Here would be my plan (though I'm not good enough to carry it out):
1. Port the dispatcher to P3.
2. Do something magic with urls.py, so you choose which version (2.X or 3.X) you want doing the query.
At this point, people can use URL plumbing to gradually port their Django projects.
3. Port the parts where Python 3 is a killer feature. String formatting, Unicode handling (maybe), function decorators, and metaclasses. This is the hard part - where would python 3 add some killer features to django?
You also need to do the db layer, the templating, and all that jazz, but that should follow once you have the incentive of actually seeing it work in your browser.
The holdup here is not technical. It's great that we've got people like Alex willing to port code, but that's not the hard part. The hard part is people who are using and want to continue using Django on platforms where Python 2.5 or even 2.4 is still the standard.
More on this forthcoming, once I've had a proper weekend off.
I started working on an app using Python 2.7. The only reason I chose 2.7 over 3.x was the lack of 3.x support in Django. I would be great to have Django support Python 3.x.
Its nice to see posts about python and django. Ive only played around with them a little bit but having options is always a good thing and right now it seems the frameworks discussion is really dominated by Rails.
I think he's aware of the formal procedures from his previous involvement with GSOC and Django. The post seems more to be stating what's what, how it's probably going to happen, etc.
This would be Alex's third time at GSoC after MultiDB support and support for non-relational databases, but hugely successful.
Plus Google decides GSoC students on a per project basis and the project's core developers get the pick the students; and Alex, already is a core developer.
[+] [-] wisty|15 years ago|reply
1. Port the dispatcher to P3.
2. Do something magic with urls.py, so you choose which version (2.X or 3.X) you want doing the query.
At this point, people can use URL plumbing to gradually port their Django projects.
3. Port the parts where Python 3 is a killer feature. String formatting, Unicode handling (maybe), function decorators, and metaclasses. This is the hard part - where would python 3 add some killer features to django?
You also need to do the db layer, the templating, and all that jazz, but that should follow once you have the incentive of actually seeing it work in your browser.
[+] [-] ubernostrum|15 years ago|reply
The holdup here is not technical. It's great that we've got people like Alex willing to port code, but that's not the hard part. The hard part is people who are using and want to continue using Django on platforms where Python 2.5 or even 2.4 is still the standard.
More on this forthcoming, once I've had a proper weekend off.
[+] [-] prog|15 years ago|reply
[+] [-] sigzero|15 years ago|reply
[+] [-] clojurerocks|15 years ago|reply
[+] [-] danssig|15 years ago|reply
[+] [-] briancurtin|15 years ago|reply
[+] [-] scorpion032|15 years ago|reply
Plus Google decides GSoC students on a per project basis and the project's core developers get the pick the students; and Alex, already is a core developer.
[+] [-] vipivip|15 years ago|reply