7 ms·
Has anyone considered a hostile fork of .org? Set up an alternative root that initially mirrors PIR. Cache existing domains until their registration expires. En
by whoopdedo 7y ago
Has anyone considered a hostile fork of .org? Set up an alternative root that initially mirrors PIR. Cache existing domains until their registration expires. Encourage domain owners to register with the fork using enticing terms. Like offering to rebate the cost of registering with PIR to be in both databases. Or offer much longer terms. Solicit ISPs to point their DNS to your root. Set a cut-off date that after which you'll stop resolving new domains registered with PIR. Give the internet the ability to vote with their feet and wallets who gets to be in charge of the root.
- dependenttypes 7y agoThe bigger "fuck you" to ICANN and the best option in general which would prevent similar issues from occurring in the future would be to move away from DNS and move to .onion addresses or something like namecoin instead. We don't even need the majority of the world to start using them, we just need browsers to add support.
- michaelt 7y ago> Has anyone considered a hostile fork of .org? There have been various attempts at alternative DNS root [1] by people unhappy with ICANN. In the past, these attempts haven't been able to convince ISPs to point their users to the forked service - even if they mirror every existing domain free of charge. And you don't just have to convince domain registrants and end users; you need to bring mailserver administrators and certificate authorities along too. I'm not saying it's impossible - Google could force things unilaterally with a change to Chrome, for example - but history shows users are slow to vote with their feet. [1] https://en.wikipedia.org/wiki/Alternative_DNS_root https://en.wikipedia.org/wiki/Alternative_DNS_root
- troquerre 7y agoThat's kind of what Handshake.org is trying to do, although instead of forking existing TLDs Handshake lets people register their own TLDs outside of the 1500 TLDs the ICANN has already provisioned. There have been many alternative root attempts in the past, but I think the only way for one to succeed is if it starts off by augmenting the existing system instead of making users choose between one or the other (that can come later).