[rfc-i] Summary: use case for 2119 markup

Nico Williams nico at cryptonector.com
Fri Jun 20 15:11:39 PDT 2014

On Fri, Jun 20, 2014 at 4:52 PM, Ted Lemon <mellon at fugue.com> wrote:
> On Jun 20, 2014, at 5:07 PM, Nico Williams <nico at cryptonector.com> wrote:
>> Maybe I'm just tired and slow today.  What's the nuance that's being added?
> Is it capitalized?   If so, does that mean normative?   Is it lower case?   Does that mean not normative?   Is it marked?   Does that mean that all keywords not marked aren't normative?   Even if they are capitalized?

RFC2119 doesn't say, does it.  We've lived with it as it is.
Therefore it's either a) time to make sure it (a successor) does say,
or b) this is not a process issue.

> Anyway, can we please drop this?   I really don't want to fight this out here.   I'm just pointing out that at least from my perspective as an AD who reviews a lot of documents, this is a significant change, and, again from my perspective, what this change means isn't clear.   If you think I'm just talking nonsense, feel free to ignore me, but Heather asked if she'd listed all the issues that came up, and these are two issues that weren't in her list that I consider significant.

No, we can't drop it yet unless the RSE chooses to drop it
prematurely.  If it is a process issue then it's one that we should do
something about.  If it isn't then it isn't and then you can drop it
if you still want to.

At least two people have posted that they think this proposal is a
very big deal from a process perspective.  Appeals to authority aren't
and shouldn't be sufficient to shut down this thread.  You're
appealing to authority ("... as an AD ..."), therefore, as the
authority you're appealing to, it is your responsibility to explain in
detail and sketch a way forward.  Neither the RSE nor other
participants should be made to feel that they should give up on this
or anything else on account of appeals to authority.

I have to say that appeals to authority by the same authorities are a
lot more of a real and serious process problem (a very big deal
indeed) than anything else we've seen in this thread.


More information about the rfc-interest mailing list