Re: wrt: draft-ietf-dnsop-inaddr-required-00.txt

[Lars-Johan Liman <liman@sunet.se>]

briansp@walid.com:
> Is this requirement going to scale reasonably into the day when
> every appliance in my kitchen is connected to the network and has
> its own address?

I see that as less of an issue. If we require that people do it, they
will make it scale. By the time your stove reads e-mail, DHCP and
dynamic DNS will be well deployed.

> I've never really been convinced that the reverse lookup stuff is that
> helpful.

Well, if you're a network admin, traceroutes with text in them is
_really_ helpful at times, provided that the rev-DNS is to some degree
correct. Being able to see that my packet from Sweden go through
Washington DC to reach Spain, gives me a lot of info. With regards to
hosts, its useful when the forward DNS is wrong. "ping foo.org" tells
you "trying bar.net" and you realise that there might be more to
investigate. So there is benefit, but not for the party providing the
information - that's the quirk.

> My first reaction to this document was 'so what', but I see the
> merits of recommending this as a BCP.  Using DNS for this purpose
> isn't going to be a panacea for identifying responsibility domains
> for addresses, but I don't see what harm it could do, and would at
> least provide a standard document for reference purposes.

Someone is going to question the "rules" if there is no motivation for
them. "Why should I put costly manpower into maintaining this if it's
just for someone else's convenience?"

				Cheers,
				  /Liman