3 ms·

2004. Almost 20 years ago. Was Python even a teen? Python does not attract smart young iconoclasts anymore. It's the scripting language everyone writes becaus

2004. Almost 20 years ago. Was Python even a teen? Python does not attract smart young iconoclasts anymore. It's the scripting language everyone writes because everyone writes Python. What languages do attract smart young iconoclasts? Nim? Elm? Whatever you're doing that is not yet old enough to drink by the time you comment?

treeform 4y ago
My vote is on Nim. Python like syntax with C++ speed. Unknown enough, but still not too new or obscure.
- girvo 4y ago
  My vote is for Nim too, but I'm decidedly biased. It's allowed us to do things in embedded that would've taken 3, 4 times as long and with more developers to achieve in C++.
- xhevahir 4y ago
  The community gives me pause. Nim's BDFL seems to have a habit of driving away prominent users of the language. Programmers who had contributed to the compiler have made their own hostile fork, and two of the three people who have written book-length introductions to the language have either given up on Nim or been tempted to do so. (The third is the BDFL himself.) If anyone cares, here's some comments from one of those authors: https://github.com/StefanSalewski/NimProgrammingBook/issues/10 https://github.com/StefanSalewski/NimProgrammingBook/issues/... . And from the other: https://twitter.com/d0m96/status/1592827547582332929?t=IqKBzSp9Gum1NuAnN2GO3w&s=19 https://twitter.com/d0m96/status/1592827547582332929?t=IqKBz... . These quarrels with Araq differ obviously, but both are pretty discouraging as far as Nim's prospects are concerned.
  - VectorLock 4y ago
    Asserting that the English language can't evolve is a weird take from someone who is trying to evolve a programming language.
  - zmmmmm 4y ago
    Interesting. I saw shades of this when I tried porting a tool over to Nim, found it was much slower than trivial code in Python and Groovy, then stumbled into this discussion: https://github.com/nim-lang/Nim/issues/9026 https://github.com/nim-lang/Nim/issues/9026 Nothing as strong as the above but it definitely rubbed me the wrong way. So much advertising about Nim being efficient/fast and the default way to read a file is incredibly slow and inefficient .... and they don't care.
    - planetis 4y ago
      That issue is about readLine specifically right? Not reading a file in general. The only time I get to ever use readLine is to solve adventofcode problems.
  - planetis 4y ago
    That's really petty argument against nim and it makes me sad really. 'Hostile' fork isn't true, just a fork. I wish you the best.
    - xhevahir 4y ago
      Most language fail to gain traction for reasons that don't have anything to do with the design of the language itself--even "petty" ones, such as the tactlessness of the people in charge. I don't want to use a technology which few other people use and which has, in my estimation, an unpromising future.
number6 4y ago
I am old.
nequo 4y ago
It would be an interesting future in which Koka[1] becomes the new Python. Or Lean 4,[2] and we end up with formally proved ML code in 2043. [1] https://koka-lang.github.io/koka/doc/index.html https://koka-lang.github.io/koka/doc/index.html [2] https://leanprover.github.io/documentation/ https://leanprover.github.io/documentation/
chillaranand 4y ago
2014 Almost 10 years ago, I translated this article to Telugu. https://avilpage.com/2014/12/python-paradox.html https://avilpage.com/2014/12/python-paradox.html
crabbone 4y ago
The hype around Python didn't age well. Python is on its way to become the new Java, almost literally trying to match it feature for feature. Interesting how it started to attract people with the same mindset that likes Java: great care for trivial correctness guarantees s.a. those ensured through type hints or producing programs through ticking boxes in existing templates. And disregard to large-scale design issues, something that was initially thought to be helped by more concise syntax and greater freedom to explore / less fear of novelty. However, the thought that dedication to a (marginal) language will help you hire better programmers isn't new, and has some supporting evidence. I remember similar argument coming from Linus Torvalds, when he didn't want C++ programmers working on Git, for example. I also happen to think that a lot of new(-ish) languages today are trying to ride the hype wave, and by doing so are trying to become instantly more popular rather than appealing to the tastes of a small but dedicated community. Eg. Rust is trying to be a "better C++", but in a way C++ programmers would most likely not be repulsed by it. Both superficial decision (s.a. "familiar" curly bracket syntax) and more fundamental ones are there to ensure that large audience of C++ programmers would have an easy migration path. I think that, today, if you wanted to match the effect the author saw created by hiring Python programmers in 2004, you'd have to go after hiring people dedicated to undeservedly forgotten languages s.a. Ada, Scheme or Erlang. Rust is posed to become just a slightly different flavor of C++ with the large volume of low-skill programmers pumping out low-quality software from big corporate shops. For now, it may attract enthusiastic people, but that's not going to last.

treeform 4y ago

My vote is on Nim. Python like syntax with C++ speed. Unknown enough, but still not too new or obscure.

girvo 4y ago
My vote is for Nim too, but I'm decidedly biased. It's allowed us to do things in embedded that would've taken 3, 4 times as long and with more developers to achieve in C++.
xhevahir 4y ago
The community gives me pause. Nim's BDFL seems to have a habit of driving away prominent users of the language. Programmers who had contributed to the compiler have made their own hostile fork, and two of the three people who have written book-length introductions to the language have either given up on Nim or been tempted to do so. (The third is the BDFL himself.) If anyone cares, here's some comments from one of those authors: https://github.com/StefanSalewski/NimProgrammingBook/issues/10 https://github.com/StefanSalewski/NimProgrammingBook/issues/... . And from the other: https://twitter.com/d0m96/status/1592827547582332929?t=IqKBzSp9Gum1NuAnN2GO3w&s=19 https://twitter.com/d0m96/status/1592827547582332929?t=IqKBz... . These quarrels with Araq differ obviously, but both are pretty discouraging as far as Nim's prospects are concerned.
- VectorLock 4y ago
  Asserting that the English language can't evolve is a weird take from someone who is trying to evolve a programming language.
- zmmmmm 4y ago
  Interesting. I saw shades of this when I tried porting a tool over to Nim, found it was much slower than trivial code in Python and Groovy, then stumbled into this discussion: https://github.com/nim-lang/Nim/issues/9026 https://github.com/nim-lang/Nim/issues/9026 Nothing as strong as the above but it definitely rubbed me the wrong way. So much advertising about Nim being efficient/fast and the default way to read a file is incredibly slow and inefficient .... and they don't care.
  - planetis 4y ago
    That issue is about readLine specifically right? Not reading a file in general. The only time I get to ever use readLine is to solve adventofcode problems.
- planetis 4y ago
  That's really petty argument against nim and it makes me sad really. 'Hostile' fork isn't true, just a fork. I wish you the best.
  - xhevahir 4y ago
    Most language fail to gain traction for reasons that don't have anything to do with the design of the language itself--even "petty" ones, such as the tactlessness of the people in charge. I don't want to use a technology which few other people use and which has, in my estimation, an unpromising future.

number6 4y ago

I am old.

nequo 4y ago

It would be an interesting future in which Koka[1] becomes the new Python. Or Lean 4,[2] and we end up with formally proved ML code in 2043. [1] https://koka-lang.github.io/koka/doc/index.html https://koka-lang.github.io/koka/doc/index.html [2] https://leanprover.github.io/documentation/ https://leanprover.github.io/documentation/

chillaranand 4y ago

2014 Almost 10 years ago, I translated this article to Telugu. https://avilpage.com/2014/12/python-paradox.html https://avilpage.com/2014/12/python-paradox.html

crabbone 4y ago

The hype around Python didn't age well. Python is on its way to become the new Java, almost literally trying to match it feature for feature. Interesting how it started to attract people with the same mindset that likes Java: great care for trivial correctness guarantees s.a. those ensured through type hints or producing programs through ticking boxes in existing templates. And disregard to large-scale design issues, something that was initially thought to be helped by more concise syntax and greater freedom to explore / less fear of novelty. However, the thought that dedication to a (marginal) language will help you hire better programmers isn't new, and has some supporting evidence. I remember similar argument coming from Linus Torvalds, when he didn't want C++ programmers working on Git, for example. I also happen to think that a lot of new(-ish) languages today are trying to ride the hype wave, and by doing so are trying to become instantly more popular rather than appealing to the tastes of a small but dedicated community. Eg. Rust is trying to be a "better C++", but in a way C++ programmers would most likely not be repulsed by it. Both superficial decision (s.a. "familiar" curly bracket syntax) and more fundamental ones are there to ensure that large audience of C++ programmers would have an easy migration path. I think that, today, if you wanted to match the effect the author saw created by hiring Python programmers in 2004, you'd have to go after hiring people dedicated to undeservedly forgotten languages s.a. Ada, Scheme or Erlang. Rust is posed to become just a slightly different flavor of C++ with the large volume of low-skill programmers pumping out low-quality software from big corporate shops. For now, it may attract enthusiastic people, but that's not going to last.