Re: Ordered 'opqque' validators

> Note that neither your solution nor my proposed heuristics require any
> protocol elements (though they may require language describing this use
> of the Date header), and that either solution could be implemented and
> be consistent with the protocol -- in other words, all we have to
> convince ourselves of is that no new protocol is required, and leave
> specific optimizations for others to worry about.

Yep, though what I'd really like is the words to put in the spec so
that we can transfer our group's accumulated understanding about
all these niggling details into something an implementor can work from.

......Roy

Received on Wednesday, 7 February 1996 00:42:29 UTC