I understand what this PR is trying to do but think it makes users worse off. For the most part, people understand the UTC offset sign conventions in terms of being east or west of the prime meridian. While this isn't 100% true, it is a good approximation and mental model.

It would be better to leave the East/West positive/negative references in place and just add a note somewhere in the docs explaining the exceptions to the rule.

Hmm, for the docs, what about instead adding an east/west note on the timedelta object, something like:

'if the time being compared is east of UTC, timedelta is positive. If west of UTC, then timedelta is negative'

and keeping the east/west removal under the offset function? This way, other functions that refer to timedelta (like dst(), which doesn't have anything on east/west, but do in the docstring) can all refer to one spot.

Regarding the docstrings, I'm pretty indifferent and would be fine with keeping east/west as-is if it reads better there.

I'm going to mark this as closed, but another core dev can reopen if they feel strongly about it.

AFAICT no reported user problem is being solved here. IMO it is a nitpick that while technically correct will make the docs less usable. People do mentally time offsets with east and west. That is a useful first approximation.

To misquote Kronecker, "god created the east/west timezone offsets, the rest of the funky timezone lines are the work of man." ;-)