When and why are you planning to upgrade to Python 3.x?

Question

Python 3.x (aka Python 3000, Py3k, etc) is now available. When and why are you planning on porting your project or code to the new Python?

edit: I'm particularly interested in any features that don't exist in 2.6 that make porting worth it. Right now seems like a lot of negatives (x hasn't been ported yet), but I don't know what people see as the positives. Regarding "when", I'm interested in people's thoughts that the first step to porting is to have "excellent test coverage" which seems a bit optimistic for some projects.

Answer 1

The when and why are the same for me. When there is a feature in 3.0 that I need, which does not have an equivalent in 2.6. The problem with being on the bleeding edge, is that its bloody.

Edit:

After reading your edit, and http://docs.python.org/3.0/whatsnew/3.0.html (Thanks Brian R. Bondy), I see no compelling features in 3.0 which would make me port from 2.6. Of course as it matures and becomes 3.1, 3.2, etc. I am sure some feature will arise that I feel I must have.

Having said that the only negative I see in 3.0 is the non-truncating behavior of 4/3. I don't think it is bad in and of itself, but I do think it will take me a while to remember to use 4//3.

Answer 2

When MYSQLdb has been ported. :)

Answer 3

already have

Answer 4

Apres le deluge... Or perhaps whenever my system administrator installs it. I would like to use it because change is good, and by good I mean inevitable.

Answer 5

My roguelike game is written to 2.5 and I'll probably keep it there for the foreseeable future for the sake of the four or five people who play it :P

For my personal stuff, I plan to switch 3.0 posthaste since it's the wave of the future.

I have some scripts at work that run on IronPython so whenever that project is 3.0 compatible I'll probably start moving them over.

Answer 6

I'm waiting at least one calendar year from the initial release of Py3k just to make sure its stable enough for production use... been burned to many times jumping to the next generational release. As it stands I don't have a need for anything in Py3k yet for my projects.

Other things holding me off: All the libraries I depend on to transition first.

Answer 7

Not anytime soon. Right now some modules aren't even ported to 2.6. Py3K rules, but language without libraries is useless.

Answer 8

When Setuptools have been ported - how else am I going to deploy my stuff!

Answer 9

The big "why" is sane Unicode support. 90% of the python I write is one-off scripts, so there is no "when" for me. I use them both.

Answer 10

I'm currently still learning Python using 2.6 so probably not for a while. Once I feel comfortable using 2.6 without having to resort to asking for help as frequently I will probably try to understand the changes used in 3.0 and see how I do with that. So:

When: Once I am able to use 2.6 proficiently with minimal help and a better understanding. Why: Solely because understanding how to use 3.0 as well as 2.6 will help if I were to look for a job and it requires knowledge of one but not the other.

Answer 11

At least not before Python 2.6 is available on the Linux distribution I am using. After that I need to think how to proceed.

Having also 3.0 available from the Linux distribution would help.

Answer 12

When and why are you planning on porting your project or code to the new Python?

When all of its dependencies will be ported to 3.0 and will be stable.

Answer 13

when Django does

Answer 14

When SQLAlchemy has been ported and has had some time to bake. The feature I'm most looking forward to is the change so that strings are all Unicode.

Answer 15

Once numpy and scientific python have been ported....but that Looks like it could be a long time coming.

Answer 16

I like Python 3.0's features a lot, but I simply can't migrate until, at the very least, Twisted, SQLAlchemy, PyOpenSSL, Setuptools and PyCrypto are stable on 3.0. Several of 3.0's features have been back-ported to 2.6, making the transition less urgent.

Answer 17

As with everyone else, I am waiting for my favourite modules to upgrade before I do. In my case, pygame. Until then, I am sticking with 2.5.

Answer 18

When <insert c library binding here> gets ported. In my case PyGTK.

Answer 19

PyQt4 is a must for me.

Answer 20

when everyone else is moved

Answer 21

I really like the Unicode support, and that alone is enough reason to move. When to move?..is a tough question. As an application developer it is not feasible for me to move till the frameworks (Django, web.py), and 3rd party packages I use have been ported and tested.

I plan to evaluate migration costs when these packages have moved and are stable. Till then I am sleeping over this question. :-)

Answer 22

I have upgraded.

I like the new features in Python 3.

For example 1/3 will return 0.3333333, a float. I also like the fact that no more long, but only int exists. Last but not least, map() and filter() return iterators instead of lists. Isn't working with iterator better than list?

By the way, my programs do not depend on any third party modules, except the one that comes with Python.

Answer 23

I think I'll move when numpy, matplotlib, and wxPython get 3.0 support.

Answer 24

When the default install of most GNU/Linux distros does.