Thanks for that excellent comment. I must say I haven't watched the presentation (yet!), but your comment touches on something that really irks me when people

Thanks for that excellent comment. I must say I haven't watched the presentation (yet!), but your comment touches on something that really irks me when people who mostly use unityped languages comment on staticially type-checked languages. It very often turns into this weird thing where you just hear complaining about "types getting in your way", whereas I personally tend to see it as "the types are there to help you understand what you're actually doing". Maybe it stems from a difference in approach. I'll usually practice what's been called "typeful programming" where I just write all the type declarations for data/functions and use "undefined" ("???" for Scala folks) as the implementation just to see if the whole thing fits together before trying to code up any of the details. The types often lead to an "obvious" implementation of any given function. This is very similar to top-down programming. It seems to me that most "dynamic" programmers (esp. of Lisp-derived languages) tend to do things bottom-up. I think this may be at the root of a sizable portion of the "gets-in-the-way"/"there to help you" schism.