combinePubKeys should not allow for n_pubkeys==1

[ysangkok]

There is this line in combine_pubkeys:

ARG_CHECK(n_pubkeys > 0);

Given that it doesn't make sense for n_pubkeys to be 1, I was wondering if this check should be changed to be "> 1"?

[gmaxwell]

why wouldn't it? n_pubkeys = 1 is well defined and has an obvious sensible behaviour. If the interface forces more than one then a caller needs to handle an additional corner case that would otherwise just work.

[real-or-random]

Hm. You can also combine the same pubkey n times with itself. I don't have a use case for this either but it's not broken or wrong in any sense, so I tend not to put restrictions here. But you're right, it's somehow confusing. I'm indifferent.

[ysangkok]

Ok, if the current behavior is to be kept, it would be nice to document it. How about a footnote in musig.md that says that the API supports a single signer? The footnote could be referenced each time a plural like "participants" or "public keys" is used.

[apoelstra]

I'd rather not complicate the exposition for the sake of this usecase. The current text makes sense for n=1 (aside from being ungrammatical) and it never occurred to me that n=1 might not work. Though perhaps I am just too close to this.

[sipa]

I don't understand what's the issue here. Does the function not do exactly what you'd expect with 1 input?

Look at Python's sum function (https://docs.python.org/3/library/functions.html#sum). It also does not list a special case for only one input.