[Python-Dev] [Python-3000] RELEASED Python 3.0 final

[Brett Cannon]

On Thu, Dec 4, 2008 at 13:07, Nick Coghlan <ncoghlan at gmail.com> wrote:
> Terry Reedy wrote:
>> and this could give some people a mis-impression, most likely negative,
>> as to the magnitude and nature of the change.  Most of the code I am now
>> writing would, I believe, run with 2.5 except for print(..., file=xxx).
>>  And I know that there was concern for backward compatibility to the
>> point that some changes were rejected (renaming builtins) or delayed
>> (deleting duplicate test asserts) for that reason.  So I would soften
>> the statements to "... version of the language that is partially
>> incompatible with... " and "were made without being bound by backward
>> compatibility,"
>
> I would agree with Terry - while there are backwards incompatibilities,
> they aren't gratuitous.
>
> Then again, Guido does seem to want to discourage people from trying to
> target the common subset of the two languages instead of using 2to3 as a
> compilation step from the python3 version.
>

It makes sense if your code would have required jumping through hoops
to keep the base use-case. But if the only major difference is
something easily covered by a __future__ statement (think
print_function or unicode_literals, I believe although that __future__
statement is not documented anywhere according to Google), then I
honestly think it's okay to try to target the subset.

-Brett