[Confiks]

As with my comment above: you trust the content that is ultimately (up the hierarchy) signed in an IPNS object with a valid signature.

And there must indeed by some central authority giving 'ultimate trust'. This is Zooko's triangle. By default IPNS gives you 'decentralized' and 'secure', but you can also opt to tie an IPNS name to a DNS TXT record, and lose 'decentralized' (at least for the initial lookup) but gain 'human-meaningful'.

[reit]

I think there is a level where DNS -> IPNS ID is even then still better than DNS -> IP address. For DNS -> IPNS you could have a facility in there saying 'This DNS TXT record should not change, if it does something is wrong'. You could lock down the record for a period of time, such as the duration of the lease from the registrar.

Besides, the idea is that IPFS could happily support Namecoin[1], so there's the decentralized, secure, and human-meaningful DNS service. The only 'non-secure' aspect there is the unlikely event of a 51% attack.

[1] https://namecoin.info

[ryan-c]

A 51% attack will let you double spend and roll back transactions. If one wanted to steal an established name on Namecoin, they'd need to either roll back to before it's registration, roll back to before it's previous renewal and force an expiration, or mine past it's next renewal to force expiration. Renewal limit is 36,000 blocks. For a name that consistently renews early every 18,000 blocks a 51% attack would need to go on for months to steal it. Still possible, but very difficult.

Namecoin is merge-mined with Bitcoin by many miners, so if you wanted to 51% it you'd need a pretty large chunk of Bitcoin's hash power. An evil pool could do this, however it would be visible on the Bitcoin blockchain that they were doing it (although not what transactions are present/not present in the attack chain).