Author jeff.allen
Recipients fwierzbicki, jeff.allen, kprussing, zyasoft
Date 2019-08-17.16:21:38
SpamBayes Score -1.0
Marked as misclassified Yes
Message-id <>
Ok, now I see why this matters.

We lost the code that used to check out (just) the license file from CPython when we dropped use of Subversion, but we still have the code that expects it to be there. 
What puzzles me is that we have managed several releases since and not noticed this.

The hg checkout is right, of course (except CPython is now on GitHub), but I'd quite like the build to work without depending on hg. If I've got the code once already, why clone again? Maybe I'm missing something.

A copy may be unnecessary (IANAL) but the easy solution is to put a version under source control and copy it to where this failing task is trying to put it. Seems like it's been expected there since at least 2.2. Our license refers to a Lib/LICENSE as if once upon a time we simply checked in (and copied) the CPython license along with the stdlib.
Date User Action Args
2019-08-17 16:21:38jeff.allensetmessageid: <>
2019-08-17 16:21:38jeff.allensetrecipients: + jeff.allen, fwierzbicki, zyasoft, kprussing
2019-08-17 16:21:38jeff.allenlinkissue2330 messages
2019-08-17 16:21:38jeff.allencreate