Discuss mailing list

[Quinn Wood <wood.quinn.s AT gmail.com>]

On May 11, 2014, at 4:22, Calum McAlinden <calum AT mcalinden.me.uk> wrote:
>> On 11 May 2014 02:54, Quinn Wood <wood.quinn.s AT gmail.com> wrote:
>> 
>> * This is a misleading proposal, as no peering is taking place. You
>> want to perform one-way (OpenNIC to Namecoin) access. As I've already
>> mentioned this- at best- isn't useful in any way I find obvious.
> 
> Sorry, "peering" was the wrong word to use.  It is useful where
> OpenNIC users would like to access .bit and OpenNIC using a reliable
> DNS server near them. Setting up a local DNS resolver is too difficult
> for some users.
> 
Have you seen a real world desire for this that is best fulfilled by every T1 
operating this access? 1.8G is a lot of space to use [on a nameservers of 
OpenNIC's scale] for something that's only interesting to you. Someone who 
knows better than me may correct me, but this would probably be larger than 
all our other zones combined.

I don't want to see OpenNIC become a one-way gateway to every DNS alt-root 
out there. Small, unorganized alt-roots are bad.

>> * There's no charter. I don't know that one is possible, because
>> nobody here is in control of the top level zone. What are you going to
>> do, refuse to let people resolve using your DNS server if they
>> register a domain you don't like or use their domain to crack/spam?
>> Tough, their domain is still there and kicking, all with the glorious
>> implicit support of OpenNIC that the anti-cracking/spaming policy was
>> intended to prevent.
> 
> I do not intend to modify any domains. We still resolve the ICANN
> domains, many of which contain illegal/cracking/spam websites.
> 
Then you don't need this proposal. If we aren't going to offer information 
and registrations/administration (though NewNations may be guilty of this as 
well, if they actually still operate and we peer with them) on an 
OpenNIC-accessible site we don't need to have a mandatory, community-wide 
access system.

>> * There's no (proof of) an information, registration, and
>> administration service (web-based or otherwise.) This is possible with
>> Namecoin, just as you can proxy between typical DNS clients and
>> namecoind.
> 
> Correct, this is not necessary as domains are registered through the
> Namecoin client or through a variety of web based services.
> 
Yes it is necessary. Otherwise there's no point in us running it. See above. 
To expand upon my requests for evidence of real demand, as well as notable 
sites people would be gaining access to- and in an absence of contrary 
information- I submit that people interested in Namecoin will probably be 
able willing to use it properly. In a way that doesn't undermined the 
decentralization and validation it offers. Using the client, a helper proxy 
DNS server, and their own DNS settings.

>> * There is no (proof of) a service we as a community can rely on to
>> be stable and secure.
>> * There is no (proof of) an admin we as a community can depend on, get
>> a hold of, or boot in the ass when things are broken.
> 
> I can't really answer this. I have only been around for several
> months, but I am a trustworthy server operator. You have my full
> permission to "boot me in the ass" (metaphorically).
> 
Good. You should still have proof of a running service that you've 
theoretically worked the starting kinks out of before asking other people to 
approve a community-wide deployment of said service.

>> You want to provide a service, do it on your own time- just like
>> DNSCrypt and operator-specific bridges/peerings.
> 
> If you're talking about tier2s, you said yourself (with regard to
> voluntary support from individual tier 2s):
> 
> "As always, I am against this kind of bypassing of consensus. We do it
> or we don't do it."
> 
That comment isn't remotely directed towards independently-run features. It's 
directed towards the "Hey everyone, add this TLD without proper consensus and 
then let's vote on it when it's already the norm to run it!" attitude we were 
getting a while back as well as any future, unrelated consensus-bypassing 
weaseling in new TLD matters.