Article author here. Just wanted to add that if any of you work for companies that send engineers to the IETF, there is an Internet-Draft RFC that would normalize the use of several P2P pseudo-TLDs such as .gnu, .zkey, .onion, .exit, .i2p, and yes, .bit, which is what Namecoin is squatting on. Would be great to see everybody supporting it.<p><a href="http://tools.ietf.org/html/draft-grothoff-iesg-special-use-p2p-names-02" rel="nofollow">http://tools.ietf.org/html/draft-grothoff-iesg-special-use-p...</a>
I'll plug BitShares DNS--a Namecoin competitor I am working on--here: <a href="http://nmushegian.github.io/dns/" rel="nofollow">http://nmushegian.github.io/dns/</a><p>Whitepaper and FAQ are not quite up to date but you get the idea. From: <a href="https://github.com/nmushegian/dns/blob/master/whitepaper.md#comparison-to-namecoin" rel="nofollow">https://github.com/nmushegian/dns/blob/master/whitepaper.md#...</a><p>- Namecoin issues new coins to miners as a reward for performing merged mining with the Bitcoin network. The namecoin supply is being inflated at nearly 30% per year for several more months, then over 10% for the next several years. Domainshares only ever shrink in supply, when fees are destroyed as implicit dividends.<p>- Namecoin attempts to service multiple namespaces at once. .p2p is highly specialized for servicing the .p2p TLD namespace. The use case is the same as Namecoin's "d/" namespace, which is used for the .bit TLD.<p>- Namecoin's name registration price is fixed at any given time and is independent of the name itself. Domainshares utilizes an auction-like mechanic to incentivize price discovery for names, making sure the final owner pays what it is actually worth. The majority of the final cost will have gone to the network as dividends by the time the auction is over, with a small fraction having gone to bidders as a reward for price discovery.<p>- As a result of the fact that domains are expensive and there are dividends on shares but not domains, there is a high opportunity cost to squatting: holding a domain without making good use of it.
I expect it will obsolete ICANN as effectively as AlterNIC, New.net, OpenNIC, etc.<p>The technical fact that it is blockchain-based really doesn't make much difference as it's incredibly unlikely to be adopted worldwide, due to the network effect of the already-established domain name system.
I am using namecoin to resolve bittorrent sync secrets in my distributed browser Syncnet:<p><a href="https://github.com/jminardi/syncnet" rel="nofollow">https://github.com/jminardi/syncnet</a><p>It was fairly easy to integrate
I have been looking at *coins for a while and the implications of the blockchain-based concensus system are interesting.<p>The system really maps the real world onto computers very well, in that it reduces what used to be technical issues to "political" issues. These systems work on concensus, and as such, require a significant amount of interested parties to work in the expected way. They are very much subject to network effects that only occur after critical mass is reached.<p>Namecoin, in particular, was subject to a major security issue last year: <a href="http://www.reddit.com/r/Bitcoin/comments/1ohyom/fatal_flaw_in_namecoin_found_doesnt_enforce_some/" rel="nofollow">http://www.reddit.com/r/Bitcoin/comments/1ohyom/fatal_flaw_i...</a><p>Blockchains are being used to implement solutions to different problems, and they could really solve some significant problems such as decentralized identity and reputation management (!). The difficulty lies in creating a significant enough "currency" so that miners will become involved and make the blockchain stable and reliable.<p>Bitcoin is a currency in a much stronger sense than namecoin or any other of "not-really-money-coins" around. I wonder if piggybacking on bitcoin might actually be the solution for this situation (i.e introduce other information in the bitcoin blockchain instead of using a brand new one).<p>Sadly, adding external information to the blockchain could be construed as "spamming the blockchain" and therefore not deemed worthy for inclusion in the bitcoin blockchain by miners. So there is a big challenge there.<p>If you are interested in this topic and want to work on related projects feel free to reach out (google my username).
Note that you can use namecoin blockchain with your DNS and HTTP tools with dnschain [0]. Check this out:<p>$ dig @dns.dnschain.net okturtles.bit<p>$ curl <a href="http://dns.dnschain.net/d/okturtles" rel="nofollow">http://dns.dnschain.net/d/okturtles</a><p>Both of them will resolve to whatever info is stored in d/okturtles domain.<p>With the (soon-to-be) DANE support (for those who forgot: DANE is about distributing TLS keys through a channel you trust (it comes from the domain you're visiting) but that is not the same as the final application (it's DNS, not HTTP/SMTP/IMAP/XMPP/etc), so you can prevent MiTM), I don't see what's missing technically to have our own internet.<p>[0] <a href="https://github.com/okTurtles/dnschain" rel="nofollow">https://github.com/okTurtles/dnschain</a>
Given the news around bitcoin one could imagine that stealing 'google.bit' out of the google wallet[1] would be a pretty lucrative target.<p>[1] No I don't really think Google would let that happen but Bob's Bakery Buns might and it could cost him his business.
How long does it take to resolve a namecoin key=>value pair? Considering the variable speed in these sorts of decentralized networks, is there a noticeable lag when you initially request a .bit website?
To me the real problem with namecoin for the purpose stated in the title of this post is that it doesn't emphasize delegation to near the degree necessary to achieve it. The current implementation is much more like the older name system in that it requires the list of canonical names to be distributed as widely as possible.<p>There is a reason the root nameservers only delegate the act of name lookup at the top level. It's just not practical for them to have a complete list, and it's not even particularly desirable for users of it to have their list of names completely public (think internal servers).<p>It is possible to name a delegate nameserver through namecoin, I believe, but last I looked it was a bit iffy and it doesn't require any kind of authentication of results from the delegated nameserver a-la dnscurve.
Why do we asume that domains have to be 'unique' like trade/brand names? I have a name, shared with my grandfather, father and son; in the right context, you would never confuse one for another. Furthermore, my name is shared by probably thousands in my country; again, I have never been confused with any of them. So, if the whole point is to translate human readable (and memorable) words to IPs, and in the event that the DNS returns more than one result, can't the browser display a search-results-like page letting one know there are options? Wikipedia does it and that is THE descentralized knowledge store of the planet. Maybe if I type homes.com I mean local homes first, like Google rankings, but is ok if someone else is using homes.com elsewhere.<p>Think about it...
Has anyone created an easy to use name coin registration system so non-technical people can register names?<p>It seems like that would be a good opportunity?
"(...) the cryptographic-decentralization Zeitgeist makes it an exciting time to have and use names"<p>Yes. Very true. For anyone interested in the above statment i'd highly recommend checking out <a href="http://twister.net.co/" rel="nofollow">http://twister.net.co/</a> - a decentralized micro-blogging spin-off of bitcoin. Unbelievable innovation is happening!
MeowBit is system service for Windows which allows for resolution of dot-bit domains across all applications.
<a href="http://meowbit.com" rel="nofollow">http://meowbit.com</a>
Could we just get rid of naming <i>authorities</i> all together? Why not just let trusted <i>listing agencies</i> handle it and users can simply pick the one(s) they wish to use. For instance, let's say I want to publish a website under the name "sony". Okay, that's fine, but it's going to get a lower priority then the offical Sony, Inc. site b/c any listing agency is obviously going to serve up that site when a user puts "sony" in their browser's address bar. However, a user could opt for an alternate listing agency --maybe one the offers no commercial listings, and then maybe my site would come up through them. Browsers could make it easy to switch between listing agencies and prioritize them.