Static Typing Where Possible, Dynamic Typing When Needed | Hacker News Reader

Static Typing Where Possible, Dynamic Typing When Needed

lambda-the-ultimate.org

22 pointsArgorak9 comments

At least for me, the link to the actual paper is broken on that site. Here is another link: http://research.microsoft.com/en-us/um/people/emeijer/Papers...

Thank you. I was on a train at that time and sadly didn't bother to check all links using my constantly failing EDGE connection.

The impulse for making static typing less partial and more complete causes type systems to become overly complicated and exotic as witnessed by such concepts as "phantom types" and "wobbly types".

"Modern type theory is hard. Let's go shopping."

I doubt that Eric Meijer, one of the co-authors of the Haskell 98 spec, personally finds modern type theory too complex to deal with. He seems to be proposing that it's not necessarily a positive language feature to expose a very complex type system to the programmer (or at least, to make it mandatory).

If I was having a type theory argument there are few people in the world I'd rather have on my side. Few people understand the bridge between type theory and language design so well. His joint paper on type classes should be required reading in PL courses: http://research.microsoft.com/en-us/um/people/emeijer/Papers...

I agree with this comment: http://lambda-the-ultimate.org/node/834#comment-7653

I suspect that the objectionable paragraph was written by Drayton, that the note atop the introduction is not at all overstated, and that that paragraph would not have made it into a more polished version of the paper.

I really hate this habit of academic papers to never include a date.

<sigh>

So... when was this article published? It lists references from 2004, so between now and then, unless Erik possesses a time machine (I wouldn't put this past him).

I find this habit healthy. I regularly read papers from the 60s, 70s, and 80s and I judge them by their content. There is no point dismissing a good article because it's old.

The URL contains 'RDL04', so I'd guess 2004. Looks like it was presented at this workshop: http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.67.5...

Reply on news.ycombinator.com