Sorry for being too lazy to read all previous posts...
But has anybody actually proposed using bitcoin algorithms to create decentralized DNS to Peter Sunde yet ?
As far as I know the argument against using bitcoin is that if you lost your keys to access the domain you would not be able to recover them.
How does bitdns get around that?
I think this is a good argument against almost any peer-to-peer distributed database of any kind. This isn't something unique to just Bitcoin.
All of this is sort of strange, as I thought the whole point of the exercise was to try and decentralize the DNS database. Instead, there is going to be a central point of control for the whole thing and an official "authority" who is going to be acting as a gatekeeper. Really, this seems to be a betrayal of whatever it is that they were trying to accomplish and they are now becoming an alternate to ICANN, not something which is really a genuine distributed domain repository. There may be some differences as it may become a sort of Napster of domain registration, but it still has that point of attack, whatever that attack may be from either a digital, legal, or physical kind of attack.
It is too bad that they aren't rethinking the entire issue down to its roots, which is an attempt to create a "directory" of the internet. They are also getting IMHO hyper-paranoid about trademarks, as if coca-cola.com isn't sufficient to identify a particular corporation that most people know anyway. So what if somebody creates coca-cola.p2p? Or for that matter wikipedia.p2p?
There also seems to be a mindset that it must follow the convention of:
<computer> <dot> <domain> <dot> <top level domain>
There was an historical reason for the original divisions for .com, .mil, and .edu. They represented three distinct and separate networks originally called "COM-NET", "MIL-NET", and "EDU-NET". It is sort of like how telephone systems used to be completely separated between business exchanges and residential exchanges. When the three networks were finally linked, they had to use some way to identify how to access computers on one network to another, so the "TLD" system was originally set up and some other side "TLDs" were made for "miscellaneous organizations" (.org), "international organizations" (.int), and generic "network resources" (.net). When people outside of America were participating in adding their networks to the internet, they were identified by their country code , now the various "country code top level domains".
Mind you, when the codes were handed out, there was the .us domain which seemed redundant (it was) but those already having .com didn't want to go to a longer .com.us.
Anyway, I contend that the whole notion here of a top level domain is something that completely unnecessary, and that viewpoint has been long advocated by Karl Auerbach
. If you've never heard of this guy, it should be important that he was the one and unfortunately only official elected representative
of North America to ICANN, and sort of a major rebel. He was one of the "at-large" directors that had to get a California court to force ICANN to actually show him their books (he was one of the members of their board of directors and supposedly recognized as such by the organization) and because of his efforts his position was dissolved by the appointed board members made up of people hand-picked by major corporations. The whole experiment of democratically elected representatives to ICANN was killed by the U.S. Department of Commerce... isn't that something?
Anyway, if anybody knows DNS backward and forward, it is Karl Auerbach, and one of his positions when he was on the ICANN board was that there was no need to even have a top-level domain system at all, and certainly the number of top level domains could certainly number in the millions or more. It is simply a directory.This history paper
about DNS goes into the gritty details of how it finally got started. This really is a seminal paper on the history of DNS in terms of how we got to where we are today.Another paper by Karl Auerbach
is certainly worth reading so far as to see what a very well thought out objection to the current DNS system as implemented by ICANN was and how when the whole system was created had people proposing some significant alternatives. Since supposedly we are talking about alternatives to the centralized DNS, I think this ought to be at least one of the key papers worth reading. It shouldn't be surprising that the problems he identified over twenty years ago are still present in the current system, and it is important to note that there was opposition back then even to the current system. If you read between the lines, he was even upset over how IP addresses were allocated... which is a completely different beast altogether.
I don't think that those working with this new "dot P2P" effort are really reading stuff of this nature. I find it too bad as well, as it shows the immoral and arguably illegal actions taken by those who set up the current system almost as if it was a sort of Jekyll Island episode of the Internet. There are a whole bunch of common parallel themes between the formation of ICANN and the Federal Reserve. In this sense, perhaps Bitcoin really is a good fit as it sort of is a merger of the two systems that have been corrupted by aristocrats in a supposedly egalitarian and "officially" classless society.
In short, I think that the "dot P2P" effort is doomed to failure if the continue down the path they are going and not addressing the real problems facing the current system. Many of those problems are being addressed by BitDNS, even if it is a bit marginalized at the moment. And I also think that the argument about lost keys is not something nearly as significant either compared to the major issues about what is wrong with the current DNS system.