Nathan Willis

We have 'Federation' ... we have 'Planets' ... where's the 'United'

Nathan Willis at

Following up on an earlier thought about Mastodon/identica/pumpio/pubhubsuburb/etc...

I suppose that I am less convinced than are others about the notion that 'federatable services' ought to be identical with 'everyone run your own server'. In truth, what I want from a communication platform is the ability to _choose_ plus the existence of some service provider that I can align with.

How many email providers are there? Statistics are a bit hard to come by, but this bundle of stats: suggests that there are eleventeen or a dozen that, collectively, account for the majority of the email sent every day.

That's fine; that's enough to ensure that they are interoperable, and for the majority of people who cannot and should not spend their time worrying about adminning their own MTA, they have choices. The choice threshold also means that there are niche players, such as the lavabit-like options for those with specific technical requirements.

I would greatly prefer it if the next iteration of federated microblogging would focus on standing up a good service and allowing their competitor to do the same. No significant advantage comes from there being 700 Mastodon services to choose from, and my gut feeling is that the initial "sign up for and get shuttled to for the rest of your life" deployment strategy hurt more than it helped.

It would have been better, in my estimation, to find a sustainable home for as a high-quality hub, and to allow other hubs to grow up on their own if their communities warranted them.

Now we have Mastodon, repeating that same initial-barrier-to-entry mistake: "want to use the service? First sit through a boring lecture about network protocols, then go look at a list of third-party services that we claim no responsibility for. Good luck, sucker".

And what's the gain from forcing new users through that cheese grater? DO they get a better service or experience at the other end of it? No, certainly not. Do they end up rolling the dice and choosing a service that they then are unhappy with hundreds of messages later, either because it goes down or changes its T&C? Maybe. And that's a minus.

Bottom line: If we had eleventeen good, interoperable microblogging services available on the web today, that'd be plenty. It's a shame that we don't have any large, freedom-centered organizations willing to run one — instead we have the debilitating 'choose-your-mom-and-pop-adventure' problem and a new protocol suite / software stack every five years.

But maybe that's just me.

Carol Chen, Mike Linksvayer, Charles Stanhope, Scott Sweeny and 1 others likes this.

Scott Sweeny, Stephen Michael Kellat shared this.

Show all 6 replies
Email is  a little different from social media since it is default private. You can all sorts of incompatible people with different values and interests using the same email provider, and none of them care because they are largely unaware of each other.

Social media is default public, and it seems we're struggling with how to best negotiate this space. It seems clear having a single, venture capital backed, host of a social network doesn't work. Heaps of abuse and other anti-social behavior abound, and users don't appear to have much influence beyond being the subjects of experiments. But it's not clear to me that only a dozen federated networks is the sweet spot. However, if it is, perhaps a larger set of federated networks is the appropriate starting point. They will get winnowed down from there? After all, isn't that how email progressed?

Charles Stanhope at 2017-10-16T00:29:11Z

Nathan Willis, zykotick9 likes this.

By default private, I meant my use of an email service is mostly invisible to others unless specifically addressed, and the contents of my email conversations are not visible even if you happen to know my email address. If you don't know my email address, you may not even be able to discover it unless I happen to leave it someplace public. You can use email in a public manner, but that's not the default, and generally requires deliberately addressing or using external services (unless communicating with a small group).

I didn't mean to quibble about the numbers too much. I'm just reflecting on what I've seen on social networks, and reflecting on past social networks like forums, chat rooms, and other things. It seems like there has always been a large number of them regardless of the technology used. Facebook and Twitter seem like anomalies.

Having said all that, I also don't mean to argue against the idea that the large number of places to join of dubious quality is an obstacle to adoption. I think it is, and you bring up important issues. These issues were certainly on my mind before I created another account.

Charles Stanhope at 2017-10-16T17:04:27Z

for services in which you don't want to run your own server, I'd rather it be serverless (P2P, like twister, or maybe like secure scuttlebutt) than having to depend on someone else's infrastructure.  if you're not going to keep your data to yourself, it's a lot better to have your account and data in a resilient distributed network than in a single node that's prone to fail and leave you out of service, more so if it's hard to migrate out of it once you started using the service

Alexandre Oliva at 2017-10-16T18:55:18Z

Nathan Willis, Charles Stanhope likes this.

Alexandre: very true. The back-end network plumbing is every bit as important as the front connection point.

Nathan Willis at 2017-10-17T11:02:15Z