There are now many technologies for P2P file sharing:

  • Usenet
  • Bittorrent
  • IPFS
  • Tenfinger
  • more?

And also several technologies for anonymization:

  • VPNs
  • Tor
  • I2P

AFAICT the most popular option is still Bittorrent with various public and private trackers and indexers. Private trackers seem to be the most effective option that communities have converged on.

I’ve tried all of the above solutions over the years and while I think something like I2P sounds great in theory, it doesn’t seem to actually work well in practice. Because it’s not terribly popular, it’s hard to know what to expect.

We clearly still haven’t found the ideal model that can provide both good performance, discovery, access, availability, and anonymity. But do you think we will actually make progress on these problems? We seem to have stabilized on a local maximum.

  • MEtrINeS@lemmy.dbzer0.comEnglish
    11·
    16 hours ago

    then every other system that lets people upload files is P2P, including things like Rapidshare, Google Drive, forums, etc. That wouldn’t make sense.

    P2P means peer to peer.

    Here you have a graphical description of usenet: https://en.wikipedia.org/wiki/Usenet#/media/File:Usenet_servers_and_clients.svg

    From wikipedia https://en.wikipedia.org/wiki/Usenet

    When a user posts an article, it is initially only available on that user’s news server. Each news server talks to one or more other servers (its “newsfeeds”) and exchanges articles with them. In this fashion, the article is copied from server to server and should eventually reach every server in the network

    Please pay attention to: Each news server talks to one or more other servers (its “newsfeeds”) and exchanges articles with them.

    hmmm… it sounds like p2p.

    Here you have the fucking RFC for the command to share articles between servers: https://www.rfc-editor.org/info/rfc3977/#section-6.3.2

    Even if you check the article headers you can see the path the article it took to reach to you. If the article wasn’t uploaded to your server, you can see that some articles that you are downloading came from different servers until it reached your server.

    Tell me, if rapidshare, mega, google drive, etc works like this. Well i don’t think so, because first you don’t even have different servers from different providers… You just have their servers, so they can’t share shit with anybody else’s servers.

    It’s not p2p for the enduser, but that’s why i said it was server p2p.

    OR maybe you operated a NNTP server with only 1 feed? (which i doubt).

    The reason P2P systems exist is that they avoid things like takedowns (since you’d need to take down every user instead of a central server), and Usenet doesn’t have that advantage.

    If you had used usenet before obfuscation took place, you could see that some articles where removed from some server but still available on another ones. That’s why there was a rule of having servers from different backbones https://cdn.rexum.space/usenet/Usenet Providers and Backbones (Version 2025-06-30).svg

    • dan@upvote.auEnglish
      2·
      10 hours ago

      Please pay attention to: Each news server talks to one or more other servers (its “newsfeeds”) and exchanges articles with them.

      hmmm… it sounds like p2p.

      This is describing federation and decentralization, not P2P. The servers communicate with each other. P2P is when users communicate directly.

      Would you consider Lemmy or email as P2P? They’re also federated and decentralized.

      It’s not p2p for the enduser, but that’s why i said it was server p2p.

      Your definition of “server P2P” doesn’t really make sense. Any CDN would fit this definition for example, because CDN edge servers fill data from origin servers when it’s not cached locally, and a lot of that data would be user-uploaded, but I’m not sure anyone would describe Akamai, Fastly, or Cloudflare as P2P.

      Usenet has companies running the servers, and you download from and upload to the company’s servers. Nobody would reasonably describe that as P2P, regardless of how the servers are implemented.

      • MEtrINeS@lemmy.dbzer0.comEnglish
        1·
        9 hours ago

        P2P is when users communicate directly.

        P2P means peer to peer!

        Definition of peer:

        2. countable noun
Your peers are the people who are the same age as you or who have the same status as you.
His engaging personality made him popular with his peers.
Synonyms: equal, like, match, fellow [old-fashioned]   More Synonyms of peer

        “The same status” , Synonyms: equal.

        This is describing federation and decentralization, not P2P.

        It’s very similar but not the same. Using lemmy as example, when i request posts from another server, my server will request only those posts. It won’t request everything. Do you see the difference?

        A CDN is centralized in control but distributed in delivery: one content owner controls the authoritative data, while a CDN operator replicates it across its infrastructure. In p2p nobody controls the authoritative data - if one peer refuses to distribute the data, my server can still get the data from another feed.

        Both federation or CDNs don’t work in a P2P manner. They either have selective replication and/or centralized control.

        Usenet has companies running the servers, and you download from and upload to the company’s servers.

        Again this has nothing to do with being server p2p or not. And btw, stop moving the goal posts.

        EDIT: because i have more than to do than keep replying to this. Here is the output of perplexity. Check the references. Maybe it will be enough for you:

        Yes, Usenet newsgroup servers do share articles directly with one another in a peer-like fashion, and many technical descriptions characterize the server layer as having a “peer-to-peer character.” However, whether the whole system is accurately described as “P2P” depends on which layer you mean (servers vs end users) and how strictly you define P2P.

## How Usenet servers exchange articles

Usenet is built on a distributed network of news servers that exchange articles using the Network News Transfer Protocol (NNTP). Key points:

- Each news server connects to one or more other servers (its “newsfeeds”) and exchanges articles with them. [en.wikipedia](https://en.wikipedia.org/wiki/News_server)
- When a user posts an article, it first lands on their local server; that server then forwards it to its peers, which forward it further, propagating copies across the network. [alchetron](https://alchetron.com/Usenet)
- Servers use feed commands like `IHAVE`/`SENDME` to tell each other which articles they have or want, effectively coordinating replication without a single central master. [bestreviews](https://bestreviews.net/usenet-glossary/)

This server-to-server replication is inherently decentralized: there is no single authoritative Usenet server; instead, there is a changing mesh of servers that mutually share content. [en.wikipedia](https://en.wikipedia.org/wiki/News_server)

## Is this “P2P”?

There are two common perspectives:

### 1. Server layer: yes, it has P2P characteristics

Many sources explicitly describe the collection of Usenet servers as having a peer-to-peer character:

- “The collection of Usenet servers has thus a certain peer-to-peer character in that they share resources by exchanging them…” [alchetron](https://alchetron.com/Usenet)
- Some academic and technical discussions even call Usenet “perhaps the oldest and most successful P2P application,” focusing on the distributed, symmetric exchange of articles among hosts. [csd.uoc](https://www.csd.uoc.gr/~hy554/papers/iptps2004.pdf)
- NNTP-based newsfeeds are symmetric in the sense that each participating server can both send and receive articles from its peers. [en.wikipedia](https://en.wikipedia.org/wiki/News_server)

In this sense, the **server overlay** behaves like a P2P network: nodes (servers) are peers that both provide and consume content, and there’s no single central repository.

### 2. End-user layer: mostly client–server, not classic P2P

For typical end users, Usenet looks more like a client–server system:

- Users run a newsreader (client) that connects to a news server and downloads or uploads articles via NNTP. [alchetron](https://alchetron.com/Usenet)
- Downloaders are not directly connected to other users; they only talk to their provider’s server, unlike BitTorrent or Gnutella where peers connect directly to many other peers. [alchetron](https://alchetron.com/Usenet)
- The identity and address of downloaders are hidden from other users; only the server knows who is connecting. [alchetron](https://alchetron.com/Usenet)

Because of this, some analyses distinguish Usenet from “modern P2P services,” noting that:

- Content distribution among servers is P2P-like, but
- User access is via client–server, and the granularity/symmetry differs from file-sharing P2P networks. [alchetron](https://alchetron.com/Usenet)

## So can it be described as P2P?

- **Technically accurate**: “Usenet’s server network has a peer-to-peer architecture” or “Usenet uses a P2P-like server overlay.” [alchetron](https://alchetron.com/Usenet)
- **Potentially misleading without qualification**: Calling the whole Usenet experience “P2P” in the same sense as BitTorrent or Napster can confuse people, because end users don’t directly share with each other; they share via intermediate servers. [alchetron](https://alchetron.com/Usenet)

A precise description is:  
Usenet is a **distributed system** where **servers peer with each other** to replicate articles (P2P at the server layer), while **users interact in a client–server manner** with their chosen news provider. [en.wikipedia](https://en.wikipedia.org/wiki/News_server)