When should Python 3.3 become the default?

[Barry Warsaw]

Since we're all raring to go now, this seems like a good time to ask for
feedback on some Python 3 decisions for 13.04.  Yes, we have a UDS-R session
on this[1] but I expect the archive will open before UDS, so it would be
better to have some direction earlier if possible.  I've chatted briefly with
Matthias and Scott K, but I don't want to accidentally misrepresent their
opinions.  Instead, I'll express my own and let them follow up.  Please let us
know what you think.

Some quick background: Python 3.2 is the default in 12.10.  Python 3.3 is
available but it is not a supported version, which effectively means that
pure-Python packages installed for 3.2 should be importable[2] in 3.3, but no
extension modules are built.  So for example, `import gi.repository` will fail
in 3.3 on 12.10.

Continuing on discussions we've had at previous UDSes, and looking ahead to
the probably schedule for Python 3.4, it is almost certain that we'll carry
Python 3.3 as the default Python 3 for the next LTS (14.04).  Whether we'll
also have Python 3.2 or 3.4 in 14.04 LTS is an open question we don't need to
decide right now.

Clearly we're going to enable Python 3.3 as a supported Python 3 version when
the 'raring' archive opens up.  The question is whether 3.3 should be the
default Python 3 as soon as the archive opens, or whether we should keep 3.2
as the default, enable 3.3, and watch for any fallout before making the switch
later in the 13.04 cycle.

There are good arguments on both sides, and I can probably convince myself to
be either aggressive or conservative.  One of my biggest concerns against
switching early is that we still have a lot of code that needs to be ported to
Python 3 (any version)[3].  If fixing problems related to Python 3.3
significantly detracts from the porting effort, then it might be better to
wait.  Then the question becomes, wait until what?

I am slightly leaning toward switching early, so that we can begin to take
advantage of some of features[4] of Python 3.3 that will benefit Ubuntu (and
Debian of course).  Two big ones are PEP 420 (namespace packages) and PEP 393
(flexible string representations).  Others that could prove quite useful
include: PEP 414 (reinstatement of explicit Unicode literals), PEP 405
(built-in virtual environments), PEP 417 (mock in the standard library), and
PEP 3151 (Reworking the OS and IO exceptions).

What do you think?

Cheers,
-Barry

[1] https://blueprints.launchpad.net/ubuntu/+spec/foundations-r-python33
[2] which is different than 'works' since of course the package itself must be
    compatible with Python 3.3
[3] http://tinyurl.com/8rvx67m
[4] http://python.org/dev/peps/pep-0398/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: not available
URL: <https://lists.ubuntu.com/archives/ubuntu-devel/attachments/20121018/4e4a9908/attachment.pgp>