[pypy-dev] External RPython mailing list

Maciej Fijalkowski fijall at gmail.com
Tue Sep 14 11:39:30 CEST 2010


Speaking from a personal perspective here, I would help people write
standalone executables using RPython. This has been tried (even with
success) for small examples and works. However, for most places where
it was tried, it was an ill-chosen tool for that purpose (where slight
python optimizations or using JIT would work equally well) with
RPython having sometimes bizarre limitations that we're not willing to
work on.

The second part (writing Python extensions) I think is not a very good
target, but can be done. However, I don't want people telling me that
I should work on it. If some people want to implement this, they can
get my help.

I think this defines rough outline where I'm (personally) willing to
help or not to help people using RPython.

On Tue, Sep 14, 2010 at 1:36 AM, Saravanan Shanmugham <sarvi at yahoo.com> wrote:
> I believe Hart's Antler has done quite bit of work in RPython.
> Yeah, that said, more work to do there.
> My goal with my RPython thread, is that I believe that there is an implicitly
> static subset of Python that can be compiled into standalone executables and
> DLLs without needing JIT or VMs.
> Can serve 2 purposes.
>   1. Make standalone executables just like C/C++ code.
>   2. Write Python Extension modules that can be compiled into shared DLL
> modules for CPython and PyPy
> Looking through the various threads on PyPy, Shedskin and Cython, I believe its
> just a matter of time.
> Sarvi
> ----- Original Message ----
>> From: Antonio Cuni <anto.cuni at gmail.com>
>> To: Maciej Fijalkowski <fijall at gmail.com>
>> Cc: pypy-dev at codespeak.net; Armin Rigo <arigo at tunes.org>
>> Sent: Mon, September 13, 2010 1:50:34 AM
>> Subject: Re: [pypy-dev] External RPython mailing list
>> On 13/09/10 10:27, Maciej Fijalkowski wrote:
>> > Is it really about  interpreters? (what's interpreter-specific after
>> > all in RPython) or is  it just that it's hard to use and does not
>> > integrate with CPython  well?
>> my point if that it's definitely good enough for writing  interpreters. For
> the
>> rest, it's a bit unknown (in the sense that nobody has  ever tried), and we
>> don't care about knowing  :-)
>> _______________________________________________
>> pypy-dev at codespeak.net
>> http://codespeak.net/mailman/listinfo/pypy-dev

More information about the Pypy-dev mailing list