Re: [dnsext] WGLC: draft-ietf-dnsext-dnssec-bis-updates-16

[Samuel Weiler <[e-mail]>]

On Wed, 1 Feb 2012, Paul Hoffman wrote:

> 5.10 is long, scary, and useless for most environments because most 
> environments will have just one trust anchor.

It is long and scary.  Earlier discussion on list discarded several 
ways of helping with that.  David and I came up with a new one: we 
stuck most of the text in an appendix.  It's still long, but maybe it 
will now be less scary.  I think this change is purely editorial; if 
we need to back it out later, we can.  Please let us know if you like 
it.

> 5.6 (setting the DO bit in replies) suggests resolvers should "be 
> liberal in what they accept". That's a bit vague. Instead, say ... 
> "Because some implementations ignore this rule on sending, the rule 
> for receivers is now that they MUST NOT expect the DO bit to be set 
> as it was sent."

We have added normative language.  I know Andrew was uneasy with that, 
having only heard from three of us (you, me, and David Blacka), but I 
continue to contend that this is the clearer way to say what we were 
saying anyway.  Andrew, if you need to flag this in the proto 
write-up, feel free.

-- Sam