Re: [DNSOP] meta issue: WG to discuss DNS innovation (was Re: draft-hzhwm-start-tls-for-dns-00)

[Suzanne Woolf <[e-mail]>]

On Feb 18, 2014, at 11:57 AM, Andrew Sullivan <[e-mail]> wrote:

> No.  This was precisely my point.  For most of the stuff people want,
> the work should be in a WG that does not have "DNS" in its name.
> That's the _key_ point.  We protocol weenies need to get out of our
> comfy chair and go learn why in the world people want to put
> _anything_ in the DNS.  Our failure to do that is how we ended up with
> the situation we have: we keep stamping our feet and saying that the
> DNS works _just fine_ with new RRTYPEs, for instance, except that many
> who want to use them cannot.
> 
> My view was that if there was something that was DNS-specific that
> needed attention, people should have a BoF.  And look!  We have 2 BoFs
> on specific DNS topics in London.  The questions are narrow and
> focussed, not big floppy "gee, this looks like DNS, so it should be
> swatted over to the DNS weenies" ones.  

I'm sympathetic to this view, assuming I actually understand it, but not entirely sure what it says in terms of our charter discussion: should DNSOP go away, continue to address specific topics per the current charter, or add some kind of v6ops-like requirements/coordination-with-the-people-who-really-have-the-problem function? 

As I see it, we have another problem, complementary to the one you're highlighting, which is the number of situations arising where working groups across the IETF want to do something that makes certain assumptions about, or demands upon, the DNS, without DNS protocol weenies in the room at all. This strikes me as….suboptimal.

Is that a problem you (generically, not just Andrew) see as a concern? If so, do you see any general principles for navigating it? 


thanks,
Suzanne