People have been trying to make alternative DNS name systems and alternative root nameservers a thing for 20+ years now.
Has never caught on. On the other hand I 100% endorse sci-hub having a presence as a tor service.
Has never caught on. On the other hand I 100% endorse sci-hub having a presence as a tor service.
- There is no centralized party that owns Handshake. This allows for any party to step in and contribute to the protocol. We've already seen this play out in building Namebase (I'm the CEO) as numerous unrelated parties have come together to further Handshake adoption
- It's not just about opening up the TLD namespace. Handshake's main technical goal is to improve the security and censorship-resistance of DNS by shifting the root of trust from CAs to a distributed ledger. My article on the technical improvements Handshake can provide was previously discussed on HN [1]
- TLDs aren't sold for a set price. There's a vickrey auction which awards the TLD to the highest bidder. This creates a better distribution of names than selling at a set price or to the first buyer
- The Handshake coin (HNS) provides an incentive for miners to provide security to the network and for holders to support ongoing adoption of the protocol, similar to how Bitcoin holders have put in massive efforts to evangelize it (I recognize that on HN people may find the Bitcoin evangelists annoying but I believe Bitcoin wouldn't be where it is today without them)
for real money, USD or EUR or such? where does this money go? or do you have to first mine some cryptocoin, or buy it with real money, and it then goes to the entity 'namebase'?
scroll down here to the 'how it works': https://www.namebase.io/
this quite honestly looks really sketchy to me. I'd much rather take my chances with registering a domain name among the myriad of ICANN root nameserver gTLDs and ccTLDs than buy some weird, obscure cryptocoin, which can then be used to buy 'domain names' that only 0.000001% of the client devices presently operating on the planet can successfully resolve into IPs.
https://learn.namebase.io/starting-from-zero/buy-hns
all of this just looks like somebody has grafted 'blockchain' and 'crypto coins' onto the same alternative root DNS ideas that failed twenty years ago: https://en.wikipedia.org/wiki/Alternative_DNS_root
"Purchasing HNS with USD is currently limited to those with a US passport or ID."
Historically, blockchains haven't been terribly firm on the "irreversibility" point. Ethereum reversed the DAO transactions. The BTC rollbacks in 2010 and 2013. Etc etc.
In practice, whoever owns the client update server and whoever has 51% of the hashing power can make arbitrary changes to the chain. For most cryptocurrencies, a meeting of these stakeholders could fit into a phone booth.
51% attacks on blockchains used as a store of value are bad because an attacker can spend their coins (ie BTC) on an exchange, withdraw their profits, then perform a 51% attack on the previous block to take back their spent coins.
On Handshake, that same attack exists for HNS, but DNS updates on chain take 36 blocks (about 6 hours worth of transactions) to propagate, which is significantly more expensive and unlikely than pulling off a normal 51% attack on a single block (it gets exponentially more unlikely). Furthermore this wouldn't even be an attack per se. An attacker may be able to undo a DNS update, but they wouldn't be able to falsify DNS records because only the owner of the name who controls the private key would be able to submit valid UPDATE transactioins.
Sure, sites like TPB and Sci-Hub need to switch to a different domain from time to time bringing some annoyance to their users, but that's about it. Only if all domain registrars at once started actively removing "shady" domains would an alternative DNS ever come to spotlight.
I have been in a few places with governments that qualify as "regimes." I have used the internet in some places that have a reputation for internet censorship.
What I have observed is that most of those billions of people you say would benefit from a censorship-resistant internet have little desire to say anything worth censoring. Free speech is a particularly American preoccupation, and this value is not universally shared. The overwhelming majority of people living under such "regimes" are not engaged in active resistance to them. They just want to get on with their lives, make money, go out with their friends, etc.
And they don't speak English anyway so it's irrelevant to them if some American websites are blocked. Those are not the websites they want to look at. The people who do want to look at them use VPNs, which are tolerated as a loophole because total blocking with no possibility of circumvention is too much trouble and would have negative economic consequences.
And if they ever do want to say something that would attract the ire of the censors, what would be the point of saying it to a tiny audience locked behind a wall of obscure tech? They say it on systems that have mass adoption in those countries, and they accept the consequences, which are usually small, because by the time they put it on their social media accounts thousands of others are saying the same thing, and there is strength in numbers.
Sure, some people do want this tech and use the current iterations of it. Most don't. It simply does not solve a problem they feel they have.
In any case, the conversation was about censorship of the DNS system. Most censorship via DNS that I have experienced was either enacted by private companies who decided they didn't like being the registrar for sites they objected to on the basis of values, not legality, or it was enacted by Western states at the behest of private companies who wanted to defend their intellectual property.
And censorship is not all that useful. It happens a bit, but why bother when amplification of the desired message is much more effective and generates better press by definition?
[0]: scihub22266oqcxt.onion
Plus there's a myriad of questionable TLDs. In my world these are 'alternative DNS'.
Also, something like DNSCrypt can work for clients regardless of DNS root servers.
I get what you mean though, and in that regard you're right.