Are we decentralized yet?

Keep the needle pointing north. Towards the center of that dial.

Too decentralized, and you can't find anything. Nobody uses it.

Too centralized, and censorship takes over. Nobody can speak freely.

I don't disagree but I do wonder if a.) discoverability is really so intractable in a decentralized environment if you're willing to throw a lot of resources towards indexing and b.) if that middle ground isn't like balancing a pendulum upside down - a very fragile equilibrium. A bunch of decentralized units might join together, or a large centralized unit might fail, pushing the pendulum to either side.

You can think of the golden age of blogs and search as an example of both. Search engines formed a centralized hub with blogs, forums, etc. forming the spokes. For a while that worked well before it was degraded by spam and consolidation of disparate forums etc. into a handful of major platforms (fueled partly be acquisitions).

You're presupposing that discoverability requires some degree of centralization.

That's a good point. Thiel's "Zero to One" makes it.

In economics, a market needs several reasonably strong businesses to get price competition. An EU study indicated that the minimum number is about 4. Below 4, price competition seems to disappear and you have oligopoly, or, at 1, monopoly.

In areas where there's no inherent effect like distance to stop centralization, markets tend towards oligopoly. Look at the number of browsers, the number of big banks, the number of cellular phone companies, and so forth. They're all between 2 and 4. The stable state seems to be around 3 big players.

This probably applies to social networks. There's only so much attention available.

“Too decentralized” critics should form a non-profit where public hosts can register and build an opt-in index for all the decentralized content. Then there is no discoverability issue.

I wouldn’t be surprised if Facebook tries to eventually capture that data with Threads.

We haven't seen a distributed Google yet.

It's not impossible, but each distributed component would have to be at least a small data center.

Yes, I think that would be smart of them.

Google search does a lot more complex processing (crawling, historical weighting) and it does it on unstructured pages. Discoveravility in a microblogging system doesn't have to be a lot more than indexing/collation of structured posts, and end-user clients can be designed to participate in that. "Retweets"/"boosts" are already a form of that.

Sure, but fediverse numbers are pitiful at this point. Reality is that 99.9% of users don't care about decentralization, so it ends up being a "this has to work as well as a centralized system" does.

Informed Voluntary choice is what I want to see. Give me the option to pick, centralized, decentralized or hybrid balanced.

The fediverse folks are violently against any efforts for discoverability. They like the high bar for discovering and joining. Any attempt gets brigaded and shut down quickly

> We haven't seen a distributed Google

https://yacy.net

They're against opt-out discovery, not discovery in general.

Mastodon (the company) is building one: https://www.fediscovery.org/

> In economics, values below 100 are considered "Highly Competitive", below 1500 is "Unconcentrated", and above 2500 is considered "Highly Concentrated".

Fediverse is almost straight left, and it's already 690. Straight up would be 5000. This is non-linear scale presented linearly.

Some are. That doesn't mean they can do anything about it.

Remember, "the fediverse" is a bit like saying "the internet". "Internet folks are against centralization." Are they?

You can index your own stuff and propagate that index to others. This is what people would have thought good when Gopher was a thing.

Fediverse makes the indexing question interesting because some people don't want it deeply indexed: they point to the practice of dredging up old opinions on Twitter as an anti-pattern that the tooling should not support. Indexing without permission is met with hostility / defederation over there, and both individuals and server owners have tools to switch fine-grained indexing on and off.

(It is, of course, fundamentally impossible to keep people from indexing a default-open network, but if one does it, one does not advertise doing it outside the service-supported mechanisms).

Given how toxic big tent communities like Twitter can get, I think that makes perfect sense for some communities. Some plants thrive in full sun, some plants thrive in the shade. Some social interactions happen in the town square and some happen at more intimate functions.

You can have decentralized discovery via flooding. Everybody gets everything. That's how USENET worked. Torrents sort of work that way.

This hits scaling problems. USENET ran into that.