Elliott Sales de Andrade

Results 335 comments of Elliott Sales de Andrade

Only failure is on `push`, which I can't restart, but the `pull_request` build did pass.

I do not understand all the extra work that needs to be done here. On Cygwin x86_64, I installed Python 3.9, `gcc-c++`, `make`, and I think `python39-matplotlib`, and was able...

Why are we building wxPython now; there should be wheels?

But there _are_ wheels for wxPython? https://pypi.org/project/wxPython/#files

Yes, theoretically it should be possible, but might need to do a bit of low-level surgery to expose it.

I dunno, "Better error messages" sounds too vague to me, while "Fix some typos" is vague enough to not need specifics...

This doesn't now seem to fix the original issue? I don't see minor ticks with ec8ae4b as you do with `cbar.ax.set_yscale('log')`.

The original issue looks like this on `main`: ![main](https://user-images.githubusercontent.com/302469/181094194-ff6f46e2-dc23-4e0f-9d4d-3a2c1fa8f38d.png) and like this on the final commit of this PR: ![pr-final](https://user-images.githubusercontent.com/302469/181094257-59b8e1ff-7414-4cbd-a879-4281e475ba5d.png) Whereas, the second-last commit looks like this: ![pr-second-last](https://user-images.githubusercontent.com/302469/181094322-eccc4766-695c-4f25-8bc3-b7891713be9b.png) So with...

In Cartopy, we removed `.git_archival.txt` entirely: SciTools/cartopy#2057; I wonder if either approach is broken somehow. @dopplershift @greglucas I think wheels were always okay; the problem is in #23325 where installs...