3 ms·

> Scala as a better Java any more than I would Python as a better C The difference between Python and C is huge because C is not interpreted and doesn't do aut

> Scala as a better Java any more than I would Python as a better C The difference between Python and C is huge because C is not interpreted and doesn't do automatic memory management, which is actually the deal breaker for many apps. Saying that such a difference exists between Scala and Java is simply wrong. It's wrong starting from the fact that both are garbage collected, running on top of the same VM, having the same performance and thus targeting the same application profiles. > Scala is maybe a "worse Haskell", with extensive Java interoperability Haskell is a language I like and that I'm in the process of grokking, being a continuous inspiration, but to tell you the truth, to the contrary I think Haskell is a less practical Scala. Scala follows the pragmatism of the ML family. OCaml is another pragmatic language. Clojure and Racket are too. Empirically speaking, Haskell is good at being a trend setter in the FP community, but besides really cool libraries and concepts, I don't see much in the way of applications built by Haskell folks. Some big companies have built complex systems on top of Scala, not to mention projects like Kafka and Spark, which can also credit Java and the JVM. For OCaml, even with a much smaller community than both Scala and Haskell, you have some really cool stuff like MirageOS. What does Haskell have to show for, besides badly documented libraries and compiler implementations? Don't get me wrong, correlation does not imply causation, maybe Haskell simply has a popularity problem, but the onus is on the Haskell community to prove that Scala, OCaml and others are inferior alternatives to Haskell by showing some results. Because at the end of the day it's the final artifacts that really matter.

myth_drannon 10y ago
Elm compiler is written in Haskell.
pmilot 10y ago
> What does Haskell have to show for, besides badly documented libraries and compiler implementations? Off the top of my head, the Darcs revision control system, the awesome Pandoc tool (http://pandoc.org/ http://pandoc.org/), and the Xmonad window manager. All of these are fully-featured and battle-tested software that don't require any knowledge of Haskell to use on a day-to-day basis. Facebook also uses Haskell to write parts of its software stack, but most of that code is proprietary and not available for reading. There are many more examples: https://wiki.haskell.org/Haskell_in_practice https://wiki.haskell.org/Haskell_in_practice https://www.quora.com/What-are-the-most-famous-software-written-in-Haskell-and-why-arent-there-more-I-know-of-xmonad-and-that-Standard-Chartered-used-it-in-production-Are-there-others https://www.quora.com/What-are-the-most-famous-software-writ...
pinkythepig 10y ago
Haskell has lots of large projects written in it. Besides the ones others have mentioned, Haxl is probably one of the bigger known industry uses of it. It is used at facebook as their spam filter: https://github.com/facebook/Haxl https://github.com/facebook/Haxl If you scroll to the bottom of that github page, you can find several papers and articles that go into detail about it.
- saosebastiao 10y ago
  "Lots" is pretty relative. Haskell has lots compared to what? Arc[0]? It's not in the same league as Scala, that's for sure. And that also means that it is orders of magnitude away from the likes of C, C++, Java, Python, Perl, Ruby, etc. Haskell is and always will be a niche language for more experimental or research-y purposes that somewhat occasionally have real world uses. That shouldn't be a bad thing. Haskell, like Prolog, SML, Simula, Smalltalk, and Lisp, have influenced the world of programming far more than one could ever expect if we only measured success by large project usage. I don't see why Haskeller's need to defend their meager industry adoption so much. It has some pretty good adoption compared to most research languages, but its industry adoption is not that great for a general purpose language, and it doesn't need to be. [0] http://paulgraham.com/arc.html http://paulgraham.com/arc.html
cutler 10y ago
A lot of the early development of Perl 6 was implemented in Haskell.
TheCoelacanth 10y ago
Haskell was designed by academics to be used for programming language research. With that in mind, being a trend-setter seems to be a much better metric for success than actual industry adoption.

myth_drannon 10y ago

Elm compiler is written in Haskell.

pmilot 10y ago

> What does Haskell have to show for, besides badly documented libraries and compiler implementations? Off the top of my head, the Darcs revision control system, the awesome Pandoc tool (http://pandoc.org/ http://pandoc.org/), and the Xmonad window manager. All of these are fully-featured and battle-tested software that don't require any knowledge of Haskell to use on a day-to-day basis. Facebook also uses Haskell to write parts of its software stack, but most of that code is proprietary and not available for reading. There are many more examples: https://wiki.haskell.org/Haskell_in_practice https://wiki.haskell.org/Haskell_in_practice https://www.quora.com/What-are-the-most-famous-software-written-in-Haskell-and-why-arent-there-more-I-know-of-xmonad-and-that-Standard-Chartered-used-it-in-production-Are-there-others https://www.quora.com/What-are-the-most-famous-software-writ...

pinkythepig 10y ago

Haskell has lots of large projects written in it. Besides the ones others have mentioned, Haxl is probably one of the bigger known industry uses of it. It is used at facebook as their spam filter: https://github.com/facebook/Haxl https://github.com/facebook/Haxl If you scroll to the bottom of that github page, you can find several papers and articles that go into detail about it.

saosebastiao 10y ago
"Lots" is pretty relative. Haskell has lots compared to what? Arc[0]? It's not in the same league as Scala, that's for sure. And that also means that it is orders of magnitude away from the likes of C, C++, Java, Python, Perl, Ruby, etc. Haskell is and always will be a niche language for more experimental or research-y purposes that somewhat occasionally have real world uses. That shouldn't be a bad thing. Haskell, like Prolog, SML, Simula, Smalltalk, and Lisp, have influenced the world of programming far more than one could ever expect if we only measured success by large project usage. I don't see why Haskeller's need to defend their meager industry adoption so much. It has some pretty good adoption compared to most research languages, but its industry adoption is not that great for a general purpose language, and it doesn't need to be. [0] http://paulgraham.com/arc.html http://paulgraham.com/arc.html

cutler 10y ago

A lot of the early development of Perl 6 was implemented in Haskell.

TheCoelacanth 10y ago

Haskell was designed by academics to be used for programming language research. With that in mind, being a trend-setter seems to be a much better metric for success than actual industry adoption.