2 ms·
OK, let's take this apart and use this thread to analyze the alleged failings of torrent model. He raises: (1) A security problem ("something safer"), presumab
by contingencies 13y ago
OK, let's take this apart and use this thread to analyze the alleged failings of torrent model. He raises:
(1) A security problem ("something safer"), presumably regarding identification prior to raids and repression.
(2) A speed (asks for "something faster").
(3) Centralization (best not to "depend on a few persons").
My random thoughts:
(1) Tor offers a reasonable approach for torrent metadata accrual and distribution, but will limit availability. As for actual content, spoofed origins may be possible on some networks, otherwise intermediaries are necessary and then you have Tor or a multi-VPN money-draining complexity fetish. Moving away from the open internet exclusively to Tor/I2P would be a great way to educate the user base and bring support to those projects. Good luck moving all torrent tracking sites though... it may just force people elsewhere. Still, the education benefit would be significant... but there's strength in numbers and fleeing to darknets does not play one of the greatest hands: solidarity, or social strength.
(2) Transfer speed challenges perhaps largely derive from the random distribution; itself an asset. Sourcing content more locally (using AS based logical topology mapping, community-reported speeds between ASs, etc.) may provide some benefit, but ultimately may cause greater harm than good due to increased risk of local distribution and thus prominent node identification and the aforementioned raids and repression. With regards to initial distribution specifically (versus general redistribution after initial seeding), it may be possible to add some sort of priority, anonymous queue for newly posted / incoming data that is based on an anonymous but cryptographically preserved uploader reputation metric (somehow adjudicated by trackers?).
(3) Infrastructure costs money, which is perhaps a major part of the centralization problem with TPB - someone needs the keys to the proverbial kingdom to keep it running. The web in general needs decentralization of hosting, which at the moment generally both exposes identities and is too difficult and expensive for the average user (largely therefore: Blogger, Twitter, Facebook, Flickr, Picasa, etc.) Read earlier today someplace Firefox has a project going in this area at the moment based upon altering consumer browsers to participate in third party data hosting, perhaps this could be co-opted carefully and legally armoured by to preserve an inability for individuals to identify which content they are in fact hosting. Unfortunately its motivator for hosting is the wont to publish one's own data .. elsewhere .. and most people right now are net consumers rather than publishers, so this may be brittle.
Others' takes?