Spec issues with consensus' new 'package' field
Hi Nick. Brought these up on irc but moving to a ticket. Our consensus' new unused packages field has some spec issues...
- It isn't very future proof. It's fine for now, but I'm not sure how it could validly have new attributes in the future.
- Presently it necessitates at least one DIGEST entry. Think that's probably a mistake.
- 'one or more non-=, non-" " characters' => That's too imprecise. For instance, it allows newlines which I'm sure wouldn't really be valid.
- It would be nice if it said if a DIGESTTYPE can appear multiple times. I'd like to model this as a hash but I'm not sure if DIGESTTYPE are unique.