this post was submitted on 10 Nov 2023
29 points (100.0% liked)

Fediverse

17665 readers
3 users here now

A community dedicated to fediverse news and discussion.

Fediverse is a portmanteau of "federation" and "universe".

Getting started on Fediverse;

founded 4 years ago
MODERATORS
29
submitted 11 months ago* (last edited 11 months ago) by [email protected] to c/[email protected]
 

I've been using mastodon for a month or two now. I never used twitter but thought I'd try it out for fun since I love this new fediverse experiment.

Then my mastodon instance started experiencing some downtime and I wondered what happens in this scenario. It seems if the goal is to have lots of smaller instances and decentralize social media, then instances, particularly those not run a big companies (who can reliably fund things for years on end and sell ad space on their platforms), will come and go and users will lose their identity or home base each time this happens along with all their followers and their connection to the wider social graph. This seems not great.

It seems that nostr might actually be a fix for this. In nostr:

  • You publish to multiple relays (essentially instances) and anybody from any relay can follow you.
  • Your messages are signed by your key so you can prove they are authentic.
  • If your relay goes down, people can still follow you via other relays
  • You can change between relays without losing your identity. Your post history and followers follow you, not [email protected].

Doing some reading, it seems people's main criticisms of nostr are:

  • Interface isn't as pretty. Looks like this has come leaps and bounds in the past six months but of course could always use more work
  • Populated by crypto bros. This seems like not an issue long-term, there's plenty of crypto bros on mastodon, you can just not follow them if you don't want to see them. The idea that you can "tip" with a tweet or whatever nostr's term for that is seems pretty interesting.

Basically, at a protocol level, nostr seems better in some important ways and the cons don't seem protocol related but userbase and UI related.

What am I missing on the pros/cons list? Anybody got experiences to share?

you are viewing a single comment's thread
view the rest of the comments
[–] Die4Ever 5 points 11 months ago* (last edited 11 months ago) (1 children)

But you said you can post to multiple relays, and you can choose what relays to post to freely? Doesn't that mean duplicate work for all the relay admins?

I have thought about nomadic social networks before based on public keys, I do generally think it's a good idea. I'm just not sure if the Twitter/Mastodon style is an ideal fit.

[–] [email protected] 4 points 11 months ago* (last edited 11 months ago) (1 children)

Not really. If you are a ban-worthy user on mastodon, other instances may ban your user id, or ban your entire instance if they think your instance's mod actions aren't enough generally. I imagine a "common blocklist" like currently exists in e-mail, mastodon, and other federated networks would emerge so that the actions of one mod can be done more-or-less automatically by mods on other relays.

E-mail is federated and much less moderated than fediverse and even though spam exists it seems like a manageable signal to noise ratio. Anti-spam tools are pretty effective.

[–] Die4Ever 4 points 11 months ago (1 children)

A database of bans with tags for the ban reasons could work, like https://gui.fediseer.com/instances/censured

I wouldn't use email as a great example for spam filters, considering how much it caused Gmail to take a monopoly and how often it blocks emails you actually want. You practically can't self host email anymore.

[–] [email protected] 2 points 11 months ago

Agreed. The lack of self-hosting ability really comes down to the refusal of the wider web to upgrade away from SMTP. If you follow all the latest backwards-compatible protocols (DKMS etc) you can still get a decent outbound delivery rate.

There were many, many elegant solutions proposed to stop spam but none of them got implemented to avoid breaking backwards compatibility with SMTP. Then again, at this point, most people have moved away from e-mail to other forms of communication anyways due to e-mails problems (spam included). Unless the younger generation gets a sudden, renewed interest in e-mail, it will probably not really exist in another 50 years.

I do think blockchain will probably solve the issue of assigning senders a "spam score" universally once and for all instead of our current system which is a grab bag of blocklists plus each provider's secret sauce. Once you have a universal blocklist every e-mail provider can use and contribute to, it becomes easy to identify most senders as "safe" and new senders will just have to spend a bit of time earning their trust.