Currently there're no

* `void vprint_unicode_locking(string_view fmt, format_args args);` and

* `void vprint_nonunicode_locking(string_view fmt, format_args args);`

overloads (where stdout should be the omitted FILE* argument).

Should there be an LWG issue for this? @jwakely @Dani-Hub

I noticed the omission in P3107R5 as well, but it looked deliberate to me. @jwakely What do you think?

We don't want convenience functions for using those with stdout. They're low level, internal implementation details. They don't need to be convenient.

I notice that the paper P3107R5 did not show a header where to add those _locking functions.

@jwakely , is the change proposed here (not the add-on stdout overloads) editorially-obvious?

Yes, I think the new functions should be in the synopsis.

@jwakely , care to approve, then?

@tkoeppe @jwakely This PR is outdated due to P3235R3. Could you review the new PR?