Hacker News Reader

Comment by richhickey | Hacker News Reader

I make no claims to novelty, and the blog post does link to http://www.haskell.org/haskellwiki/Enumerator_and_iteratee, the most similar work and a clear influence. If more people knew about Iteratee, it would be worth spending more time talking about the connections and contrasts, but they don't, and knowledge of it is not prerequisite to understanding reducers. No one helped me write the code.

View on HN

Isn't foldr/build fusion much closer? A collection is represented by a "build" function that takes the reducer, and list transformers become reducer transformers. The main difference is that it's applied automatically by list library using rewrite rules, so it's not as obvious, the reducer is supplied as a pair of "cons" function and "init" value rather than a variadic function, and there's no parallel fold.

http://research.microsoft.com/~simonpj/papers/deforestation-... The first paper (yes, that's a .ps.Z - check the date)

http://www.scs.stanford.edu/11au-cs240h/notes/omgwtfbbq.html some recent slides, which also include a bit about a "stream fusion" which isn't yet in the standard library

http://darcs.haskell.org/cgi-bin/gitweb.cgi?p=packages/base.... The details from GHC's libraries - it's all in the RULES.

> and there's no parallel fold.

When working up from the bottom it might seem that this is just manual stream/list fusion. But the parallelism is the key prize. We need to stop defining map/filter etc in terms of lists/sequences, and, once we do, there is nothing to fuse, no streams/thunks to avoid allocating etc, because they were never essential in the first place, just an artifact of the history of FP and its early emphasis on lists and recursion.

Yes, foldr/build is almost exactly reducibles, but not foldables.

Iterators do nothing for parallelism either.

First, thank you for implementing and promoting this approach to collections.

I did see that link, because I liked the concept enough to read it all the way through without knowing Clojure, but that's not quite what I had in mind. You're right that most people have never heard of that library, which is why the way you presented it will leave most people with no idea that it was an influence (even if they read that far). That's something you could have just said, in one sentence, without getting into detail about how it was different.

I'm not really trying to criticize you for saying or not saying certain things, and I don't think you did anything wrong. Not really acknowledging influences is just a symptom of what turned me off. I feel like this post was written from a sort of aggressive fighting-for-popularity mindset that I'm uncomfortable with in a language.

EDIT: missing word

This can't really be true because there has never been anything "clear" about enumerators and iteratees. Maybe a dozen people understand what's going on there.

Also, most of what's special about enumerators is the gymnastics needed for Haskell's type system, which isn't relevant to Clojure at all...

Reply on news.ycombinator.com