Py3k(python3) more than one year on - 0.96 % packages supporting py3k.
renesd.blogspot.com
renesd.blogspot.com
What's the point of new features when hardly anyone uses them because not only their own code but also most of the important libraries don't support the new version?
What ends up happening is everyone keeps using the old version of Drupal until they're forced to upgrade or abandon it due to security concerns (old versions don't get patches from the security team).
It's a difficult matter: Let the language evolve and break things, or make it grow old and stale?
Careful ponderation should be made.
The only way I can see Py3k successful is if there are clear advantages to move to Py3k, like improved performance or extra vital features (like improvement of inline functions (lambda)). The improved performance is also a big question as teams like PyPy and Unladen Swallow are working on improving Python 2.6 and not Python3k.
I would not consider this a vital feature. You can already nest arbitrary functions, you just have to name them. However Tail-Call-Optimization would be a vital feature. (Alas, it's not going to happen.)
Uh not necessarily, and even if you are there are many situations where it's interesting and useful.
> or using variables inside your closure, so you might as well name it, too
Does not follow.
I think Python might end up in a situation like PHP4 and PHP5 - where tons of people still use PHP4, even if PHP5 holds tons of improvements and is around 5 years old.
Breaking backward compatibility to fix some essential things is an ok move, but I think it's only going to work if Python3k has a much better offering than it currently has.
I hope Python will go through the same process. It's not that Python 2.x is dead, but that we've scheduled its retirement.
yeah but programs which ran on 2.4 kernels still run on 2.6 right? So the situation is totally different.
It's dead in that precious "The technorati are planning on moving away from it in the next couple of years" sense, which has, like, virtually no relevance to anything else.
See also: Windows (and Microsoft in general), iPhone App Store, and Twitter. All dead, baby, dead as a doornail.
No. Python 3.0 was never expected to see widespread adoption, nor was it expected that libraries be ported to it. Ever. Python 3.1 is expected to have a better adoption rate, and library porting to start in earnest during its time and even more with Python 3.2 next year, especially with tools such as 3to2 appearing.
Everything is going according to the plan, and this plan is the reason why 3.1 was released less than 7 months after 3.0 (3.2 will probably be released early next year, putting barely a year between 3.0 and 3.2). That's also why 2.7 is in the works (and was already planned before the 2.6/3.0 releases), and there might even be a 2.8 (though that's unlikely).
edit: when you get down to it, those claiming Python 3 has failed are either morons or FUD-spreader, Fox-style. I'm not saying Python 3 has succeeded, but we won't know whether or not it has failed before the end of 2010 at the earliest. Right now, everything's going according to the plan.
Please remember the planning for Python 3000 started for real during the 2003 State of the Python Union. Meanwhile the release of Python 3.0 was in december 2008.
Well I don't know if I love you but you sound like somebody I could love.
One way to make it substantially better is to get the new [Distribute](http://pypi.python.org/pypi/distribute) finished and fast-tracked into 3.x (Distribute is a replacement for setuptools).
It's a believable number