Syntax with Purpose in a Programming Language
5 points by schonfinkel
5 points by schonfinkel
I like the idea this expresses but I don't think the rant does it justice. I believe I saw a more thought through presentation of this idea from Slim Lim at our colloquium.
I'm the guy in the video! I am still writing an article on it that attempts to put it more in depth using semiotics.
Your reference seems fantastic. I wasn't aware someone was talking on this. I am actually glad.
I'm not exactly sure what they are ranting against here. Is the example from OCaml that let is used for both binding expressions inside a function and for binding inside modules, and that conflating the two by reusing the keyword is bad? I also don't think they articulated well the problem they see with the Clojure syntax compared to the ELisp syntax (I'm not even sure they claimed that there was a problem?).
Perhaps that was not so well articulated, but I wanted to say that expressing the idea of a list of bindings pairwise without any delimiters may sound like an impoverishment of meaning, but I believe it's warranted since the context is enriched with the usage of different symbols (in the case of Clojure a vector) and their composition comes to convey a different quality worthy of attention. In the F# and OCaml case, both usages, although different, are commingled into the same case with poor locality.
Thank you for taking the time to respond. I still don't get it, but I'll be awaiting your blog post :)