doc build failure mystifies me cause builds locally

And it's an important doc build since it's the style example and unicode_minus keyword :/

Note to self:

• add a figures equal test that plt.xkcd and with style(xkcd) produce the same figure.
• changename to__name__ orpatheffect or use list[(name, kwarg dict), (name, kwarg dict)]

A concern with using{'name': 'func_name, **func_kwargs} as the structure there is a collision between keys that the API here controls ('name') and the keys that belong to the function that is going to be selected. While it is true none of the functions currently inpatheffects have a keyword named 'name', but it seems unwise to leave that resolvable conflict / special case on the table. Picking another name might reduce the risk of a collision, but usingTuple[str, dict] seems like the safest way to encode this and drops the need for a bunch of text + magic key in the input.

https://www.youtube.com/watch?v=iKzOBWOHGFE has a good section on inband vs out-of-band encoding of information.

but using Tuple[str, dict] seems like the safest way to encode this and drops the need for a bunch of text + magic key in the input.

Technically with this format I can do Tuple[str, list, dict] for(name, args, kwargs) but I dunno if that saves anyone work.

Technically with this format I can do Tuple[str, list, dict] for(name, args, kwargs) but I dunno if that saves anyone work.

If you wantTuple[str, list, dict], it should likely beTuple[str, dict] | Tuple[str, list, dict], because I anticipate that the first case is the fare more common one, and empty args'withStroke', (), {'linewidth': 4} is somewhat annoying.

But I suggest to keep it simple and only support kwargs. One could always add the other option later.

It'd be really helpful if#26989 could get in b/c that would help me debug the unicode minus glyph not working error

builds usinghttps://github.com/dummy-index/xkcd-font/raw/brushup/xkcd-script/font/xkcd-script.ttf so do I need to verify that it's okay to use? liscence is cc and using this file b/c it has an extended character set

Also, we may want to ask if we can vendor that file since it's on an unmerged branch of an unmaintained fork if we're gonna use it in doc builds? ETA: or fork that fork or the original project and add that fork?

And do we wanna deprecateplt.xkcd if this goes in? shadowed all the existing xkcd tests so continuity is maintained.

Can you please put theHumor Sans work in its own PR so we can get that merged and backported ASAP?

#27299

follow up from call is to make a note in xkcd docs about plt.style as alt

notes:
fix doc build - is probably xkcd-font missing symbol
support third party using fully qualified name