That is actually a great UX question. For federated services how do you think we best solve this? I mean everyone wants to just not pick one. But different loss suits people better in terms of location/interests/something else.
UX folks, what patterns would you suggest to help solve this?
Funny enough, choosing a node wasn't such a show-stopper for users ten years ago and before. I've used local email services in my backwater town, hanged out on local forums. The entirety of Fidonet was built on people connecting to their acquaintances—and it was rather popular here almost until the 2000s.
As for the present, I'm sorta baffled that the anarchic dream of decentralization turned into not-quite-that-cool federation where you're tied to an instance, instead of the identity working across the net. If my chosen server goes down, I'd love to be able to just move to another one with my keys.
That's easily done now with personal backups, and can be added onto any federated network that gains enough popularity. Mastadon is probably approaching that.
Email is a mature example of a federated system, and so if there is an answer to the general question, I'd expect we'll have figured it out in the context of email.
So... what's the "general strategy for picking an email provider?"
> So... what's the "general strategy for picking an email provider?"
Pick Gmail unless you have a good reason not to. And if you don't know any reasons pick Gmail.
Federation sounds nice until you realize that most people don't have particularly unique needs and a 'best for most people' will eventually emerge and dominate the market.
But instances aren't competing for users. There's no account fees, no ads, seems most big instances rely on voluntary donations to keep running.
You are also free to set up your own instance(s) and you can follow anyone.
Otherwise, stick to an instance with people like you (i.e. don't put your account on a gardening or knitting instance if you are primarily going to post and read about FPS gaming as it will mess up the instance timeline for everyone else and you'll have little use of the instance timeline.)
Until recently most people just took the provider that cane with their isp because they were usually tied together. It was either that or AOL.
Nowawadys it’s either hotmail/office365/whatever other name it has or gmail. It used to be a federated system and still sort of is, but your experience is vastly diminished if you aren’t using one of the biggest providers because you’ll have deliverability problems.
I've had these issues myself, and know the "stereotype comes from somewhere", but...does anyone work off of anything besides anecdata in discussion re: the stats behind this position? It's very defeatist.
I worked in email deliverability for years. It's absolutely true. Mainly because the big providers work together to make sure that their mail is successfully delivered to each other, and they also work with big corporations.
I was working on it when I worked for ebay and PayPal. I was working with Yahoo (who was big at the time) on what became DKIM, which was created to make sure that Yahoo mail was delivered successfully and that ebay and PayPal mail would be delivered successfully, among lots of other big corps.
Sure, anyone can use those technologies, but only the big players run interoperability tests on a regular basis, and provide reporting to each other, and other big corps.
Email is a different kettle of fish. It got in there first and only really competed with snail mail. Other alternatives were arguably more complicated. There was no easier alternative to fall back on. If you wanted to communicate on the internet, you had to use email, and by the time there were alternatives, it had a large enough user-base which meant that it was never simply going to be displaced.
I could be wrong here, but I believe that for most ordinary people, AOL messaging and even CompuServe boards were in common use before Email. The email "standard" had to displace those.
I never used CompuServe forums but my understanding is that they were public internet forums, not exactly something that could be used for personal/private messaging. Newsgroups (often used with an email client) pre-dated CompuServe forums by a decade and I'm pretty sure they had a much larger user-base.
AOL Instant Messenger didn't come out until 1997, email was already in heavy use by then.
Scroll a bit down and you'll see that you can filter through the servers based on your language and interests, as well as see the amount of users each server has. https://joinmastodon.org/
> For federated services how do you think we best solve this?
For distributed services, the solution must be such that the question "Which pod do I choose?" makes no sense. (There may not be a solution, in which case this may be distributed solutions' Achilles' Heel.)
Not the domain name, but the provider definitely makes a difference. What's the spam filtering like? What's the privacy policy? What's the sending reliability (are you sharing the platform with spammers?) What's the uptime like?
Unless the auth database is also distributed and you get to log in as yourself from any pod, it does matter.
That is actually a great UX question. For federated services how do you think we best solve this? I mean everyone wants to just not pick one. But different loss suits people better in terms of location/interests/something else.
UX folks, what patterns would you suggest to help solve this?