Re: [dnsext] getting people to use new RRTYPEs

[Doug Barton <[e-mail]>]

On 04/26/2013 05:14 AM, Andrew Sullivan wrote:
> I am speaking here as an individual and as an operator.
>
> On Thu, Apr 25, 2013 at 06:41:31PM -0700, Doug Barton wrote:
>
>> 1. Insert the ability into the interface to add freeform stuff
>> 2. Run the equivalent of named-checkzone prior to committing the change
>> 3. Profit!
>
> That's preposterously naive.

Careful, that's dangerously close to ad hominem. :)

> Step 2.1 is "Find that customer who has
> no theory of the mystifying DNS arcana screwed it up, so you can't
> publish, and now you have to contact a human.

No, you don't. You write the documentation in advance about how the 
known type codes work, and when the thing fails you send them to a page 
that says, "I think you meant to do 'this' which should look like 
'that.' If that's not what you wanted, go 'here' which has a big list o' 
DNS records and their proper format."

Not to throw resumes around or anything, but I used to be in that 
business. I've written management platforms for DNS. When I left Yahoo! 
we had over 600,000 domains in our registrar-reseller business. My 
current business is enterprise DNS/DHCP/IPAM management solutions. Prior 
to that I was working for a company that does hosted solutions for mom 
and pop's all the way up to major e-commerce sites.

I'm not talking out of my ass here, I'm explaining that there is a real 
solution to this problem. The real issue (as you point out below) is 
that it isn't free, in either sense (speech or beer).

> Stop.  Invoke expensive
> off-page customer service process."  In some significant number of
> cases, we never get to step 3.  In the DNS business, the margins are
> small.

Thank you for confirming the real problem, which as I have pointed out 
in the past, is that the operators, especially hosting providers that 
cater to naive users, don't want to make ANY changes because it costs 
money on both the front and the back ends.

> This list -- and indeed, much of the IETF, as is often lamented -- is
> heavily populated by people who either are not operators or else are
> not consumer-facing operators, though there are notable exceptions.
> It sseems at least possible that this introduces a bias of ignoring
> problems that others report (or that are observable in deployment)
> because they're not the problems of the people who happen to be here.
>
> I have to agree with John Levine that the usual answer of the DNS
> community to provisioning problems -- in caricature, "This is easy,
> you must be morons."  -- is not helpful, and is indicative of a closed
> mindedness that is unbecoming.

I'm not close-minded, and I'm not naive. I'm also not willing to allow 
all progress on new DNS type codes to be held hostage to providers 
unwilling to invest in their own infrastructure.

Doug