I posted this on reddit but I feel the need to repeat myself because nobody seems to understand:
The reason Python3 isn't ready is because most modules that the community needs are not supported by it and are 2.* only; Django by itself does not an entire python ecosystem make (other options exist but they have the same flaws of not being supported for python3, among other issues), and unless I can DEPLOY my app - good luck finding a host that natively supports it as not a SINGLE "big cloud" company supports a very recent version of 3.*! - this is really an academic discussion.
To repeat this because its important: Python 3.3 support for personal small projects is all well and good, but unless I can actually DEPLOY my app in a production system supported by "the enterprise" in the "cloud" and use the modules I expect and depend on (and right now cant because they are not currently compatible with python 3), this is an academic discussion.
Django is great as an ORM, but it by itself does not make a python ecosystem, especially given that it refuses to play nice with others in the community. I agree that its a driving force, but even its own support for python 3.* is not perfect. There are so many other projects that deserve attention (Pyramid, Flask, Tornado, to name a few) by the community.
The fact is, Python3 as a language is NOT supported by the most important places: Where we deploy apps. Good luck trying to deploy python 3.3 to AWS or Azure without a lot of hassle, because only 2.7 is supported natively for things like what Elastic Beanstalk provides (Read: I select it from a drop down and click submit, or its the default install so I have no extra work) on both. Sure, I can buy hosting with my fiends uncle or manually set up a server of my own to support it, but until I can press a button and use my credit card to get cheap, effective hosting for my python3n apps that scale with demand like I can with PHP, etc its just not going to take off. And this is further complicated by the total lack of support for Python3 for most core projects/modules (although thats less of an issue thanks to the python wall of shame/superpowers)
You want better python3 support? The open source/python community needs to get its butt in gear and do whats needed. No more "5 year plan" crap. Lets do a 1 month plan. And Red Hat needs to stop being theretarded distro with the oldest version installed, as they are actively harming the dev community since between them and centos and fedora and all thier other copycat distros thats a lot of installs with old python libs as the oldest python version in the newest version of redhat.
The python3 dev community should pick a project thats standing in the way of python3 adoption and port it. When its done, move to the next one. What project should they attack this month?
We need better support for newer versions by default in Centos Linux, Redhat Linux, Ubuntu Linux, and the Debian Linux Spinoffs so the cloud providers have to do less work to support it by default. Its getting better but its still not there. Amazon and Microsoft need to get their heads out of the ground and support it without hassle or work needed by their customers, but for Amazon at least if the linux distros support newer versions, so will they by default. If the Python community wants to win against other languages (or at least get better support for Python3.3) it needs to start there and get itself the default version of python to be 3.3+ in all of them
After that, getting it supported in AWS, Azure, etc so apps in Python3 can actually be deployed will be a lot easier!
So to summarize:
* Tell RedHat to update and stop drawing the rest of the community down with its antiquated version of python (2.6) since the default version of python used in redhat is also the default version of python used in AWS, most VM hosting companies, etc due to its centos and fedora connection, and that sets the minimal bar that is currently way too low.
* Get the other linux distros to keep updating (Ubuntu seems to have this handled, props to them!)
* Find and port the major module blockers for python3+ use in a targeted, en-mass porting spree.
* Get status in a month and keep momentum going.