Stackless Python, eventual merge?

[Martin v. Loewis]

Ian Bicking <ianb at colorstudy.com> writes:

> I don't think this represents the current state of Stackless --
> Christian has tried to get it into the standard Python, and it is not
> for lack of desire or effort that that did not happen.  I think it could
> be considered a difference of opinion between him and Guido -- in part,
> Stackless introduces more complexity which Guido did not want to
> maintain, and in part because Stackless is not portable to forms of
> Python (not to Jython in particular, and certainly any .NET Python --
> probably a Parrot Python as well).
> 
> So the issue is primarily technical/political, and not a matter of
> resources.  

It's definitely not political, and certainly not the "current state".

I looked at the old Stackless patches once, and spotted significant
amounts of unexplicable code - code that did not serve any obvious
purpose, was not documented, or clearly should have been removed.  So
the old Stackless patch was incomplete for integration into Python -
it would have taken efforts to complete it, and to work with Python
maintainers to integrate it. Nobody went through this effort (and
again, I don't blame anybody for not doing that).

>From a shallow inspection, the code merely looked complex - but it
really was complicated, and complex is better than complicated.

However, this is *not* the current state. The current implementation
(which I have not reviewed, yet) supposedly is much simpler, so past
experience with trying to integrate it does not apply.

Regards,
Martin