Forking baazar to add Python 3.x support

[Richard Wilbur]

On Tue, Mar 11, 2014 at 10:49 PM, Ben Finney <ben+bazaar at benfinney.id.au> wrote:
> Richard Wilbur <richard.wilbur at gmail.com> writes:
>> In other words,
>> the revised module would require Python 2.7 or 3.3+ in order to run.
>> Will that cause anybody problems?
>
> This is a good question to ask. But I think it will need to be answered
> by a representative sample of the Bazaar user base, most of whom are
> *not* reading this forum.
>
> As I understand it (I don't have any hard evidence for this), Bazaar's
> user base has many members in private projects within corporations.
> Those private projects have no public face, and a survey of the internet
> won't find them.
>
> It would be a big win to get feedback from those users on the future
> development of Bazaar. How will we find and engage them?
>

I've been thinking about this for awhile.  It seems like a pretty
thorny problem.  The best solution I have thought of so far is to
continue back-porting fixes to at least 2.6.x (maybe 2.5.x and/or
2.4.x, if needed) and making bug fix releases from time to time.  This
way people will have a more useful, stable platform while we work on
upgrading and porting the code.

This would keep the bzr 2.6.x releases requiring Python 2.6 while the
new development on bzr 2.7+ would require Python 2.7/3.3+.

> Another important benefit of breaking those components out is that they
> could be improved separately, and migration to Python 3 would likely be
> more tractable (easier to test, easier for developers to understand) for
> each since the separate projects would be smaller.

Good points.  I'm certainly in favour of making the porting process,
testing, and developer understanding easier.