Creating an alternate internet. Sort of.

bits and pieces of nerdom.

All content © NFGworld, unless otherwise noted, except for stuff we stole. Contact the editor-in-chief : baldbutsuave@thissitesdomain, especially if you are an attractive young female willing to do nude photography modelling. All rights reversed. 979

Forum Search Members The Bedroom About

Forum: NFGworld Forums › Science, Tech and Such

Creating an alternate internet. Sort of.

(Administrator) 2010-12-01, 13:11 #1

Member since May 2011 · 2203 posts · Location: Brisbane
Group memberships: Administrators, Members

Show profile ·

Link to this post

Subject: Creating an alternate internet. Sort of.

Given the recent American penchant for being asshats and arbitrarily seizing control of parts of the internet, it's interesting to see brokep (of The Pirate Bay fame) suggesting it may be time for an alternative, non-American root system to assign domain names and generally keep the internet functioning. Distributed, international, and hopefully outside the control of asshats, American or otherwise.

I popped into their IRC channel recently. I love listening to (reading) the things smart people say. The discussion was fast and technical and fascinating. I asked a couple of questions, and got a couple of replies.

<NFG> Given that this sort of alternate root dns has always been possible, and that it has been tried before with very minimal success, what's the confidence of success this time? Does it realy feel like the time has come and this time it might work? Or is it still in the 'let's throw the idea out there and see if it floats' stage?

<Gekz> NFG: both, really.
<Gekz> NFG: because it's not just a competiting root
<Gekz> it goes over the current roots head
<Gekz> if it works, it would be the first successful distributed dns solution
<interfect> An alternate root needs a central person to run it
<dnsSoldie> NFG: even the idea itself is frightening to those like HS who would shut down DNS and domain names with censorship.
<Ve-> It's not "competing" per se. There's no competition. It's just other.

<NFG> so the key differentiator this time is its inherent uncontrollability? That seems to be completely at odds with its ability to be functional. Sounds like a good challenge.

<Gekz> is it a problem? I don't know.
<Ve-> Depends on the design.
<Gekz> we've never had it happen before.

<BitSlayer> we've never had p2p-dns before either
<Gekz> mhmm.
<Ve-> Indeed.

<Anubis_> Is Justin Biber aware of this project ?
<Gekz> Anubis_: I certainly fucking hope not

It seems the internets are in capable hands.

BLEARGH

(Administrator) 2010-12-01, 15:10 #2

Member since May 2011 · 2203 posts · Location: Brisbane
Group memberships: Administrators, Members

Show profile ·

Link to this post

Subject: DNS-P2P Discussions

A user in the channel sent me some logs of technical discussions that took place before I arrived. It makes for interesting reading.

Some background: When you type an internet name into your browser (facebook.com, for example) the name must be resolved first, converting it into a numeric address that the internet uses to communicate. For example, facebook.com resolves to 69.63.189.11.

This name resolution is performed by name servers, special purpose computers that typically do nothing else. Because they play such a significant role, you can imagine they're sort of important.

Well, the internet's root servers resolve the name servers, so they're one level up and as you might expect are even more important. There are only 13 root servers in the whole wide intertubes, and they're all under the control of ICANN, an American organization that purports to work for the whole world but tends to favour American interests.

This is what the DNS-P2P project is all about: de-centralizing the DNS system so that it cannot be controlled by any single source. It's a challenging but not insurmountable job. If it's to be done and gain any kind of popular traction; if you or your grandmother is ever to find themselves using it instead of or as well as the existing system, it had better work and it had better be trustworthy.

So this is what they discuss: an alternative, decentralized, accountable but uncontrollable, trustworthy alternative to the increasingly untrusted ICANN. (ICANN has always been an organization controlled by greed, BTW. This isn't new).

So, on with the logs:

On the issue of Cybersquatting:
Someone asks: it's probably impossible to both prevent squatters and allow everyone to register whatever domain he wants -- any thoughts on this aspect?

No solution seemed to present itself, only more problems. by emulating the ICANN way of designating a tld to a specific use, we are opening ourselves up for the same squatting problems.

Since there can only be one coke.com, the discussion touched on the value of top-level domains (TLD, like .com, .edu, .org etc). I fail to see how the squatting problems are any different if you have a tld or not. I concur. Karl Auerbach, one of the 'good guys' who was part of the ICANN board at one time, in a sort of attempt to infiltrate the den of bastardry and maybe do some good, once suggested that there's no real reason not to have an endless supply of TLDs. coke.com becomes less valuable when Coke controls dot coke. drink.coke. ICANN could have done that, so says Karl, but was too addicted to the money to try and move away from their incredibly lucrative .com monopoly.

Talking about potential new TLDs:
Which brings us to the new system of TLDs: squatters are going to get *everything*.restraunt and *everything*.place and *everything.porn. It seems at this time the focus will be on dot p2p until until we can get our shit together and expand..

On the issue of distribution:
This is the whole point of the system. Everyone that uses the service will [be] a node, if I'm understanding right.

On the point of it all:
Their wiki spells out the goals pretty clearly. To start, it seems, the focus is not just the .p2p TLD, but the promotion and continued functionality of peer to peer services at threat by short-term thinkers like politicians and media executives. I think that's the way to go. If you want the best torrent throughput, you'll need to run P2PDNS. I think the torrent client app authors would be onboard quickly.

hmm is .fur broken? Firefox can't find the server at register.fur

On the future:
Honestly I think we should be very careful with messing around with real tlds
agreed

For the time being then it's just .p2p. Start small, make it work, make it secure. Then start plotting domination.

On conflicts:
The internet proves one thing beyond a doubt: griefers will be griefers. They'll rig the system, exploit the loopholes and basically be dicks because they can. How can you deal with this? How can an open system smack the assholes down without turning into another ICANN, arbitrarily smacking those it disagrees with? an idea for the authority-problem: a 'crowd authority' could be implemented like this: if someone comes across a fake/wrong whatever site it could be voted down, and these voting count finds a reflection in the node's entry

then someone will create bots to downvote some sites

Not an easy answer in sight.

On real-world takeovers:
obviously you couldn't have a TLD that already existed like .com or etc., but what would happen if ICANN made a .free .bbs etc domain in the future that already existed on this system?
already occuring, and not much you can do unless you got deep pockets

BLEARGH

RSS Feed for this thread

We love UNB by Yves Goergen!