[Python-Dev] [Python-checkins] r42116 - python/branches/release24-maint/Lib/unittest.py (original) (raw)
Georg Brandl g.brandl-nospam at gmx.net
Fri Jan 20 22🔞20 CET 2006
- Previous message: [Python-Dev] [Python-checkins] r42116 - python/branches/release24-maint/Lib/unittest.py
- Next message: [Python-Dev] [Python-checkins] r42116 - python/branches/release24-maint/Lib/unittest.py
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Barry Warsaw wrote:
On Fri, 2006-01-20 at 21:43 +0100, Thomas Wouters wrote:
I don't believe this belongs in 2.4, since it can, actually, break code. Code that depends on the current situation, TestCase_attributename. Fragile code, to be sure, but still. If there were a compelling reason to backport, I guess it could be hacked to work right-ish, but subclassing TestCase in this way, while convenient, isn't important enough to warrant this (IMHO). Exactly right.
You're right. Next time I'm going to decide in favor of not backporting. I reverted both checkins, this and the urlparse one.
For 2.5, this would be a good opportunity to add additional schemes to urlparse. Candidates?
Georg
- Previous message: [Python-Dev] [Python-checkins] r42116 - python/branches/release24-maint/Lib/unittest.py
- Next message: [Python-Dev] [Python-checkins] r42116 - python/branches/release24-maint/Lib/unittest.py
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]