4 ms·

Maybe Lush? http://lush.sourceforge.net/lush-manual/f0288067.html http://lush.sourceforge.net/lush-manual/f0288067.html Rather than a Lisp that runs on the JVM

by hhm 19y ago

Maybe Lush? http://lush.sourceforge.net/lush-manual/f0288067.html http://lush.sourceforge.net/lush-manual/f0288067.html Rather than a Lisp that runs on the JVM, I'd be very interested in knowing a Lisp that somehow runs as Python bytecode. Does such a monster exist?

danohuiginn 19y ago
Googling turns up a few lisp-in-python implementations: <a href="http://www.biostat.wisc.edu/~annis/creations/PyLisp/ http://www.biostat.wisc.edu/~annis/creations/PyLisp/">Pylisp</a>, <a href="http://www.xs4all.nl/~yduppen/site/psyche.html http://www.xs4all.nl/~yduppen/site/psyche.html">psyche</a> and <a href="http://hkn.eecs.berkeley.edu/~dyoo/python/pyscheme/ http://hkn.eecs.berkeley.edu/~dyoo/python/pyscheme/">pyscheme</a> (the last two being scheme, obviously). Haven't tried using them; maybe I will the next time I feel like frittering away a couple of hours.
- danohuiginn 19y ago
  clearly html links are the wrong way to go around here ;)
- hhm 19y ago
  Thanks a lot! I had googled for that, but couldn't find anything interesting.
henning 19y ago
Lush is great for numerical prototyping, much better than MATLAB/Octave. Its matrix/tensor engine is one of the most awesome (as in being in awe) pieces of software I've ever come across. It is definitely mature and well-tested in that niche. It was created by leading neural network researchers. Outside of its niche, it's pretty useless.
- viergroupie 19y ago
  I don't quite get the faith Lush seems to inspire in some of its users. It's terribly documented, has antiquated semantics (no closures), and isn't even type safe (you can segfault if you pass an object of the wrong type). I might consider using Lush when performance was absolutely critical, but otherwise I don't think it holds a candle to Matlab for prototyping.
jey 19y ago
Why? So you have access to the Python libraries from Lisp? That would probably be better accomplished by writing a layer that used the Python/C API to fbnd the functions into Lisp, rather than compiling the whole program to run on the (slow) python interpreter.
- hhm 19y ago
  No, I just was thinking in some Lisp implementation that didn't require to install the JVM, but that could use the bytecode of a lighter language, like Python. Also because I suppose if a Lisp implementation is generating code for a specific platform, its documentation will refer to such platform, and I'd prefer to read about Python bytecode than about JVM. But maybe I'm wrong and it's just useless prejudice of my part.

by hhm 19y ago

danohuiginn 19y ago

Googling turns up a few lisp-in-python implementations: <a href="http://www.biostat.wisc.edu/~annis/creations/PyLisp/ http://www.biostat.wisc.edu/~annis/creations/PyLisp/">Pylisp</a>, <a href="http://www.xs4all.nl/~yduppen/site/psyche.html http://www.xs4all.nl/~yduppen/site/psyche.html">psyche</a> and <a href="http://hkn.eecs.berkeley.edu/~dyoo/python/pyscheme/ http://hkn.eecs.berkeley.edu/~dyoo/python/pyscheme/">pyscheme</a> (the last two being scheme, obviously). Haven't tried using them; maybe I will the next time I feel like frittering away a couple of hours.

danohuiginn 19y ago
clearly html links are the wrong way to go around here ;)
hhm 19y ago
Thanks a lot! I had googled for that, but couldn't find anything interesting.

henning 19y ago

Lush is great for numerical prototyping, much better than MATLAB/Octave. Its matrix/tensor engine is one of the most awesome (as in being in awe) pieces of software I've ever come across. It is definitely mature and well-tested in that niche. It was created by leading neural network researchers. Outside of its niche, it's pretty useless.

viergroupie 19y ago
I don't quite get the faith Lush seems to inspire in some of its users. It's terribly documented, has antiquated semantics (no closures), and isn't even type safe (you can segfault if you pass an object of the wrong type). I might consider using Lush when performance was absolutely critical, but otherwise I don't think it holds a candle to Matlab for prototyping.

jey 19y ago

Why? So you have access to the Python libraries from Lisp? That would probably be better accomplished by writing a layer that used the Python/C API to fbnd the functions into Lisp, rather than compiling the whole program to run on the (slow) python interpreter.

hhm 19y ago
No, I just was thinking in some Lisp implementation that didn't require to install the JVM, but that could use the bytecode of a lighter language, like Python. Also because I suppose if a Lisp implementation is generating code for a specific platform, its documentation will refer to such platform, and I'd prefer to read about Python bytecode than about JVM. But maybe I'm wrong and it's just useless prejudice of my part.