On Sat, Jan 26, 2019 at 09:19:11AM -0500, Matthias Felleisen wrote: > > > On Jan 26, 2019, at 9:03 AM, Sorawee Porncharoenwase > > <sorawee.pw...@gmail.com> wrote: > > > > Matthias, where can I find this "History of Clojure"? I searched for > > "It is better to have 100 transducers ..."
I'd love even to know what he *meant* by that. I always thought 10x10 made for more modularity. > > and found no result (besides this very thread). I also searched for > > "History of Clojure" and only found this tweet > > <https://twitter.com/richhickey/status/1058083678029537283> from Rich > > Hickey in Nov 2018, and it seems he was still writing it back then. > > Is it available for public to view now? > > > > > Unpublished notes. But I believe he said something like that during > his StrangeLoop presentation of transducers in Clojure. > > > > If we take an argument that we can encode a tuple with a list and > > get all benefits from list operations, why don't we (define-values > > (bool/c true false) (values int/c 1 0))? Why don't we (Church) encode > > everything to lambdas? > > > Functions are infinite data, lists are finite. The theoretical difference between the two lies in the Turing undecidability theorems. Functions in Racket are, however, finitely represented. The practical difference lies in efficiency. > You need to develop good taste for good compromises. > > I am the one who pushed PLT Scheme/Racket into the static corner > (e.g., structs). But the place in the middle is where the sweet spot > is. The Buddha's middle way. Everything in moderation, even moderation in moderation. -- hendrik > > — Matthias, radical centrist with nuanced opinions often stated in > provocative ways :) > -- You received this message because you are subscribed to the Google Groups "Racket Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to racket-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
