3 FD94  A00F 3569 61A2 0C8B FD03

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://www.haskell.org/pipermail/haskell-cafe/attachments/20130109/850a8d80/attachment.pgp>