Wait, so the initial expectation was that small? I knew “avoid success at all costs”, but I was not aware of the lack of ambition.
Alright, then start your own thread - “kindaro’s consolation corner” - and we’ll redirect all the repetitious outpourings of unhappiness there for you to attend to.
It isn’t about “feeling good”, it’s about realising that constantly bemoaning problems are not a solution - it can potentially have the opposite effect: the constant distraction caused will not help the problem be solved any more quickly; the risk being the ongoing utterances will instead be treated as “background noise”, to be ignored along with the associated problem.
But, if after all that you’re still not convinced, ask parents who have taken children on long journeys what they think of hearing this:
- Are we there yet?
- Are we there yet?
- Are we there yet?
- Are we there yet?
-
Are we there yet?
⋮
1 Like
Julia already has this. Julia can link up to Python and R, as well as run its own libraries.
Clojure was built like this, also. Clojure is hosted on the JVM, and so can use the Java libraries as well as its own.
In both cases, the language has a lot more libraries from the off.
Anyone for Haskell on the JVM?
As I dimly recall, the main reason running Haskell on the JVM in any major way is encumbered by the fact that the JVM wasn’t designed to be a multi-language platform - it was more-or-less “tailor-made” for the Java language, which definitely isn’t functional.
Yes. That SPJ video that @jaror links to is his (rather personal) take on the vision.
It’s not lost; it’s still ever-present. But you can’t just produce a non-strict higher-order-typed fully-featured language (which was more or less done by ~2000) and then rest on your laurels.
(Not sure what you mean by “that small”.) As of early 1990’s who would have guessed Haskell would still be going 30 years later? All the mainstream languages were procedural with strict semantics. Landin had written about “The Next 700 Programming Languages” already in 1966. By 1990’s you could talk of another 700 and now another.
(The vision does seem to be lost on several of the more morose commenters here. We have already had several iterations of this – frankly – aimless bemoaning. Go write some beautiful code to remind yourself of the vision. Perhaps we could turn this thread into an appreciation of the vision.)
And I thought lisp was mainstream then.
It is it’s own skewed view on the Haskell ecosystem, completely ignoring the significant improvements of it wrt tooling.
And yet, people are constantly rehearsing these historical disputes, apparently not knowing that today, they pose no significant meaning anymore, because WE have moved on and there is in fact cross-pollination of ideas, which was not possible 5 years ago: RFC: Cabal support for LTS Snapshots · Issue #7556 · haskell/cabal · GitHub
So yes, get over it and make productive suggestions.
Haskell had no goals, except for being a research bed of ideas. What are you talking about?
There’s a Haskell Foundation, that was built to increase Haskell adoption in industry and facilitate said growth. Did you talk to them for participation?
I don’t see any new information here, that the vast majority of the community isn’t already aware of: we want to grow, yes. This topic is about the claim that Haskell is dying, which we all vehemently disagree with.
You haven’t made any productive arguments about how we can grow better.
What makes you think the lesson has not been learned? What is even the lesson? That there are competing ideas of usability? Yes there are, the main two are: 1. full control (unix APIs) and 2. batteries included.
Your claim we haven’t learned anything about that is, again, absolutely false. I repeat:
- RFC: Cabal support for LTS Snapshots · Issue #7556 · haskell/cabal · GitHub
- Let the extension manage GHC/cabal/stack? · Issue #540 · haskell/vscode-haskell · GitHub
- Haskell / ghcup-hs · GitLab
Fragmentation is (a sometimes sad) part of open source, but it’s an integral part of it. There is no way around it unless you start paying people and run the show like a company.
And competition is sometimes very very necessary.
Neither Python, nor Rust have any strong visions anymore, yet they are thriving a lot. Those are simply languages that work well.
And yet, Haskellers think we constantly have to chase more visions and technical fantasies to grow.
I think that is the main hindrance as to why Haskell is not growing more rapidly: it isn’t focusing on stability, performance or very boring library quality, but on visions.
But even if that changes, it will at best be the number one in functional programming (maybe it already is), not the number one across all programming paradigms. Beating imperative programming is not that easy.
4 Likes
After a bit of search it seems lisp was never really mainstream outside of AI:
I don’t believe that has happened. Lisp was never “popular”, except in very specific domains. In the AI world of the 80s it was popular but that was never a big community. ~quora
Largely because of its resource requirements with respect to early computing hardware (including early microprocessors), Lisp did not become as popular outside of the AI community as Fortran and the ALGOL-descended C language. ~wikipedia
Interesting, I thought AI was huge then. Guess it wasn’t. Anyway, python, with its ease of learning and scribbling nonsense, took the domain once dominated by lisp…
LISP is procedural. What do you think SETQ is about? And COND with its side-effects as it examines conditions. You can’t be an effective LISP programmer without thinking procedurally.
Also LISP was never mainstream – except in AI. And how many AI researchers were there then? And where were they so-called ‘working’?
Being thoroughly and resolutely non-strict and functional is a huge mindshift vs how everyone thought in the 1990’s. And a huge risk to even contemplate building such a language. And then how on earth to interface to C-language/foreign/procedural functions? to get all the low-level goodies and the vast intellectual capital from trusted libraries?
Perhaps you don’t realise the Haskell Prelude was nearly all written before there was even a working compiler? And a huge proportion of it just worked first time: that’s how much of a shared vision there was.
there was never a domain “dominated by LISP” except AI. Python is nowhere near the same domain. Python took over from BASIC – I mean Dartmouth BASIC (and similar), not VB; and BASIC was dumbed-down FORTRAN for the PC world.
Is this thread all because you don’t understand the history of programming languages?
Let me put that less harshly: because you don’t understand the context of how programming languages evolve, and how difficult it is to start a new language, let alone keep it going. Of the 700 + 700 + 700. Nearly all are dead. The only ones that survive 30 years have huge commercial backing.
And there seems little logic as to which survive/which don’t. It’s still a mystery to me why C wasn’t strangled at birth: it started as an undergraduate prank, it’s only got worse.
I see, btw the thread was expressing my feeling/concern about haskell. That haskell/FP might have been dying for past 10 yrs. As it seemed so in my small country.
This is a fascinating and wide-ranging conversation! I am learning a lot from it. I hope people don’t feel it is too fraught a conversation. The feeling that something we love is “dying” can cause stress and anxiety that can easily go unnoticed. Remember that it’s OK to take a step back, check out for a bit, have a breather, recuperate, before jumping back in to discuss this wonderful programming language that we love so much! (And that’s a reminder to me as much as to everyone else.)
12 Likes
Could you share the discord gamedev link? I’m interested in making games as an education tool. Mostly cloning.
1 Like
I don’t really like these split ecosystems. I don’t like the idea of a bunch of OO c# mixed with f#, etc. I am really unsure what sharing between clojure, scala, f#, and haskell would actually look like in practice. I don’t think they would want to share libraries.
Maybe the sharing would be more in just idea cross polination. Maybe its just stopping any language specific conferences, and only having functional “zoo” focused conferences. Maybe its participating in a shared Discourse instance for pure FP or emphasizing the FP slack over any language specific slack / zulip instances. Maybe it’s having working groups that go across languages, so a data science group that had members from clojure, haskell and scala talking about efforts within each of their languages.
An enormous slice of the nice pure-Scala libraries are “we adapted something that someone else did in Haskell” (which is usually a mindscrew because of goofy subtyping and strictness by default and various JVM weirdness). I don’t see as much going back the other way, but it might be interesting to see how much grows organically over there and how much is just frustrated Haskell programmers who need the safe paycheck of JVM compat.
3 Likes
Why do you think that you have a privileged view as to what will be beneficial for the public discourse, even while at least one person (me) disagrees with you directly and others behave in a way that is contradictory to your view? Whence your desire to contain dissent? You do not get to speak as if you can decide what helps and what hinders the evolution of the Haskell community and ecosystem, what should be rewarded and what should be punished, contained, eradicated. Your attitude of routine dismissal towards honest expressions of the thoughts and feelings of others makes me think you should be the last person to be given any such authority.
You are also likely to be wrong. The evolution of public discourse is a hard topic. Long term predictions are hard to make. You have not given any theory nor evidence in support of your claims. Rationally, what you are saying is just noise. In your own words, a distraction.
We are not talking Software Engineering here. We should be twice careful. Maybe your position can be repaired. But you are not making a case for it. I see myself as protecting the common freedom and respect, and yourself as attempting to steal the authority that does not belong to anyone. This is not the way I want to go. We can easily de-escalate this situation by agreeing that speech should be free and respect and gratitude should be the default attitude towards anyone who is kindly and generously sharing their experience and perception.
The link is not working. Says «Invite Invalid».
