Python 3.0 unfit for serious work?

[Paul Boddie]

Steven Bethard wrote:
>
> Well, Python 2.4 code will work on Python 2.6 and 2.7 so just because
> your code isn't yet compatible with Python 3.0 doesn't mean you should
> give up on Python.

Perhaps the most important concern in the context of Python 3.0 is
what the term "Python" will come to mean to the different communities
using the language. Some people will be maintaining software that
needs to be compatible with older releases; these people aren't
technologically backwards: they just find that such older releases
provide sufficient capabilities for the implementation of their
solutions. For such people, "Python" will be the language they've
always used, with a gradual accumulation of features, some known
quirks, and some relatively minor incompatibility issues over the
years.

Meanwhile, the risk is that the core developers will only consider
Python 3.0 as "Python" and that people doing anything with older
releases will be on their own. If the gap is too wide between 2.x and
3.x, any lack of maintenance in the 2.x series will be perceived as an
abandonment of "Python" for certain kinds of user, and the result will
probably be a loss of confidence in both variants of the language.
Although I believe that some of the recent attempts to lower the
disruptiveness of Python 3.0 may have helped to maintain "Python" as a
single narrow continuum, I think some people overlook how fortunate
the relationship between Python's evolution and momentum has been, and
how easily such a relationship can be broken.

Paul