Re: [DNSOP] [[e-mail]: New Version Notification for draft-sullivan-dns-class-useless-01.txt]

[Suzanne Woolf <[e-mail]>]

On Mar 18, 2016, at 3:10 PM, Paul Vixie <[e-mail]> wrote:

> Shane Kerr wrote:
>> Mark Andrews<[e-mail]>  wrote:
> 
>>> Just because we can't think of a good way to use class today is not
>>> a good reason to shut down the registry.  It really doesn't cost
>>> to maintain a registry that is not being actively updated.
>> 
>> Closing the registry sends a clear message: don't use class. That
>> matches the best understanding today.
>> 
>> If someone thinks of a good way to use class tomorrow, the registry can
>> be re-opened, right?
> 
> i think mark is right-- the cost of this registry is low, and the reservations for HS and CHAOS are in use, though interoperability is low. we should leave the registry in place until there's some need beyond cleanliness to remove it. (for example, if we revise STD 13 in a way that makes $class useful.)

<Chair hat>

As a matter of process, "close the registry" doesn't mean "remove it from the IANA repository". We don't do that, for exactly the reason you stated. It means "state formally we have no process for updating it further."

</chair hat>

IMO a reason beyond "cleanliness" to close the registry is to pave over an entry point for bad ideas. Periodically, $thing_someone_wants_to_do looks like something they could do with a new class, at which point they try to suggest it, are angry when DNS experts say "This only looks like a way to do what you want, it won't work," etc.  

There's an argument to be made that "People considered carefully and don't believe any amount of tinkering will make an idea that depends on CLASS workable, so the option isn't available" will result in shorter conversations about those ideas.


Suzanne