[Distutils] status check on PEP 517

[Nathaniel Smith]

So one consequence of this is that every time a new release of flit (for
example) adds a new dependency on X, or one of flit's dependencies adds a
new dependency on X, then this is technically a backwards incompatible
change, because any existing packages that use flit and happen to have a
top level directory named X will stop working.

I guess the most obvious example of when this would occur is: suppose click
switches to using flit for builds, and then flit switches to using click
for command line parsing. Now there's a bit of a chicken and egg problem
where 'pip install click' will end up importing flit with the click source
tree on the path, and this tree of course contains a directory named
'click', so unless special measures are taken flit will end up importing
the code it's trying to build.

But of course this can happen for lots of reasons; most packages have names
that you wouldn't expect to randomly encounter at the root of a source tree
very often, but with 100,000+ packages on pypi I expect it will happen
eventually.

This doesn't happen with setuptools because setuptools traditionally has
few or no dependencies, but obviously we're changing that; the whole idea
here is that now your build system has full access to pypi.

Anyway, when this happens, what's the recommended mitigation? Should flit
and other backends take special measures to delay imports until after their
hooks are called and they have a chance to play games with sys.path? Should
we put the source tree at the end of sys.path instead of the beginning, so
that site-packages masks the source tree instead of vice-versa? That would
be different from what people are used to from setup.py and other scripts,
and seems like it might cause its own problems, but maybe it's better?

Or am I worrying about a non-issue and it's fine if flit imports click from
the source tree?

On Jul 29, 2017 2:31 AM, "Thomas Kluyver" <thomas at kluyver.me.uk> wrote:

> On Sat, Jul 29, 2017, at 09:47 AM, Paul Moore wrote:
> > +1 from me on just adding the project root to sys.path.
>
> Ditto. I don't expect most projects to use it, but it seems like a handy
> bit of flexibility to have available. The potential clutter is a bit
> ugly, but I don't think it's worth the trouble of specifying an extra
> key to avoid that.
>
> Thomas
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/distutils-sig/attachments/20170729/ba8e2624/attachment.html>