Author ned.deily
Recipients anthony-flury, bbayles, berker.peksag, eric.araujo, ezio.melotti, michael.foord, ned.deily, rbcollins
Date 2018-09-14.15:07:21
SpamBayes Score -1.0
Marked as misclassified Yes
Message-id <1536937641.31.0.956365154283.issue32933@psf.upfronthosting.co.za>
In-reply-to
Content
The potential change in behavior affecting existing code was one issue.  But the other was the fact that people writing tests to make use of the new behavior can't depend on that behavior being there for 3.7 or 3.6 without checking the patch level, for example, 3.6.7 vs 3.6.6.  That's one of the main reasons we generally do not backport behavior changes unless they are a clear bug; as I noted. this particular issue seems somewhere in between a bug and a feature.  Given how far along we are in the 3.6.x cycle, I think we definitely should not backport to 3.6.  Since 3.7 is near the beginning of its support cycle, I would not object if we did backport this for 3.7.1.  I'll leave it up to Berker.
History
Date User Action Args
2018-09-14 15:07:21ned.deilysetrecipients: + ned.deily, rbcollins, ezio.melotti, eric.araujo, michael.foord, berker.peksag, anthony-flury, bbayles
2018-09-14 15:07:21ned.deilysetmessageid: <1536937641.31.0.956365154283.issue32933@psf.upfronthosting.co.za>
2018-09-14 15:07:21ned.deilylinkissue32933 messages
2018-09-14 15:07:21ned.deilycreate