Albert Y. Shih

For SunPy, we have switched our RTD build back to conda to resolve this for our docs (sunpy/sunpy#6261)

I've put up a PR (astropy/sphinx-automodapi#152) that finally fixes this bug

Upon further testing, I discovered that the true bug is in Sphinx after all, and has nothing to do with sphinx_automodapi. See sphinx-doc/sphinx#10576

I'll point out that even after Sphinx accepts the bugfix for current releases, the Astropy docs are currently being built against a rather old version of Sphinx. You'll almost certainly...

Yeah, I'd just define a subclass of `BaseCoordinateFrame`. In fact, I'll point out that `BaseEclipticFrame` and `BaseRADecFrame` are two examples of transform-free frames already defined within Astropy.

My stance hasn't changed, so I'll simply reiterate that I want the justification for the decision to be rooted in aspiration (running towards a "we are one" philosophy) rather than...

There's been back and forth on the Sphinx PR (sphinx-doc/sphinx#10614), but it hasn't been merged yet.

1. That's really a feature request rather than a bug. See #7186. 2. We aren't particularly interested in fixing issues with `solar_rotate_coordinate()`. Can you use `propagate_with_solar_surface()` instead? ```python >>> with...

> for 1. great - missed that PR. Heh, you didn't miss anything. I created that PR in response to this issue. > for 2. absolutely, so is the issue...

I'll just point out that #12317, followed by #13497, is the reason for the finiteness check.