gTLD

Other projects
Post Reply
andyfletcher
Posts: 3
Joined: Fri Jun 13, 2014 9:22 am
os: linux
Location: .qa .om
Contact:

gTLD

Post by andyfletcher »

ICANN closed their current round of applications for gTLD some time ago, see http://newgtlds.icann.org/en/ but it is probable that they will start another round in a year or so. If the Namecoin community want a chance to register .bit (hasn't been bid for it in the current round) then there is a lot of planning to do in advance for the application.

They have published an applicant guidebook http://newgtlds.icann.org/en/applicants/agb which is well worth a read.

This planning falls into several categories,

Organisational
There will have to be an organisation which will act as the registrar. This could be the bitcoin foundation, a-to-be-formed Namecoin foundation or some other group. They need to show a proper governance model and have an appointed board of identifiable persons. Also they have to show that they speak for the community.

Technical
There are a stack of requirements in the applicant guide book. The most technically demanding are colecting data for the WHOIS service and managing the primary nameservers. Neither are very difficult but they have implications for how registrations are collected - maybe the namecoin id/ records need expanding or a whois/ record needs creating. There will be a few other related issues but the one which will give other registrars a real headache in their applications - managing registrations - is already built-in to namecoin. We probably need to exclude certain subdomains for policy reasons.

Financial
There needs to be some money to pay for the registration and operational fees afterwards. I expect this to be fairly easy, maybe a kickstarter approach or a simple appeal for donations prior to submission of the registration. Ongoing costs could be managed in several ways including some sort of 'tax' on updates or maybe a simple payment in namecoin to an address controlled by the organisation before the subdomains are exported into the public gTLD.

Lots to think about.

Is there any interest in discussing these topics and putting work into aligning Namecoin for future gTLD integration?

Andy

phelix
Posts: 1634
Joined: Thu Aug 18, 2011 6:59 am

Re: gTLD

Post by phelix »

andyfletcher wrote:ICANN closed their current round of applications for gTLD some time ago, see http://newgtlds.icann.org/en/ but it is probable that they will start another round in a year or so. If the Namecoin community want a chance to register .bit (hasn't been bid for it in the current round) then there is a lot of planning to do in advance for the application.
Actually there has been something going on in this direction:
http://tools.ietf.org/html/draft-grotho ... p-names-02

Note that should .bit ever be stolen by ICANN it won't be a biggy as we will simply switch to .b or so.
nx.bit - some namecoin stats
nf.bit - shortcut to this forum

andyfletcher
Posts: 3
Joined: Fri Jun 13, 2014 9:22 am
os: linux
Location: .qa .om
Contact:

Re: gTLD

Post by andyfletcher »

Interesting read, it effectively says .bit should never be in the public DNS space. Still an IETF draft but should become a RFC without too much trouble, I recognise some of the authors.

Section 5.5, states that if the local resolver can handle .bit it should return DNS compatible results. If caching DNS servers receive .bit queries they should by default immediately return negative responses. If they are configured to accept .bit queries then they should not cache negative responses as this would slow down propagation of new domains.

All this pushes the use of resolvers for .bit outside the normal DNS infrastructure and discourages ISPs etc from supporting resolution of .bit queries moving the domain into the realm of a small subset of users who have local blockchain clients or special zone files pointing at dedicated DNS servers.

All the public .bit domain servers I have tried appear to be unavailable which means it is impossible to deploy a lightweight resolver reliably which keeps .bit from widespread adoption unless there is some easy way to transfer precompiled zone files to clients.

biolizard89
Posts: 2001
Joined: Tue Jun 05, 2012 6:25 am
os: linux

Re: gTLD

Post by biolizard89 »

andyfletcher wrote:Interesting read, it effectively says .bit should never be in the public DNS space. Still an IETF draft but should become a RFC without too much trouble, I recognise some of the authors.

Section 5.5, states that if the local resolver can handle .bit it should return DNS compatible results. If caching DNS servers receive .bit queries they should by default immediately return negative responses. If they are configured to accept .bit queries then they should not cache negative responses as this would slow down propagation of new domains.

All this pushes the use of resolvers for .bit outside the normal DNS infrastructure and discourages ISPs etc from supporting resolution of .bit queries moving the domain into the realm of a small subset of users who have local blockchain clients or special zone files pointing at dedicated DNS servers.

All the public .bit domain servers I have tried appear to be unavailable which means it is impossible to deploy a lightweight resolver reliably which keeps .bit from widespread adoption unless there is some easy way to transfer precompiled zone files to clients.
There is no reason for ISP's to support .bit via their DNS infrastructure. It would have no advantages over OpenNIC, and would dilute the reputation that Namecoin has as being resistant to censorship and hijacking.

You're right that we need lightweight resolvers, but there are good and bad ways to do that. ISP relaying via DNS is a bad way. libcoin's upcoming UTXO client is a good way.
Jeremy Rand, Lead Namecoin Application Engineer
NameID: id/jeremy
DyName: Dynamic DNS update client for .bit domains.

Donations: BTC 1EcUWRa9H6ZuWPkF3BDj6k4k1vCgv41ab8 ; NMC NFqbaS7ReiQ9MBmsowwcDSmp4iDznjmEh5

mightbemike
Posts: 57
Joined: Fri Apr 25, 2014 4:40 am

Re: gTLD

Post by mightbemike »

I hope the community sees this as undesirable. Going down the ICANN path is a slippery slope full of potential issues, and Namecoin has no legal staff or bankroll.

Take the financial considerations for example. It's not simply a matter of coming up with the $185k application fee, and demonstrating that the technical and financial means exist to operate the registry function for a few years. Companies can easily challenge, and escalate the costs. In the current round, it cost only $50k to challenge a string application, and if a negotiated settlement cannot be reached it goes to auction.

Case in point, .BUY gTLD. Amazon outbid Google and paid $4.6 million. Any company that saw .BIT up for grabs and wanted it would likely think they could take it; it's obvious Namecoin does not have the deep pockets to fight off any significant challenge.

While I have no insight into the motivations of any of the players, it appears to some observers that one particular company applied for many new gTLDs which were all in contention. Those observers have suggested that their intention was to negotiate for nice lucrative settlements. See .GREEN as a good case study. Expect challengers.

So why subject .BIT to that potential trouble?

If the next process resembles the current one, nobody else will be able to apply for .BIT anyway, it's safe. Regardless of the application "type" (would this be community TLD?) one of the top criteria for acceptance is that it causes no confusion in the marketplace. That criterion is very unlikely to go away, because the IP holders need this. Avoiding brand confusion is their top priority, along with requiring accurate and specific ownership data in WHOIS records.

And ICANN will require the ability to seize domains - mark my words. The GAC (national governments) is the most powerful group among ICANN stakeholders, and this is their most cherished feature.

To submit .BIT to ICANN oversight is reckless and counterproductive. ICANN has no right to determine the fate of .BIT unless this community cedes that authority to them. To do so means that Namecoin will submit to their rules on issues like compliance with WHOIS data collection requirements, which are getting more stringent over time.

If Namecoin compromises it's core values, like protection from domain seizures and censorship, then it becomes nothing more than a novelty.
NMC: NFhmGAqzRpZbGs3uCPPo7DJKuscuL4Aap2
id/mightbemike

phelix
Posts: 1634
Joined: Thu Aug 18, 2011 6:59 am

Re: gTLD

Post by phelix »

mightbemike wrote:I hope the community sees this as undesirable. Going down the ICANN path is a slippery slope full of potential issues, and Namecoin has no legal staff or bankroll.

Take the financial considerations for example. It's not simply a matter of coming up with the $185k application fee, and demonstrating that the technical and financial means exist to operate the registry function for a few years. Companies can easily challenge, and escalate the costs. In the current round, it cost only $50k to challenge a string application, and if a negotiated settlement cannot be reached it goes to auction.

Case in point, .BUY gTLD. Amazon outbid Google and paid $4.6 million. Any company that saw .BIT up for grabs and wanted it would likely think they could take it; it's obvious Namecoin does not have the deep pockets to fight off any significant challenge.

While I have no insight into the motivations of any of the players, it appears to some observers that one particular company applied for many new gTLDs which were all in contention. Those observers have suggested that their intention was to negotiate for nice lucrative settlements. See .GREEN as a good case study. Expect challengers.

So why subject .BIT to that potential trouble?

If the next process resembles the current one, nobody else will be able to apply for .BIT anyway, it's safe. Regardless of the application "type" (would this be community TLD?) one of the top criteria for acceptance is that it causes no confusion in the marketplace. That criterion is very unlikely to go away, because the IP holders need this. Avoiding brand confusion is their top priority, along with requiring accurate and specific ownership data in WHOIS records.

And ICANN will require the ability to seize domains - mark my words. The GAC (national governments) is the most powerful group among ICANN stakeholders, and this is their most cherished feature.

To submit .BIT to ICANN oversight is reckless and counterproductive. ICANN has no right to determine the fate of .BIT unless this community cedes that authority to them. To do so means that Namecoin will submit to their rules on issues like compliance with WHOIS data collection requirements, which are getting more stringent over time.

If Namecoin compromises it's core values, like protection from domain seizures and censorship, then it becomes nothing more than a novelty.
lol. Should the paper go through it only means that they will keep .bit reserved, nothing more.
nx.bit - some namecoin stats
nf.bit - shortcut to this forum

mightbemike
Posts: 57
Joined: Fri Apr 25, 2014 4:40 am

Re: gTLD

Post by mightbemike »

phelix wrote:lol. Should the paper go through it only means that they will keep .bit reserved, nothing more.
Reserved for what? There's no real chance of any application for .BIT to ever be approved by ICANN unless it came from this group.

Not causing confusion in the consumer marketplace is such a fundamental part of the process, any application would surely be denied. It would take very little effort to get an application for .BIT rejected, so there's no reason to worry in my humble O.
NMC: NFhmGAqzRpZbGs3uCPPo7DJKuscuL4Aap2
id/mightbemike

phelix
Posts: 1634
Joined: Thu Aug 18, 2011 6:59 am

Re: gTLD

Post by phelix »

mightbemike wrote:
phelix wrote:lol. Should the paper go through it only means that they will keep .bit reserved, nothing more.
Reserved for what? There's no real chance of any application for .BIT to ever be approved by ICANN unless it came from this group.

Not causing confusion in the consumer marketplace is such a fundamental part of the process, any application would surely be denied. It would take very little effort to get an application for .BIT rejected, so there's no reason to worry in my humble O.
Reserved in that they don't sell it to anybody. How would you make an application for .bit rejected?
nx.bit - some namecoin stats
nf.bit - shortcut to this forum

mightbemike
Posts: 57
Joined: Fri Apr 25, 2014 4:40 am

Re: gTLD

Post by mightbemike »

By simply stating the obvious, that .BIT is currently being used, and citing some examples, during the public comment period.

Every new application is put through a set of filters which vary depending on the type of application. i.e. the conditions which must be satisfied for community based application for a TLD like .GAY are not quite the same as those for a brand TLD like .Google, or a GEO TLD like .BERLIN. But thanks to the strong influence of the intellectual property lobby a common thread in the criteria for all is the "will cause confusion in the marketplace" concern. Allowing another .BIT would certainly cause confusion for consumers, and diminish the value of the current DNS in the process, since consumers could not count on a given name resolving consistently. Such is the thinking, as I understand it.
NMC: NFhmGAqzRpZbGs3uCPPo7DJKuscuL4Aap2
id/mightbemike

phelix
Posts: 1634
Joined: Thu Aug 18, 2011 6:59 am

Re: gTLD

Post by phelix »

mightbemike wrote:By simply stating the obvious, that .BIT is currently being used, and citing some examples, during the public comment period.
If it is so simple: all the better :mrgreen:
Every new application is put through a set of filters which vary depending on the type of application. i.e. the conditions which must be satisfied for community based application for a TLD like .GAY are not quite the same as those for a brand TLD like .Google, or a GEO TLD like .BERLIN. But thanks to the strong influence of the intellectual property lobby a common thread in the criteria for all is the "will cause confusion in the marketplace" concern. Allowing another .BIT would certainly cause confusion for consumers, and diminish the value of the current DNS in the process, since consumers could not count on a given name resolving consistently. Such is the thinking, as I understand it.
nx.bit - some namecoin stats
nf.bit - shortcut to this forum

Post Reply