from Demo/scripts/newslist.py :

# Feel free to copy, distribute and modify this code for
# non-commercial use. If you make any useful modifications, let me
# know!
#
# (c) Quentin Stafford-Fraser 1994
# fraser@europarc.xerox.com                     qs101@cl.cam.ac.uk


this is not very nice for commercial distros - we have to drop the file (for now anyway)

i'm not sure what the "right" solution for upstream is - you might want to remove the file, ask Quentin to relicense it, or just acknowledge its existence and keep it

It’s a problem for free software distribution, regardless of the commercial status, like the profiler used to be.  The Demo directory is gone in 3.2 and higher; I’m not sure any effort to fix this in 2.7 is worth it.

The only resonable fix to make would be removal from the distribution.

I don’t see any problem with removing it from the 2.7 branch.

The windows installer does not install the demo directory.  *nix distributions do what they want.  I think it too late to change the 2.7 repository.

thanks for bringing this up. I don't agree that the non-availability in the windows build should be used to close the issue.  Linux distros distribute the source, and are still affected? If it's already removed in Python3, why not remove it here as well?

Anyway, I can repackage the sources for Debian/Ubuntu

Given that it's only the demo directory (and a relatively obscure demo as well these days), I'm inclined to just fix it upstream rather than making the distros patch it.

There's also the fact that we distribute the source tarballs from python.org as well, and assert that those are under the PSF license, which this script doesn't adhere to.

If we are going to backport the removal from the distribution of Demo/scripts/newslist.py to 2.7, it might as well be done immediately, before 2.7.9 goes out.  Benjamin?

I am more concerned about separately licensed stdlib modules, such as turtle* (and others).  Do they all properly 'fall under the PSF license'?  Has this been systematically checked (and recorded)?  Is there a policy in place with respect to new modules?

* G.Lingl said in email he would sign he contributor agreement in August, but none is recorded yet.

New changeset 54af09408795 by Benjamin Peterson in branch '2.7':
nuke ancient script with dubious license (closes #12987)
https://hg.python.org/cpython/rev/54af09408795