@amiloradovsky And if I remember correctly from a talk from Edwin Brady where he was asked the opposite question about Idris, he said that the author of Miranda expected to write a lot more lists than type signatures, hence the shorter cons.

@typochon
Okay, that explains everything…

@typochon
From now on I'm going to put Haskell in the same bin with SQL, PHP, and COBOL, and also FORTRAN 77…

@newt
It's culture (often arrogant and inadequately pretentious), negligence for the ecological aspects, focus on short-term gain at the cost of long-term pain — typical industrial technology though, nothing special.

@amiloradovsky what do you mean by ecological aspects?

@newt
Externalities. Everything regarding infrastructure, sustainability, "climate" and "pollution": e.g. reaching and teaching aspects. See also "technical debt".
For instance, optimizing the language for writing by insiders, as opposed to reading by outsiders, is not a good approach in the long run.

It's hard to describe precisely, but I guess you've go what I mean.

@amiloradovsky I'm not sure I have exactly the same experience as you do. For me, learning Haskell was quite easy, though I had some Erlang and Lisp experience prior. Infrastructure isn't something I can complain about, especially since nix and stack became available. Compared to my experience with OCaml, writing stuff in Haskell is a bliss.

There as some dark corners such as lens and certain stuff by ekmett, but that isn't hard to avoid.

@amiloradovsky I must say I can agree with the technical debt part. Error handling in Haskell is abysmally bad, but so it is in almost any other language. This could change if somebody made a new base library, but it probably isn't gonna happen any time soon.

@newt
I suspect you also had quite a lot of experience with C++ before that…

My point is not that I personally find Haskell's concepts difficult to understand — no — that is how it's fans seem to hear any critique. (Although I indeed am not very familiar with it's libraries.)
I also learned Lisp and Prolog (but not Erlang), and even some Haskell, before embracing ML.
It's the constant sense of being fooled, like in a scam, that makes me uncomfortable: as if I'm taking part in a fraud.

@newt
WRT the last bit, see e.g. the answer by @typochon — guy wrecked the notation, widespread even in the literature (going back to the 1920's!), to make conses one symbol shorter… leaving lots and lots of people who studied math. logic confused. And nobody cared to fix that before Idris…
Another example is Bottom and Void types. Not to mention all the contractions and misused or unnecessary jargon. — No, I don't like that.

@amiloradovsky btw, what are the best OCaml projects to get more familiar with the language? I've been curious about it and Scala for quite some time, but I haven't had a chance to do something useful with them.

Writing a mastodon bot in OCaml, I can imagine, would be a great start.

@newt
I'm currently learning #Ocsigen, namely #Eliom. — Although this isn't my goal, I suppose it would be relatively easy to write a #Pleroma analogue in it…

@newt
P.S. If you're less into Web and more into the low-level stuff, check out #Mirage OS.

@amiloradovsky thanks, I have already seen it, as well as HaLVM and other library operating systems.