How ahead are you guys in the (Python) real world?

[Neil Schemenauer]

My company uses 2.2 and 2.3; we hope to drop 2.2 Real Soon Now.

This has been an interesting thread. There has been some discussion
on python-dev about doing another 2.3 bugfix release. Based on the
number of people still using 2.3, it looks to me like there would be
interest.

Neil

 

[Aahz]

This has been an interesting thread. There has been some discussion
on python-dev about doing another 2.3 bugfix release. Based on the
number of people still using 2.3, it looks to me like there would be
interest.

Yes; the real question is whether there is enough labor available to
make it happen.

 

[Jack Diederich]

Yes; the real question is whether there is enough labor available to
make it happen.

I came away from the thread with the opposite conclusion for similar
reasons. People would use a 2.3.6 if their OS upgraded it for them but
those are the same people who won't upgrade to 2.4.x because it involves
testing. 2.3.5 isn't broken for them or they would know it by now.
2.3.6 probably isn't broken for them but it can't help -- or they would
have noticed a bug by now.

My own servers jumped from 2.2 to 2.4 for the same reason everyone else
has given for an upgrade. We just happened to have time in the dev
cycle to do testing and did it on the version that was available (2.4.1)
at that time.

-Jack

 

[stdazi]

Since I haven't used Python at work, I am using Python 2.5 right now.
However I wonder, how fast are you guys moving from version to version
at work? As an illustration my ex-company just moved to Java 5, which
was released around... what, 2-3 years ago? (While I am running Java 6
at home)

Is it the same in the Python world? What version of Python is used in,
say, Google? Is it even 2.4 yet?

Heya, here at my job 2.3 is a must , but the code *should* be as
compatible as possible with 2.2 too.

 

[skip]

Jack> I came away from the thread with the opposite conclusion for
Jack> similar reasons. People would use a 2.3.6 if their OS upgraded it
Jack> for them but those are the same people who won't upgrade to 2.4.x
Jack> because it involves testing. 2.3.5 isn't broken for them or they
Jack> would know it by now. 2.3.6 probably isn't broken for them but it
Jack> can't help -- or they would have noticed a bug by now.

In contrast, I generally do only a small amount of testing when a micro
release of Python comes out before inflicting it on our users and
developers, precisely because I have confidence that it only contains bug
fixes. Since probably 2.2.1 there have been no new features in any micro
releases that I can recall and, up to this point at least, no regressions in
my experience. Micro updates seem like a pretty safe bet to me.

Skip

 

[skip]

(ack! forgot to push the send button several days ago - hopefully this isn't
woefully out-of-date now...)

neil> Based on the number of people still using 2.3, it looks to me like
neil> there would be interest.

aahz> Yes; the real question is whether there is enough labor available
aahz> to make it happen.

Just to make it clear for people with an interest in another 2.3 release who
might not read python-dev, here's what needs to happen. Someone(s) needs to
consider which bug fixes applied to 2.4 (and maybe CVS head) since the last
2.3 release should be applied to the 2.3 branch. That entails a few steps:

1. scan the 2.4 checkins for bug fixes (scanning Misc/NEWS might be
sufficient)

2. decide which ones apply to the 2.3 branch (some might not if they
were bugs that were introduced in 2.4 or because code structure
changes would make it too painful to add to 2.3)

3. work up a patch with test cases against the 2.3 code (simplified by
starting with the 2.4 patch in most cases) that implements the fix

4. test the 2.3 branch as patches are applied to make sure there are no
regressions against your own code base.

All these steps are necessary to some degree, though one person doesn't have
to do all four steps for any given bug fix. Any help you can give would be
appreciated.

Skip