On Thu, Jan 2, 2014 at 11:12 AM, Vinay Sajip <vinay_sa...@yahoo.co.uk>wrote:
> > -------------------------------------------- > On Thu, 2/1/14, Brett Cannon <br...@python.org> wrote: > > >> "Progamming Language :: Python :: Python2and3" > >> or some such. > > > How is that any better than specifying the Python > > 2 classifier and the Python 3 one? One should specify every > > individual version of Python that a project is supposed to > > be compatible with anyway so I don't see the benefit of > > knowing through a classifier if 2to3 is needed or > > not. > > At the moment, if a classifier indicates that a distribution is both 2 and > 3 compatible, there is no declarative way to indicate whether or not 2to3 > is to be run - only via a setuptools.setup argument. I'm not hung up on > whether a classifier is used, but I think there should be *some* > declarative mechanism. A classifier has the advantage that it could be > applied to a specific release in the index without changing its code. > I guess my question then is "why do you care?" If 2to3 is run at install time then it's a cost, but it's one-time and if you really care you can always create your own wheel of the translated code or something. I guess I just don't view the overhead of some packages using 2to3 to be enough to warrant a classifier that only some people will use (it's hard enough to try and get people to use proper classifiers for what version of Python they support as it is).
_______________________________________________ Distutils-SIG maillist - Distutils-SIG@python.org https://mail.python.org/mailman/listinfo/distutils-sig
