Hacker News Reader

My ideal UI would be having multiple accounts on different servers and a client with a slack/discord like sidebar to quickly and easily switch between them.

Then I could have one profile for coding, one for gaming, one for politics. The feed for each would only be about that topic so I can delve into what I feel like at that moment. The other big benefit is that I know everyone following me is following for that topic. On Twitter I mostly talk about tech and when I post about politics it may be uninteresting to 80% of my followers. Because one single account is supposed to represent me they have to put up with that and it creates a worse user experience.

The advantage of anonymous virtual accounts is we can have as many of them as we like, we should embrace that and build a better social network because of it.

There are a clients that make this convenient. Toot! on iOS has one of the more delightful UI innovations I've seen recently in the form of a wheel in the corner that you can rotate to flip between what looks like different copies of the app, each individually themed and associated with a different account.

what would be cool would be some sort of agent that could live on a server and process toots and take a guess as to whether or not you should or shouldn't post this toot on this server. Manually overridable, of course.

Try out Husky, it meets your listed requirements.

Husky - https://f-droid.org/packages/su.xash.husky

> So which one do I choose? I cannot even explore a server’s users and posts from the community list.

You can, right from the homepage of any instance.

You have "discover users" button, which will show you any local user that wants to be there (by default you're not visible here, you need to opt-in in the settings).

You also have "see what's happening", which will show you latest public posts from an instance.

Using Fosstodon purely as an example: https://fosstodon.org/explore and https://fosstodon.org/public .

> And then I created an account in the wrong server, and then what? Can I migrate my account to another server?

...kinda. You can "archive" your account and redirect your visitors to the new one. You can't migrate who's following you and who you're following.

Also there's nothing stopping you from having an account on multiple instances, but that's mostly unnecessary since you can follow people regardless of what server they happen to be using.

https://docs.joinmastodon.org/user/moving/

> Moving your account is the same as redirecting your account, but it will also irreversibly force everyone to unfollow your current account and follow your new account, if their software supports the Move activity. Your toots will not be moved, due to technical limitations.

I joined mastodon.social like everybody else years ago when it started. Lurked around and discovered and followed several accounts from various interesting communities. After a few months/year you can pretty much know the feel of the community by following enough people from it.

Recently I was invited into one of those communities and I just used the migration feature in mastodon. Went pretty smoothly. I guess it went easier for me as I didn't mind "losing" my toots in this new era. For those who need/want to keep everything it maybe problematic.

My motto on modern web is: don't get too attached to things you don't own. You can't own a community so you shouldn't worry about losing it.

> Can I migrate my account to another server?

There is a github issue discussing using a custom domain and pointing DNS records to whichever instance you like but it has been dormant for 2 years

https://github.com/tootsuite/mastodon/issues/2668

This kind of solution doesn't solve the problem for laypeople without the technical acumen to buy+manage domains. Which should be a concern for all the Mastodon/ActivityPub supporters who want to see federated solutions overtake centralized ones.

Additionally, DNS-based identity migration doesn't address the grandparent's real problem, which is that they feel expected to choose just a single community to live in in the first place. They want to live at the intersection or the union of multiple interests simultaneously, which doesn't fit the Mastodon model of social interaction (but does fit e.g., Twitter's).

Account migration for non-technical users _is_ being worked on in fits and starts, afaik. It's just a complicated problem with a lot of potential rough edges.

It's finished for mastodon 3.0 https://docs.joinmastodon.org/user/moving/

If it's open, a market opportunity can popup to allow technically adept folks to pay a reasonable fee to manage it for them?

$20/yr/user ought to cover DNS, hosting, and operations for a 1-2 person team as a part time project? Esp once you get past 50 subscribers?

I keep forgetting about Fiverr. There’s a way to dip toes in. Could make a .com landing page that pushes interested parties to the Fiverr job page.

https://docs.joinmastodon.org/user/moving/

The current answer is not as flexible, but requires no technical ability.

"I use Linux and I like FOSS and I am still an oldschool BSD nerd, and I also like regional communities." FOSStodon is the obvious choice https://fosstodon.org/public But I think the article actually misses the main point about Mastodon. You an follow and interact with people on any instances.

But then why is it so important to choose the right server? The author asserts that choosing the right server is essential to your experience of Mastadon, is that not the case?

Also the author says that I should choose a community between 500 and 5000 people, which fosstodon exceeds.

All in all, my point is, there are so many choices and decisions to make before I can even create an account and experience it.

It’s fairly common knowledge that people actually don’t like choice. When you go to a restaurant and see a menu with 100 choices, people actually enjoy their meal less than when there would be little choice; they have this nagging feeling of “was that other choice perhaps better?”

For me, Mastadon has the same problem. Too much choice is a bug, not a feature.

I think it's easier to discover people on the same server as you, that's all.

There's a feed for the server you're on. So if you pick a "bad" server (aka the culture there is different from what you like), that becomes useless, more or less. Of course you can always just follow whoever and only pay attention to your following feed.

Also, some nodes maintain blocklists of other nodes. [0] So while you can still follow those people and see them in your following feed, posts from users on that node will -not- appear in the public feed (posts from nodes in the federated network).

https://instances.social/ exists to help you find a node but I'm not sure how up to date or accurate it is.

[0]: https://github.com/Gargron/mastodon.social-misc

This is not the case.

As Mastadon increases in popularity, how do they plan to make the choice of server easier? If 5000 remains a rough optimal size, how will then how will the next 10M users choose from 2000 servers?

Probably my having a few (or one?) very large servers where most people join.

As much as I wished that wasn't the case, most people just want to join an established network and don't care about federation as a concept. A long as a single server doesn't take over altogether, however, I don't think there is a problem with large servers.

Surely the problem with large servers is that they're difficult to moderate, difficult to grow a sense of community, and all the other problems that lead the OP to suggest that 5000 is a vague upper limit.

If you are going to go with bigger servers, what is Mastadon doing to support them (socially, not technically)?

I think really huge servers will need a different moderation system. It wouldn't be the responsibility of the Mastodon team to manage or even design that.

Also, mastodon.social has about half a million users, and they are having performance problems. I'm guessing if you want to scale that up by a factor of 10 or 100 you're going to need new infrastructure.

Surely the Mastodon team are very concerned with moderation? Given that everyone agrees it is vital for the success of the project (or any social network), why would they consider it outside their purview?

I meant that they wouldn't consider the difficulties of moderating a multi-million user instance. I believe they have enough to deal with when it comes to moderating their half-million user instance. Also I think a large number of those users have either stopped using it, or moved to different instances.

They will need a way to migrate users between servers then. If your local feed is important, it's not a great plan to have users start on a suboptimal server and then have to make a new account on the one they want. If that's the case, I would generally expect to be able to migrate to a new server and have all my old posts, etc come with me, and that my user page on the old server would redirect to my new one.

There is a way to migrate servers. People have mentioned it in other comments.

In short, it marks your old account as migrated, and messages is sent to your followers instances, asking them to subscribe to your new feeds instead. For the most part it's pretty seamless.

I will admit, during my foray into Mastodon I tried a server at the low end of the author's size range, and I would say selecting the 'right server' has some impact from my experiences, but not in the way described:

A month after I joined, they had a disk crash and my account was gone. It looked like everyone recreated from scratch, so I did the same.

Two weeks later and without notice they restored a backup, overwriting everything that had happened since. I continued, but with rapidly declining interest to accompany my total loss of trust.

Two more months later the server was down permanently.

> It’s fairly common knowledge that people actually don’t like choice.

This is a misunderstanding of what people like.

What people like is good defaults, because choosing is work. When 90% of everything is crap, nobody even wants to see anything but the 10% that isn't. They don't even want to see anything other than the one option which is the best for them, if somebody can give them that without making them do the work.

The problem is that the best option is not the same for everybody. Even if it's the same for 70% of people, that lets you make that option the default and make 70% of people happy, but the other 30% still need the option to do something else.

And sometimes there isn't any one option that can satisfy any majority of people. Then there isn't any better alternative than giving people a list of options to choose from.

But you could still put the list in a better order. Put the "editor's choice" options near the top and the crummiest ones near the bottom. If someone is in Queens, show them the regional server for New York City on the first page and the regional server for Los Angeles on the hundredth page, and a link next to it that says "show other regions" and leads to a regional map. That sort of thing.

Remember forums? Remember how you had to have a separate account on every single one? Remember how you stuck around some forums, but not others? Treat it like that.

But it's not like that, you don't need to create a separate account on every single one. You can follow any account on any server/community from just one account. That's the big difference/selling point. It's the "local timeline", an optional discovery feature, that adds the incentive to create more accounts; but there are definitely ways to circumvent it. Many communities expose their timeline publicly and there's almost always a public profile directory so you can see who's on that server/community to follow them.

Forums had a public discovery and consumption mechanism. You didn't, for the vast majority, need to create an account to see what the community was like.

I mean that's why reddit is winning.

"Getting bored" is a feature, not a bug. Also a side effect of not having dozens or hundreds of people who have KPIs to optimize for "engagement".

Yes, but also no.

You can get bored because you aren't being fed a curated stream of engagement-boosting content; but you can also get bored you don't find an interesting community.

Mastodon has no unit of community smaller than "the local timeline", and it's really not cheap / easy to run your own (due to very high RAM requirements).

By way of comparison, 'one subreddit' is a group you can follow and post to from your main account. You can look at a single accounts post history, but the common case is to follow moderated interest groups. There aren't good ways to participate in interest-specific communities on mastodon (I'm not just whinging; I've tried to improve the situation but it's genuinely difficult at an ecosystem level).

Pleroma runs fine on a $50/year vps fwiw.

And an hour or two to stand it up, and a domain name, and an email host since you have to respond to abuse@.

A new subreddit is two minutes and there is no reason why a federated system has to be markedly harder.

An instance is not a replacement for groups, and it's most definitely not similar to a Reddit community.

What you want is, well, "groups" which is a feature that GNU Social had, and which is eventually going to come to Mastodon and Pleroma.

A lot of people seem to happily use Twitter which also doesn't have a concept of groups.

I wish Plethora had a solid docker-compose available officially because the current installation method is not great.

Try out Husky, it meets your listed requirements.

Husky - https://f-droid.org/packages/su.xash.husky