bug#37517: New rx.el breaks a number of packages - missing symbol rx-submatch-n

[Mattias Engdegård]

26 sep. 2019 kl. 16.28 skrev Eli Zaretskii <eliz@gnu.org>:
> 
>> Using internal functions doesn't make them APIs. All we have now is a single 
>> datum, which was for all I could see just a mistake. Let's wait for evidence 
>> about which, if any, internal symbols that have been used, and for what 
>> reasons.
> 
> We've heard about one such symbol, which you consider internal, but
> someone else didn't.  Why not provide that one symbol right now?  If
> no other symbol is reported, that will be the only one, but I don't
> see why it should wait for others to be granted that status.

Given that flycheck was rapidly fixed and the single internal function used was 
tied to their particular rx use (because the old mechanism for extending rx was 
under-documented, as alluded by Clément), do you still think we need to add 
rx-submatch-n for the benefit of old flycheck versions?