13

My project has deferred solution on integration C++ with Python until now. But now this question is raised again. Do anybody know status of project? On Google: (http://code.google.com/p/unladen-swallow) there is no actual info at all. Can we reckon on this project?

Dewfy
  • 23,277
  • 13
  • 73
  • 121

2 Answers2

6

Unladen Swallow is dead in the sense that activity is low and there seems little hope of being a standalone Unladen Swallow branch like Jython or Ironpython.

That said, a lot of the optimizations and advances are going to get merged into CPython. Moreover, there is still hope that the Unladen Swallow team will continue to make their optimizations, as PEP 3003 is designed to allow other branches of Python (namely IronPython and Jython but Unladen too) catch up with the C implementation.

But no, you can't reckon on Unladen Swallow. It's unlikely it will ever stand on its own two feet in the near future or, most likely, ever. You may need to find another solution for C++ integration with Python (CPython isn't that slow, so maybe try it?)

Rafe Kettler
  • 75,757
  • 21
  • 156
  • 151
  • 2
    The intent was never to make a 'standalone Unladen Swallow interpreter'. It was always a project for experimental optimizations that would be merged back into CPython. That's still planned to happen, see PEP http://www.python.org/dev/peps/pep-3146/ – Thomas Wouters Oct 04 '10 at 13:59
  • @Rafe Kettler, okay, PEP gives a hope, but this arises another question - when this PEP can be released? – Dewfy Oct 04 '10 at 14:42
  • @Dewfy: I'm not sure what you mean. Do you mean when will they resume development on Python, or when will Unladen Swallow's optimizations merge into CPython? – Rafe Kettler Oct 05 '10 at 03:00
  • @Rafe Kettler, I've meant when CPython will be optimized according this PEP. Actually we are stare at Unladen Swallow because of promises to avoid GIL. So it is also unclear if PEP supposes this kind of optimization. – Dewfy Oct 05 '10 at 10:10
  • @Dewfy: getting rid of GIL is hinted at in the PEP, so it's safe to assume that when the PEP was written someone (probably Unladen Swallow) was considering doing such a thing. – Rafe Kettler Oct 05 '10 at 12:35
4

It is now five months since this question was asked, and Reid Kleckner has posted his own retrospective on Unladen Swallow. While he isn't listed as owner, most of his observations make it likely that Unladen Swallow is finished.

Don Wakefield
  • 8,693
  • 3
  • 36
  • 54
  • 2
    Many of the unladen devs at Google went on to create the Go language. They are much more interested in that now. – Keith Mar 27 '11 at 00:59
  • it's interesting. Looks like this branch is dead. Only this http://python.org/dev/peps/pep-3146 promising something – Dewfy Mar 29 '11 at 10:10
  • @Dewfy, I don't know the minutia of PEPs, and can't find a history for this document, so maybe it has always had the section: PEP Withdrawal, where it says "With Unladen Swallow going the way of the Norwegian Blue [1] [2], this PEP has been deemed to have been withdrawn.". Is that what you intended to point out? – Don Wakefield Jun 03 '13 at 23:24
  • @DonWakefield actually my interest in Unladen Swallow based on ideathat some python implementation would allow my project resolve GIL problem. So PEP gave me some hope on it. but my bad - i miss phrase on withdrawal. – Dewfy Jun 04 '13 at 07:32
  • @Dewfy, have you looked at one of the Python variants? I understand IronPython can do threading nicely. With an adjustment of paradigm, you might be able to adopt Stackless Python... – Don Wakefield Jun 04 '13 at 14:59
  • 1
    @DonWakefield, well, IronPython is really wonderful - I had opportunity apply it on my previous project, but now there is cross-platform requirement. Actually there are 3 real requirement on python: (1)-linux/windows compatibility /so IronPython dropped-off/, (2)-Python3 compatibility /so PyPy is in beta only/, (3)-boost integration /that is why I'm not sure on Stackless - correct me if U'm wrong/ – Dewfy Jun 05 '13 at 06:53