Will Crichton

that was an interesting read, especially once it got concrete with the case studies about concrete and proximal ways of thinking but much of the work on DSLs, programming by example, direct manipulation programming, and live programming incorporates these ideas today

the fact that languages that celebrate and enable abstraction (e.g. Haskell) are niche and there is massive resistance any time anyone so much as suggests teaching them even to CS majors suggests that they are not the "standard, canonical style" in the broader CS culture

Is the resistance *because* of Haskell's facilities for abstraction though? Cambridge teaches OCaml, Brown/Northeastern teach Racket, CMU teaches SML. I would suspect resistance comes more from lack of perceived applicability in industry.

Racket is not exactly industry-applicable though. And there are schools that teach Haskell as an introductory language — I believe UBC is one, if I remember right.

UBC uses the student languages in Racket for teaching. Related: the schools that "use Racket" are not using the full language, but rather special small languages used for teaching.

Oh right! I remember you saying that now. Dunno where I was thinking of tho I'm sure I've heard of at least one US school that uses Haskell as an intro lang. Still, I think "no industry applicability" is not the main factor in Haskell's lack of use as an intro lang.

Racket, Haskell, and OCaml are all used in industry, and what tools are used in industry doesn’t really matter to teaching computer science. We’re teaching how to build, not how to use a hammer.

Right! I... feel like I've not made my point very clear haha. Novices seem to perceive that these langs are equally useless in industry, yet Racket seems to fare better than Haskell as an intro lang. So I think the issue of industry applicability is not Haskell's problem.