[Freedombox-discuss] CCC Meeting Notes
On Tue, Aug 16, 2011 at 3:11 PM, Ted Smith <tedks at riseup.net> wrote:
>
>> http://1231231231.onion/ because the http URL spec requires DNS
>> resolution there. One would need to a number
>> of onion schemes otherwise.
>
> Tor has its own lookup protocol for .onion names. The name in
> <name>.onion is actually the hash of a Tor hidden service identifier
> key. You can't resolve .onion addresses via DNS, because they don't have
> IP addresses, because they're _hidden_ services.
For those who are wondering how this actually works, the answer lies
in the fact that
Tor exposes the API of a SOCKS5 proxy to the browser, and the browser
relies on the SOCKS5 proxy to do the DNS resolution work.
The fact that Tor does no such thing, but instead performs a different
kind of lookup and makes a different connection entirely, is invisible
to the browser.
This is generally a good strategy which could be emulated by anyone
wanting to implement alternate naming/routing schemes side-by-side
with DNS/IP: make your 'names' look like DNS names, so they work in
normal URLs/browsers and then make magic happen in a proxy layer.
(I didn't get any responses to my lapcat mail from a few weeks back,
but lapcat is basically an experiment along these lines, allowing me
to connect to different names using different strategies.)
--
Bjarni R. Einarsson
Founder, lead developer of PageKite.
Make localhost servers visible to the world: http://pagekite.net/
Reply to: