docs: Start release notes for 5.0. #2219
Conversation
PR tornadoweb#1984 was based on the mistaken belief that we were already doing this (and in python 3.4+, it's true, thanks to PEP 446). This fixes a regression introduced in Tornado 4.5 in which autoreload would leak file descriptors and leave client connections hanging. Fixes tornadoweb#2057
Cherrypick FD_CLOEXEC change for 4.5.2
Update versionadded/versionchanged tags. Misc other doc updates
|
Given Python 2 support is on its way out, would suggest adding Tornado to python3statement. Admittedly this has mostly focused on the Scientific Python stack. Though there is no reason it should be limited to those Python packages. Not to mention many Scientific Python libraries make use of Tornado. So this would be good info to have there. |
|
Thanks for the suggestion; I've added a comment to python3statement/python3statement.github.io#21. |
|
Since I see that @bdarnell has left an open "TODO" for the Q of dropping Python 2 earlier rather than later, my recommendation is to go for it. I manage a much smaller but still substantial project (music21 -- about 700 people on the mailing list) and while I expected some pushback, there was nary a squeak from users when the first Py3 only versions started coming out. One grumble received personally then an hour later an email, "Oh! Feature X is 10x faster in the new version. Okay, I'll switch to Py3" Feature X was something I had never bothered to optimize because it was so hard to make work the same in Py2 and Py3; once I dropped Py2, that feature, and indeed the whole project, became a joy to work on again. I can't imagine that anyone working extensively with async code wouldn't experience that joy ten-fold. |
Update versionadded/versionchanged tags. Misc other doc updates