bug#51342: 29.0.50; remove non-CAPs from rcirc capability list

[J.P.]

Philip Kaludercic <philipk@posteo.net> writes:

>> Standard replies are quite mysterious. From what I can gather:
>
> [...]
>
> I understand the issue, but am still hesitant. If this is vague, then
> it seems better to err on the side of safety

What's mysterious and vague isn't the existence of a capability called
standard-replies. No such capability currently exists. As mentioned in
my last reply (not sure if you saw that one), the closest thing is
inspircd.org/standard-replies. Hope that makes sense.

> Or are there any real downsides to being more explicit?

No downsides at present because you request one cap per line. And you
have no interdependent caps as yet. So long as both remain true, there's
nothing to worry about. And rcirc doesn't make you accrue flood debt,
so early messages (even spurious ones) don't cost extra.

In ERC's case, we *do* have to worry because we implement 302 and have
multiple dependencies. If any one gets NAK'd, there goes the ball game.
There's also some undefined behavior [1] that can turn connection
registration into a bit of a limbo without additional planning (should
you ever decide to go that route). Thanks.


[1] 
https://github.com/ircv3/ircv3-specifications/pull/400#issuecomment-579063998