[cabfpub] Ballot 188 - Clarify use of term "CA" in Baseline Requirements

[Gervase Markham]

On 01/03/17 07:02, Dimitris Zacharopoulos via Public wrote:
> I can't imagine that Mozilla, Entrust, Globalsign, Digicert (that
> already voted "Yes") didn't read through the ballot and didn't consider
> these misunderstandings.

Your confidence in me is heartwarming :-), but sadly misplaced. I didn't
notice the issues which Ryan and Peter have raised, but having reread
this thread fairly carefully, I can see that they have a point.
Particularly about the OCSP/"good" thing, which seems like a serious bug.

I hope that you won't see it as a denigration of your hard work on this
ballot, but I'm afraid we have to change our vote.

Mozilla now votes NO.

I would like to make a concrete suggestion as to the way forward. It
seems to me like this very important task of fixing the BRs and other
documents to have consistent language falls into two parts:

a) making a sane and consistent set of definitions; and
b) making the document use them consistently.

Might it make sense to do a) as the full Forum, and get agreement on the
definition set, before re-attempting b)? And when b) is re-attempted, we
may find that it's impossible in some cases to express what the BRs
currently say using the sane and consistent set of definitions created
in a). This is likely to be a bug in the BRs. We can then decide on a
case-by-case basis whether to craft "custom language" to keep the bug
and fix it later, or have the ballot fix the bug as well as fixing the
language.

It may actually be that the work of a) turns into an RFC 7719-like
document for the WebPKI.

Gerv