setuptools 0.6c10 delayed until Monday
After giving it some thought, I've decided to hold off on pushing out the 0.6c10 release until Monday morning, since it's likely that some people have scripts or build processes that run off of ez_setup SVN or the latest PyPI version of setuptools... and who aren't aware the release is coming and might be surprised by it. It would therefore probably not be a good idea to push this out on a Friday. Also, this will give those who are using Distribute and do *not* wish to use this release, a chance to upgrade to the latest version of Distribute first. (Which, if I understand correctly, includes code to prevent accidental re-installation of setuptools.)
On 2009-10-16, P.J. Eby <pje@telecommunity.com> wrote:
Also, this will give those who are using Distribute and do *not* wish to use this release, a chance to upgrade to the latest version of Distribute first. (Which, if I understand correctly, includes code to prevent accidental re-installation of setuptools.)
Correct. There's an "if" in the one or two places where a requirement is looked up. If the request is for "setuptools", distribute will reply with "here you go, here's distribute". So that ought to work. Related: I added the code that checks whether we're not requesting a 0.7-series setuptools. So if a 0.7a1 setuptools comes out and someone requires it, it will be installed: distribute won't fake a 0.7. And I added a similar check in the (from the top of my head) on_insert() method that raises an error if you try to install a 0.7 setuptools inside a distribute-handled working set (like you suggested, iirc). Reinout -- Reinout van Rees - reinout@vanrees.org - http://reinout.vanrees.org Software developer at http://www.thehealthagency.com "Military engineers build missiles. Civil engineers build targets"
At 02:03 AM 10/17/2009 +0000, Reinout van Rees wrote:
I added the code that checks whether we're not requesting a 0.7-series setuptools. So if a 0.7a1 setuptools comes out and someone requires it, it will be installed: distribute won't fake a 0.7. And I added a similar check in the (from the top of my head) on_insert() method that raises an error if you try to install a 0.7 setuptools inside a distribute-handled working set (like you suggested, iirc).
Yep. Thanks!
participants (2)
-
P.J. Eby -
Reinout van Rees