Author r.david.murray
Recipients Nashev, Todd.Rovito, asvetlov, georg.brandl, gpolo, kbk, michael.foord, ned.deily, python-dev, r.david.murray, rhettinger, roger.serwy, taleinat, terry.reedy
Date 2012-11-02.12:00:20
SpamBayes Score -1.0
Marked as misclassified Yes
Message-id <>
The reason for our "no new features" policy is that if a program works with version x.y, it should work for all x versions (modulo failing on an earlier version because of a bug...and conversely if it works on x.y, it should work on all later versions of x, which is why we also sometimes don't fix certain bugs in maintenance releases).

In this case, if I understand correctly, there are no *programs* that can depend on the feature, just humans.  So I think putting this in to bugfix releases is more analogous to the fixes-that-look-like-enhancements we occasionally put into the build infrastructure for Python itself (as opposed to distutils, which is governed by the normal backward compatibility rules).

So, it's a judgment call and other developers might not agree with me, but I think it is OK.
Date User Action Args
2012-11-02 12:00:21r.david.murraysetrecipients: + r.david.murray, georg.brandl, rhettinger, terry.reedy, kbk, taleinat, Nashev, gpolo, ned.deily, roger.serwy, michael.foord, asvetlov, Todd.Rovito, python-dev
2012-11-02 12:00:20r.david.murraysetmessageid: <>
2012-11-02 12:00:20r.david.murraylinkissue1207589 messages
2012-11-02 12:00:20r.david.murraycreate