6 ms·

I hear this a lot but my experience tells me otherwise. For example, I worked on a Clojure project that would leverage this supposed “10x” power of lisp. Despi

by mjul 5y ago

I hear this a lot but my experience tells me otherwise. For example, I worked on a Clojure project that would leverage this supposed “10x” power of lisp. Despite hiring the even key Clojure contributors, this hypothesis was falsified. The project went from around 35 to over 400 people and it is moving at a the speed of glacial ice. Once you have a project with more than a few people a better language does not have much impact: it is all dominated by people factors and how to fit the project into its larger context.

wirrbel 5y ago
Any "10x argument" is just lazy thinking. 'Productivity' is governed by so many factors, its just unrealistic that 'choice of programming language' would have such an enormous effect (unless you benchmark CL against let's say COBOL). On a fun note: The old guard of Common Lisp (like Norvig IIRC) was very fond of Python when it hit mainstream because for them it had a whole lot of the feature set that lisp had which other popular languages lacked. [disclaimer: I am a die-hard Lisp and Clojure fan but the other fans just make me cringe at times]
- pjmlp 5y ago
  Pity that Python did not embrace the Lisp features related to machine code generation.
  - agumonkey 5y ago
    You mean having bytecode and native code ? maybe they can start that for 3.2+ .. surely it would be more interesting than optional type signatures :)
    - pjmlp 5y ago
      Yes.
      - agumonkey 5y ago
        It's interesting, lisps, [o|oca]ml .. have often both, but somehow this practice never went mainstream.. even js got huge influx of money to push vm super fast but no native code practices.
        pjmlp 5y ago
        I consider Java and .NET mainstream enough. :)
        agumonkey 5y ago
        traumatic selective memory at work
- mbrodersen 5y ago
  I agree. And I think that the constant BS claims of certain Lisp advocates actually damages the Lisp community. Especially the silly “I am smarter because I use Lisp” claims you often see in Lisp articles. My answer is always: OK great. Why don’t you instead first build some impressive stuff in Lisp and then show us how Lisp made it possible for you to build it 10x faster. Then you might actually make people believe you. Especially given that pretty much all successful software out there is written in (or run on top of) C/C++.
  - medo-bear 5y ago
    this is just personal taste but i think that GNU Emacs blows any other software ever written out of every universe that ever existed :)
    - mbrodersen 5y ago
      Ask Vi lovers if they agree :)
      - medo-bear 5y ago
        of course they do. they just chose evil-mode
        mbrodersen 5y ago
        Hahaha you are funny.
    - mbrodersen 5y ago
      Maybe because it is written in C? Hehehehe :)
      - medo-bear 5y ago
        its written in emacs lisp and c. most of the code base is in the former :)
medo-bear 5y ago
when i wrote this reply i didn't expect it to generate such a discussion. it was meant as a sort of smart-arsy opposite side to the same coin of which the other side was parent's claim that lisp is not suitable for large teams. the claim that lisp is not suitable for large teams seems ridiculous to me, especially given the prevalence of javascript and python in industry. surely lisp cannot be worse in large teams than these languages. usually people mention type safety in support of these claims (which again makes no sense given the prevalence of js and python), but this point is even more mute now given coalton's existence which allows you to have haskell-like type safety in common lisp - https://coalton-lang.github.io/ https://coalton-lang.github.io/ on the other hand i think people can be more productive in lisp provided certain criteria are met: they are motivated to solve the problem at hand; and that they enjoy and are capable to use the tooling that is offered in lisp to its full extent. but this is not surprising because in order to understand lisp to this extent you have to understand far more about computing than just knowing lisp syntax, which is trivial btw
- mbrodersen 5y ago
  Look I know Lisp quite well, have written more than one dialect of Lisp for fun, and have 25+ years of professional software development experience behind me. I don’t think there is anything you need to somehow understand better about “computing” to use Lisp. But you are welcome to convince me otherwise. Do you have any concrete examples?
  - medo-bear 5y ago
    > I don’t think there is anything you need to somehow understand better about “computing” to use Lisp I am not saying that at all. If anything I said the opposite of that. I think, given how simple Lisp's syntax is, being able to write/use Lisp is almost trivial. However being able to use Lisp is a way that has given it an almost mythical status (one being discussed in this thread is the so-called "x10 programmer") requires far more than just being able to write Lisp
    - deterministic 5y ago

by mjul 5y ago

wirrbel 5y ago

Any "10x argument" is just lazy thinking. 'Productivity' is governed by so many factors, its just unrealistic that 'choice of programming language' would have such an enormous effect (unless you benchmark CL against let's say COBOL). On a fun note: The old guard of Common Lisp (like Norvig IIRC) was very fond of Python when it hit mainstream because for them it had a whole lot of the feature set that lisp had which other popular languages lacked. [disclaimer: I am a die-hard Lisp and Clojure fan but the other fans just make me cringe at times]

pjmlp 5y ago
Pity that Python did not embrace the Lisp features related to machine code generation.
- agumonkey 5y ago
  You mean having bytecode and native code ? maybe they can start that for 3.2+ .. surely it would be more interesting than optional type signatures :)
  - pjmlp 5y ago
    Yes.
    - agumonkey 5y ago
      It's interesting, lisps, [o|oca]ml .. have often both, but somehow this practice never went mainstream.. even js got huge influx of money to push vm super fast but no native code practices.
      - pjmlp 5y ago
        I consider Java and .NET mainstream enough. :)
        agumonkey 5y ago
        traumatic selective memory at work
mbrodersen 5y ago
I agree. And I think that the constant BS claims of certain Lisp advocates actually damages the Lisp community. Especially the silly “I am smarter because I use Lisp” claims you often see in Lisp articles. My answer is always: OK great. Why don’t you instead first build some impressive stuff in Lisp and then show us how Lisp made it possible for you to build it 10x faster. Then you might actually make people believe you. Especially given that pretty much all successful software out there is written in (or run on top of) C/C++.
- medo-bear 5y ago
  this is just personal taste but i think that GNU Emacs blows any other software ever written out of every universe that ever existed :)
  - mbrodersen 5y ago
    Ask Vi lovers if they agree :)
    - medo-bear 5y ago
      of course they do. they just chose evil-mode
      - mbrodersen 5y ago
        Hahaha you are funny.
  - mbrodersen 5y ago
    Maybe because it is written in C? Hehehehe :)
    - medo-bear 5y ago
      its written in emacs lisp and c. most of the code base is in the former :)

medo-bear 5y ago

when i wrote this reply i didn't expect it to generate such a discussion. it was meant as a sort of smart-arsy opposite side to the same coin of which the other side was parent's claim that lisp is not suitable for large teams. the claim that lisp is not suitable for large teams seems ridiculous to me, especially given the prevalence of javascript and python in industry. surely lisp cannot be worse in large teams than these languages. usually people mention type safety in support of these claims (which again makes no sense given the prevalence of js and python), but this point is even more mute now given coalton's existence which allows you to have haskell-like type safety in common lisp - https://coalton-lang.github.io/ https://coalton-lang.github.io/ on the other hand i think people can be more productive in lisp provided certain criteria are met: they are motivated to solve the problem at hand; and that they enjoy and are capable to use the tooling that is offered in lisp to its full extent. but this is not surprising because in order to understand lisp to this extent you have to understand far more about computing than just knowing lisp syntax, which is trivial btw

mbrodersen 5y ago
Look I know Lisp quite well, have written more than one dialect of Lisp for fun, and have 25+ years of professional software development experience behind me. I don’t think there is anything you need to somehow understand better about “computing” to use Lisp. But you are welcome to convince me otherwise. Do you have any concrete examples?
- medo-bear 5y ago
  > I don’t think there is anything you need to somehow understand better about “computing” to use Lisp I am not saying that at all. If anything I said the opposite of that. I think, given how simple Lisp's syntax is, being able to write/use Lisp is almost trivial. However being able to use Lisp is a way that has given it an almost mythical status (one being discussed in this thread is the so-called "x10 programmer") requires far more than just being able to write Lisp
  - deterministic 5y ago