[tor-talk] Time for p2p, content addressed, pre-emptively cached web pages - Tor Project Accuses CloudFlare of Mass Surveillance, Sabotaging Traffic

[Zenaan Harkness]

On 2/28/16, Александр <afalex169 at gmail.com> wrote:
>> me at beroal.in.ua
>> You mean instead of an expected webpage you see a CAPTCHA?
>
> Yes i do. Moreover, it is an endless captcha (you hit the right
> combination, but you get a new one) OR a series of captchas that you just
> can't read.
> Of course, one can play with the restart of Tor, but it kills much time and
> ​ ​
> it doesn't always bring success.
>
> Besides, it's getting worser. In the last two years i see it on more and
> more webpages. We should definitely think about a solution, before they
> cover most of the internet (except the onion pages).

After seeding your local cache with peers by doing a torrent download
with something like Red Hat/ Fedora, then afterwards, downloading most
"magnet" links works in relatively short time - i.e., your peers have
a DHT and/ or some built in (hidden to the user) search mechanism to
download the torrent file (torrent meta data file).

That meta data file has enough information to download the pieces of
the full torrent, and your peers all pitch in.

SO, surely a similar principle could work with web pages "oh, ahoy
there me peerrrs, I'd be likin' ta download them tharr web page
assets";
"no problem, here's a few" says one peer.

Perhaps need a O2D transparent conversion algorithm. What? You haven't
heard of O2D? Well duh! Aren't you in the dark ages - that's the
Open-net to Dark-net content addressing conversion, like duh!

Actually, it's more like:
- curl URL url-xform-dir
- git add url-xform-dir
- git commit
- return HEAD as the URL meta data hash

Then all I have to do is meekly ask my Tor peers "Hey yo! Like HEAD
yo!" and BAM!!! there's firetruckin' web page!!

You know, like BAM!!!!