Comments on: IronPython 2.0 and Jython 2.5 performance compared to Python 2.5

By: Дайджест ссылок №3 | В поисках успеха

Дайджест ссылок №3 | В поисках успеха — Mon, 08 Jun 2009 21:47:03 +0000

[...] любителей Python: сравнение производительности IronPython 2.0 и Jython 2.5 по сравн…. Графики показывают, что для питонщиков жизнь на .net [...]

By: Are the New JavaScript Engines Going to Completely Change the Game? « TechWorthy

Wed, 03 Jun 2009 20:24:04 +0000

[...] week, a gentleman wrote on his blog about the performance of CPython, Jython and IronPython in creating an empty binary tree to [...]

By: Are the new JavaScript engines going to completely change the game? « Occasionally sane

Tue, 02 Jun 2009 22:23:13 +0000

[...] the new JavaScript engines going to completely change the game? Last week, a gentleman wrote on his blog about the performance of CPython, Jython and IronPython in creating an empty binary tree to [...]

By: Robert Smallshire

Robert Smallshire — Wed, 27 May 2009 14:54:03 +0000

@ottorommel Thanks - it should be working now.

By: ottorommel

ottorommel — Wed, 27 May 2009 00:56:49 +0000

@Robert Smallshire
The ‘new article’ link gets a can’t preview drafts error.

Thanks,
Otto

By: Pinderkent: Losing developer time to performance problems hidden by high-level languages.

Sun, 24 May 2009 01:03:38 +0000

[...] good example of this is a performance issue described recently with IronPython, an implementation of Python for Microsoft’s .NET platform. In [...]

By: Robert Smallshire

Robert Smallshire — Sat, 23 May 2009 11:39:45 +0000

@Jon Harrop
I’ve seriously considered switching to F# for this project. I’ve been using F# on other unrelated projects with a fair amount of success – writing small compilers for DSLs for image processing operations. I have to admit though, in my shallow investigations into using fslex and fsyacc, they didn’t seem to be as debuggable with complex grammars as PLY, which I’m using with Python.

By: Jon Harrop

Jon Harrop — Sat, 23 May 2009 11:21:06 +0000

Use F# instead of Python. It is much faster and much better suited to compiler writing.

By: fijal

fijal — Sat, 23 May 2009 02:01:37 +0000

@wolf
If you try with –gcrootfinder=asmgcc, it will be even faster (32bit linux only though, a bit experimental)

@Nick craig-wood
The pure fact that this issue is well known should be enough to fix it (it’s not
even actually that hard). Doing random hacks and taking care by hand if you have
or not circular references does not sound like a solution to me.

cheers,
fijal

By: ita

ita — Sat, 23 May 2009 00:46:05 +0000

Scons does not support jython yet but waf does it.
The benchmark does not look too good against cpython
http://code.google.com/p/waf/issues/detail?id=434&can=1&q=jython#c9