Re: [tor-talk] Status of Torrent on Onion in 2019
On 1/1/19, Fabio Pietrosanti (naif) - lists <lists@infosecurity.ch> wrote:
i'm supporting with some technical analysis a volounteer-non-profit collaborative torrent forum/site with their future-looking upgrades of infrastructure.
I would love them to bring all their public forum
That's just webwork.
and public tracker on Onion, while exporting their +160k releases database as open-data so that Torrent Search sites and Torrent Software with incorporated Torrent Search could use it.
There are probably lots of ways out there. Even maybe something like "getstrike" and "strike-search" API.
My question is about the status of uses of Torrent on Onion, not for "content downloading" but for index tracking:
"index tracking" seems a nonsensical mashup. Though one might say something like "publishing realtime torrent scrape info from trackers on index sites". Definitions... Index - Stupid forums where users publish descriptions for their infohash'es aka: torrent files, the scene. Rarely do centralized indexes use efficient plaintext search and insert interfaces instead of blingy forums and risk. Tracker - Leet tech that clients connect to to find nodes by infohash, ie: DHT, PEX, OpenTracker Client - transmission-bt, vuze, etc Protected - All traffic stays entirely within the overlay network. Exit - Some or all traffic may "exit" to clearnet
- Are there stable / long term Torrent Trackers running on Onion Addresses?
Yes, and i2p etc too. Yet trackers are just bittorrent tech. So there are plenty of indexes as well. Most of the indexes are multihomed from clearnet, some exist only in the overlays.
- How many Torrent Software natively do support Tor and Onion Addresses, only for Trackers
Assuming you're using tor+onioncat p2p v2 onions with v2 HSDir support, that yields ability to transport UDP over IPv6 over tor (many bittorrent protocols use or require UDP)... Thus all clients that support IPV6 can work over "Tor and Onion Addresses". Native .onion support (TCP only, no IP) is not as simple. Nor is split horizon exit or full exit. Same for trying to get TOR space to interop with I2P space or CJDNS space, client multihoming, etc.
(not for content-download-upload-exchange)?
This is moot as typical users don't run their clients just look at the scrape, they do it to seed and leech. Which of course they can do 24x365 over both tor and i2p, and other overlay networks. And due to rampant censorship and loss of free speech on clearnet they are doing it more and more. See all the recent bans by Youtube, Facebook, Twitter, etc etc etc. The overlays are now the only reasonably safe spaces for free speech to thrive. That's sad. Users may also want to look at DTube, IPFS, Brave Browser, BitChute Webtorrent, lists of p2p apps and overlay networks, distributed file sharing and storage, blockchain and cryptocurrency, etc for other ideas and things to read and try. Lots of new protected overlay networks and apps are now online, and more will be coming in the years ahead. Things are changing fast. Plug in and have fun :) https://en.wikipedia.org/wiki/Bittorrent https://en.wikipedia.org/wiki/Vuze https://en.wikipedia.org/wiki/Transmission_(BitTorrent_client)
On 1/1/19, Fabio Pietrosanti (naif) - lists <lists@infosecurity.ch> wrote:
a volounteer-non-profit collaborative torrent forum/site with their future-looking upgrades of infrastructure. I would love them to bring all their public forum and public tracker on Onion, while exporting their +160k releases database as open-data so that Torrent Search sites and Torrent Software with incorporated Torrent Search could use it.
Examples of indexes / forums: http://uj3wazyk5u4hnvtk.onion/ http://rutrackerripnext.onion/ http://rutracker.i2p/ http://ujyqbawolalwvdpy33v2wjuaspnu4ym3fsl3fd6mldjblzzqlvlq.b32.i2p/ The minimum fields for export / web presentation in any index / forum / dump are: key: infohash data: title Other data, including that contained in legacy .torrent files, is secondary metadata to primary search and retrieval. As seen in above examples, even the title fields are rarely conformant to any minimal spec or format, and are often childish... so as to make search and downstream use more difficult and bloated. Those having control over titles on their own releases, lint over databases, or operating insertion API's, should try to please the most picky of librarians. Some projects are also now using blockchains as bittorrent search and insertion indexes. News of one, albeit centralized, twist on that... https://www.engadget.com/2018/07/25/blockchain-company-tron-buys-bittorrent/ http://blog.bittorrent.com/
participants (1)
-
grarpamp