Op 24-08-16 om 09:06 schreef Nick Coghlan:
I think in this case, campaigning to remove ".zip" support will prove
to be a false economy, as anything you'd gain from simplifying the
toolchain implementation you'll lose arguing with folks that would
prefer for .zip to be the default instead of .tar.gz.
While it might sound very strange: at one time (2008) it helped us that
**both** .zip and .tar.gz were allowed
There was a bug in python 2.4's build-in tarfile module that failed to
extract files with a path that was exactly 100 characters long. 99 fine,
101 fine, 100 would fail to be extracted.
See http://maurits.vanrees.org/weblog/archive/2008/09/scrambled-eggs
In order to support broken 2.4 pythons, zest.releaser explicitly created
.zips for a long time :-) If I look at the changelog, the issue still
cropped up from time to time in 2012, four years later.
We stopped forcing .zip in version 3.56 (2015-03-18).
So...
a) It came in handy to have both .zip and .tar.gz once :-)
b) zest.releaser might have **massively inflated the number** of zip
files in the statistics. zest.releaser is used a lot in the zope/plone
corner of pypi and that's a very active corner of pypi. Donald, perhaps
you could look at the top 10 of zip packages to see if it has a plone smell?
Reinout
--
Reinout van Rees http://reinout.vanrees.org/
rein...@vanrees.org http://www.nelen-schuurmans.nl/
"Learning history by destroying artifacts is a time-honored atrocity"
_______________________________________________
Distutils-SIG maillist - Distutils-SIG@python.org
https://mail.python.org/mailman/listinfo/distutils-sig
