[02:32:14] <tos9> I find the current behavior awful as well, but it'd seem from that revert like it'd have to be behind a switch (and personally I'd assume one that just tells pip not to do that entirely, though maybe I don't have it all thought through there)
[02:45:06] <omry> tos9: if anything - let's restart the discussion.
[02:45:23] <omry> there are still users complaining about this 5 years after the original bug was reported.
[02:45:53] <omry> for me the difference is 45 times faster with this
[02:46:26] <omry> do you see any reasoning about the revert?
[17:02:18] <di_codes1> Has anyone ever seen a coverage.py failure like this? Only on pypy3, and with negative line numbers? <https://travis-ci.org/pypa/packaging/jobs/561272633>
[19:54:11] <omry> pip installing using pip+ssh from a github sub directory functional?
[19:59:18] <omry> trying to install using #subdirectory does not seem to work: http://paste.ubuntu.com/p/FgWzZcXYMZ/
[20:02:57] <omry> scratch that, my bad. bath subdirectory.
[20:51:19] <sumanah> di_codes: https://github.com/pypa/warehouse/pull/6207 is ready for rereview (better WebAuthn errors)
[20:51:39] <di_codes> sumanah: already approved that PR
[20:53:09] <sumanah> di_codes: you ok with merging it as is despite changes since your approval?
[21:00:34] <sumanah> woodruffw: that CI failure looks like a blip, to my intuition - now that it's out of date with master I suggest you re-push and see whether it happens again?