The ideas in Finelli’s talk are fascinating, but the talk was based much more heavily than I had realised on the paper propositions as types (pdf) which I have read most of. Finelli was not a bad speaker, but Wadler is a master of the writing art and on this one I’d recommend just grabbing the paper.
Also, it was probably a bit rude to define “type system” to exclude dynamic and unityped languages at a conference with Erlang and Clojure speakers! :D
I think the presentation makes a good point in figuring out the importance of type systems in programming.
On the other hand I guess the message was a bit lost within the process of building the relevant historical evidence.
My "aha moment" was the quote about "unityped systems", but I suppose that the presentation could be refined to better outline the main point of the discussion: the relevance of strong types as a useful tool.
At least from what I made out of the talk.
Comments
Comments are closed.
I really liked the talk.
Could have gone even a bit more into the topic of type theory. And keeping the history and introduction to it a bit shorter.
Michele's talk was really interesting and I liked a lot the way he provided us details about the context this theory was born in.
Really interesting. So we lack one true type system because there are so many logics!
The ideas in Finelli’s talk are fascinating, but the talk was based much more heavily than I had realised on the paper propositions as types (pdf) which I have read most of. Finelli was not a bad speaker, but Wadler is a master of the writing art and on this one I’d recommend just grabbing the paper.
Also, it was probably a bit rude to define “type system” to exclude dynamic and unityped languages at a conference with Erlang and Clojure speakers! :D
I think the presentation makes a good point in figuring out the importance of type systems in programming.
On the other hand I guess the message was a bit lost within the process of building the relevant historical evidence.
My "aha moment" was the quote about "unityped systems", but I suppose that the presentation could be refined to better outline the main point of the discussion: the relevance of strong types as a useful tool.
At least from what I made out of the talk.
thanks everybody for the feedback: to Michael Newton: you are abosultely true about how Wadler is a master writer
I cited him a lot, but I had the temptation to just read excerpts of his paper, instead of writing my own presentation
so, if you got anything from my talk I hope that it was the message: go and read the papers by Philip Wadler