Stackless is an alternative C python. It actually ships all of python.
Stackless is a fork. './path/to/stackless/pythonthon mymodule.py' runs
the module.
Cython ships none of python. Cython requires that you already have a
python installed, and uses python as a library. './path/to/cython
mymodule.py' does not run the module.
Nuitka ships none of python. Nuitka requires that you already have a
python installed, and uses python as a library. './path/to/nuitka
mymodule.py' does not run the module.
How helpful for people who dont already know what it is to tell them its
another implementation of cpython, when you cannot even run any python
code with them? That you need another implementation already installed
to do anything with them, and the output product of both of these tools
is binaries linked.... NOT to the project, but to the other cypython you
had to install?
Technically, these two might have a tokenizer and lexer, and to an
extent a parser, though IIRC, they use cpython's ASTs (nuitka does at
least.), but what they output isnt useful without another python
installed. they implement NOTHING of python. Including them helps NO
ONE who didnt already know about them.
On 11/5/2015 18:33, M.-A. Lemburg wrote:
On 06.11.2015 00:08, Alexander Walters wrote:
They don't implement....anything. Guido implemented it.
Actually, it's Guido and the Python core dev team who implemented it
and the cast of thousands who help by sending in patches.
but it is SUPREMELY unhelpful to list them as stand alone third party
implementations of python,
SINCE THEY ACTUALLY REQUIRE ANOTHER PYTHON TO RUN! You CANNOT run a Nuitka
module without libpython
being installed. 'standalone mode' simply links against an already existent
cpython.
So while technically It MIGHT be considered another python (It isnt. Its a
CPython tool), it helps
no one to tell them its another implementation of the language.
I'm not sure why you are getting all upset. There's nothing
to be dogmatic about.
All of the implementations I listed below under "Alternative CPython"
use Python C code and the Python stdlib in one way or another -
that's why they are listed as *alternative* CPython. Unlike the already
listed repackaged Python versions, these implementations apply
changes to CPython or extend it in various way (by e.g. adding a JIT
compiler, adding additional syntax or compiling part of your
Python program to Python C extensions).
Now, if you want to continue the discussion, please provide constructive
input. Otherwise, we can agree to disagree and stop the thread
right here.
On 11/5/2015 04:49, M.-A. Lemburg wrote:
On 05.11.2015 00:14, Alexander Walters wrote:
To be absolutely clear, Nuitka and Cython are not implementations of the
language... because you
cannot run anything with them. Can you really call it a language runtime if
you cant run anything
with them? they are tooling. It is unhelpful to new users to list this
tooling as an
implementation when they cant do anything with them alone.
Both are compilers and they do allow you to create Python
applications, just like gcc does for C (together with the
libc runtime), so I don't really understand why you would
not call them implementations of the language. They both
take Python code on input :-)
I guess putting them under a "Python Compiler" heading
would satisfy your request.
On 11/4/2015 08:24, M.-A. Lemburg wrote:
On 31.10.2015 22:23, Alexander Walters wrote:
Nuitka and Cython cant really be called alternate implementations of c python,
since they kind of
require c python to run. They are tooling at best.
I'm not sure I follow. Cython runs all Python code, plus
manages extensions for more easily compiling functions
to C extensions. Nuitka tries to implement most of CPython.
Both support use of Python C extensions.
It's true that both need libpython - that's why I listed them
as "Alternative CPython" and not as "Alternative Python"
implementations :-)
On 10/29/2015 05:21, M.-A. Lemburg wrote:
Some implementations which are not listed on the page:
Alternative CPython:
HotPy - https://sites.google.com/site/makingcpythonfast/
Pyston - http://blog.pyston.org/
Cython - http://cython.org/
Nuitka - http://nuitka.net/
Python for embedded devices:
MicroPython - https://micropython.org/
This list is incomplete.
Could you provide pointers to missing items ?
Python in the browser:
Brython - http://www.brython.info/
Skulpt - http://www.skulpt.org/
as is this one (it does not even list the python in the cloud implementation
used on the front
page
of python.org)
You mean PythonAnywhere ?
I think there's a misunderstanding here. I just listed
alternative implementations which aren't already mentioned
on https://www.python.org/download/alternatives/.
Python in the cloud:
Trinket - https://trinket.io/python
Wakari - https://wakari.io/
Python for education:
Python Fiddle - http://pythonfiddle.com/
Geometry Zen - http://www.geometryzen.org/
Evaluzio - http://www.evaluzio.net/
More alternative CPython packagings:
StaticPython - https://code.google.com/p/pts-mini-gpl/wiki/StaticPython
The wiki page lists a few more:
https://wiki.python.org/moin/PythonImplementations
On 29.10.2015 09:56, M.-A. Lemburg wrote:
On 29.10.2015 09:55, M.-A. Lemburg wrote:
There's currently a ticket open for the alternative implementations
page:
https://github.com/python/pythondotorg/issues/842
The page currently lists a few variants of Python implementations,
but is not up to date with more recent developments such as
Brython, Pyston, HotPy, etc.
Forgot the link to the page:
https://www.python.org/download/alternatives/
I think we should try to add all currently maintained variants
to the page.
Would someone like to help with this ?
We'd need ReST entries of the form:
* `PyPy <http://pypy.org/>`_ (A `fast <http://speed.pypy.org>`_ python
implementation with a JIT compiler)
for each variant.
Thanks,
_______________________________________________
pydotorg-www mailing list
[email protected]
https://mail.python.org/mailman/listinfo/pydotorg-www
_______________________________________________
pydotorg-www mailing list
[email protected]
https://mail.python.org/mailman/listinfo/pydotorg-www
