I've been trying out Mastodon for the past 2 days. I like it, but it honestly has some problems that will prevent it from growing outside of niches. The main one being discoverability. There's simply no easy way to discover other users on the platform. And from what I understand, this is by design.
You can search hashtags, but not everyone uses hashtags. And the results that appear seem to be most recent, and not most liked or most boosted. It requires a lot of "work" to find interesting people and/or friends on the platform. And most people don't have time for that extra "work."
I'd like it more if I could see what's trending, but only if I specifically seek out what's trending.
Also on my Galaxy Tab S7 in portrait mode, a breakpoint is preventing the search field from appearing. I have to shrink Firefox a bit for the search field to appear. Took me a day before I figured that one out.
Outside of those issues, it has potential. I love everyone having a chronological feed.
> There's simply no easy way to discover other users on the platform. And from what I understand, this is by design.
It also feels fragmented. An account can be on any servers, anywhere, controlled by... someone? Users can move around. Communities can just shutdown or lock people out. Etc.
It's a weird solution to a problem that I agree exists.
It's definitely fragmented. Something like mastodon.social is probably perfectly fine for the general population though. And as users become more comfortable with the platform, they can transfer to other instances if they want.
I just don't see it growing because of lack of discoverability. I'm betting most new users will hop on, not know who to follow or what to do, and never sign in again.
You discover other users by slowly walking the graph from the people you know and you find the first users in the communities you like through hashtags. It's slow but you eventually build up a pretty good list of people to follow.
Also you generally join a server oriented around the subject(s) you like so you can check the local toot stream for people to follow.
There's a hidden misunderstanding here. Phrasing Mastodon as "the platform" implies you (probably subconscious) perceive it as a "classic" single, centralized, authoritative service managed by a single (legal) entity.
As you've learned: it's the exact opposite. It's not a service: it's just software.
Mastodon fits in between (self)hosted solutions like Matrix, Postfix, IRC, XMMP/Jabber. It would be the odd one out between Twitter, Slack or Discord.
> The main one being discoverability.
And that's a valid concern to have as a dealbreaker. It's also an advantage that's innate to centralized, authoritative services. Storing millions of users in a single telephone directory vastly lowers the bar for discovery compared to a distributed system of telephone directories.
Then again, it's also a chicken and egg discussion. Centralization is a key driver that prompted hundreds of millions of users to flock together. Aspects such as ease of use, convenience and discovery are key conditions as well...
... but let's not pretend this is a net positive: we also experience the many downsides of having millions of people on a single platform that provides common denominator interfaces and policies for governing interaction.
In a federated environment with many instances, agency over governance is delegated again to the countless of small communities and tribes that generally tend to emerge organically in any social context.
Remember, back in the early days, between BBS'es, Usenet, online fora, IRC communities,... the fragmented nature of online communities in itself wasn't seen as problematic. On the contrary, if anything, the freedom to be able to start your own IRC channel or host your own forum was seen as a boon.
In that regard, I don't think that discovery should be treated as a zero sum game where the only logical conclusion is a silver bullet solution e.g. either a single centralized authority or a distributed ledger that acts as a single source of truth. Discovery is massively complex and relies on innate heuristics people have used since forever to find and interact with others. It's never a "solved problem".
That doesn't mean throwing up arms not investing in discovery tools. But rather acknowledging that discovery isn't necessarily worth a trade off against other key aspects of complex interpersonal relationships such as having agency to build your own community space, and self-reliance to implement your own set of policies allowing for governance
The same handle can be used on multiple servers within a federation. Search for @NASA and you'll find at least 4 accounts under that name on different servers making discernment as to which to follow quite a challenge.
On federated sites, you have to think of domain names as part of the username. The account isn't just @NASA. It is @NASA@mastodon.social, for example. This isn't even a new concept. NASA@gmail.com and NASA@outlook.com can co-exist. We just have to get used to it outside of Email.
I'm not sure but who really has lack of random contacts?
Maybe those vibrant meeting places could/should be a click away from my profile. In a walled garden that means in the same garden, behind the same wall. But in a free world it can be somewhere else. And as Ronald Reagan said: tear down this wall!
Strongly agreed. Also: no lists to be easily followed?
I found some account which creates curated lists for specific interests, but then you have to go through the list one by one to click follow... surreal.
They've recently added this to the iOS app. You can browse posts, hashtags, news articles and users. It shows users and posts from people you haven't followed in a seperate style of timeline. Not sure how they figure out who to show.
I hope Elon opens up the Twitter API and lets fully-powered 3rd party clients flourish.
Ideally, moderation policies and enforcement actions can be client-specific to some degree. Ultra woke or righty twitter can implement their own policies without losing out on the network effect of the greater network.
Ben Thompson has been crusading for this direction and it’s gain traction in free-speech circles
Twitter already works that way, the block feature is available to everyone. It doesn't make any difference to any small groups or to the greater network, moderation from the platform itself still has to happen to prevent damage to it.
You can just ignore what's being said in "free speech circles", those people have no idea what they're talking about. If they were actually trying to solve these problems they wouldn't be pursuing a solution that we already know doesn't work.
> You can just ignore what’s being said in “free speech circles”
The irony here is hilarious. This is the exact point people arguing against social media censorship are pointing out: if you don’t like it and the speech is protected under law, use your little block button and ignore it (rather than asking a social media company to ban it because it hurts your feelings).
The problem is everyone has been burnt by this before. The moment you build a 3rd party solution, then they pull the API away from you 1-x years from now, your dead in the water. Thats not to say people wont give it a go, but I wonder how viable you can build an eco system onto of someone else's API?
The problem of third-party clients has always been more complicated than optimism of the will can solve IMO. Client specific moderation sounds like fodder for another set of messy congressional hearings. It can be done well of course. It would be interesting to hear the blockers the team sees internally and what is higher priority in their opinion.
Filter bubbles should just be a platform-wide thing applied to your experience. Everyone gets a mandatory set of filters for illegal things in their geo. Then you can add whatever you want and share the filter set with other like-minded people. “No conservatives” can be one filter pack, “no liberals” another. “No politics at all” could be yet another. Filters can be as simple as grep, could include ML stuff, or even have moderators.
So a new wave of people can see how unusuable it is. I guess I'm not shocked that Mastodon has improved so little since I joined 6 years ago.
- The very first instance I joined within a few weeks of launch stopped existing within a year
- Maintainers of instances got into weird internet fights and started banning entire instances
- DMs can be read by practically anyone
- in 2022, still zero discovery
- in 2022, fairly poor UX on a variety of web UIs, terrible UX on mobile clients
Mastodon really isn't worth your time unless you prioritize true decentralization. Unfortunately that appears to work against the concept of "community" and good software design.
Mastodon's onboarding UX is an absolute fail... I just didn't have the patience for it do I doubt the casual users won't either. Once I had an account it got even worse, so it's very unlikely I'll login ever again. If that's supposed to be a twitter alternative I now understand there's no competition
I've been using mastodon over the last few days and the difference between tech twitter and mastodon tech instances are dramatic. People engaging with my content has gone way up compared to twitter. I feel like unless you have a ton of followers on twitter, you get completely lost in the shuffle. Joining smaller instances increases the chances that your post will get noticed.
Also, joining tech focused instances means the content you are reading is almost exclusively about tech. Twitter has mechanisms to filter the content you want, but it feels totally different and a little forced.
For example, my sidebar has entertainment news articles on Twitter. I don't want to see that crap on my page.
"Tech Twitter" has been heading towards lowest-common-denominator content targeted at beginners and wantrepreneurs, for a few years now. Those audiences engage a lot more. And it's easier to throw in click bait content or low-density information threads, because the audience cannot yet tell the difference.
How does one find the bigger Mastadon instances? They aren't even listed on this joinmastodon website. I know it is intentionally decentralized, but it seems like it can never catch on because it feels like 100s of separate websites. You have to make logins for every single one.
You don't. You join one - or host your own - and your mastodon instance is connected to the others. You can follow users on any other public mastodon instance.
You have to find a host that accepts content for all other nodes you care about. TheFoss server probably isn't carrying anthing adult. So you do need to host your own
It is federated like email so if you have an email (@gmail, @yahoo, @custom.domain, etc) they are all interoperable. (Modulo moderation / spam filters)
You don't have nor would be able to create an email address in all public email servers. (Without mentioning private/non-public email servers or personal/family ones of others)
I login to my old Mastadon account on some specific server then try to reply to someone on another one. It asks me for the domain I want to act from so I type it in. Click Proceed to reply and am greeted with an error screen saying something went wrong. I really want an alternative to Twitter, but using this site feels very abrasive.
Someday a tech giant will mainstream Mastndon by recentralizing it, either by (1) running a large enough instance that its gravity sucks in everyone else, or by (2) creating a service/client on top of Mastodon that creates a unified Mastodoniverse. Decentralized services don't become mainstream without restoring the user conveniences of centralization.
I've been on Mastodon since ~2016, and actively use it on a daily basis. I've seen how it works in detail. I've had numerous exchanges with developers and instance administrators.
I think your concern is highly valid. Mastodon is a protocol-based network, much as SMTP email, NNTP Usenet, HTML+HTTP Web, IRC-based chat, and others are and have been. It's susceptible to the same types of challenges which have either centralised or killed off those platforms.
You mention sheer size. That's only one option.
There's the old Triple-E: Extend, Embrace, Extinguish. Applied successfully by Microsoft to DOS, the Office Suite, and Web browsers. By Google to Web Search and Web Browsers, to email, and increasingly to Web design and hosting. By Slack to Web Messaging. By Facebook to social media.
There's the risk of small instances fading away (happened to me with Mastodon.cafe) or large ones being bought (Mastodon.cloud). There's the prospect of admins going rogue (witches.town). Or, from the Diaspora* world, of admins dying with no succession plan in place (pluspora.com).
There are the Four Horsement of the Infopocalypse: drug-dealers, money-launderers, terrorists, and pedophiles. That's largely what killed off Usenet: the inability to defend against either malicious use of networks or those who made hay of that use to advocate for the shutting down of those networks. (This was combined with the lack of a compelling business argument for keeping them operating.)
There may simply be the staling of the protocol and paradigm. Mastodon is fun and interesting, yes, but lacks robust search, filtering, and privacy-management tools (e.g., auto-expiring posts), as well as obvious issues in scaling. If, say, a manageable instance size is 10,000 users, and Mastodon is to scale to serve an appreciable fraction of the roughly 10 billion people worldwide --- each instance administrator needs to at least consider 10^(10-5) or 10^6 other server relationships. It's one thing to keep track of reputations of a few thousand other servers, it's another to deal with millions, absent some robust tools for management, reputation, appeal, and similar factors. (My suspicion is that some degree of hierarchy with intermediate "hub" servers would emerge at some point.)
That said: Mastodon has proved robust and resilient beyond the predictions of numerous early critics. Not without flaws and issues, but those have to date been surmounted.
If that 30,000 is mostly the kind of people that like to be shouty about their worldview being so correct while gloating at the removal of voices they don't like or won't tolerate, then I'm sorry for your gain.
Curious to see if this holds or if there's a deficit of ~30,000 in three months time...
Imho mastodon's got an onboarding problem since before you even start you have to decide where to make an account, a decision with unclear ramifications. That'll scare a lot of people off at step 1.
Readit News
You can search hashtags, but not everyone uses hashtags. And the results that appear seem to be most recent, and not most liked or most boosted. It requires a lot of "work" to find interesting people and/or friends on the platform. And most people don't have time for that extra "work."
I'd like it more if I could see what's trending, but only if I specifically seek out what's trending.
Also on my Galaxy Tab S7 in portrait mode, a breakpoint is preventing the search field from appearing. I have to shrink Firefox a bit for the search field to appear. Took me a day before I figured that one out.
Outside of those issues, it has potential. I love everyone having a chronological feed.
It also feels fragmented. An account can be on any servers, anywhere, controlled by... someone? Users can move around. Communities can just shutdown or lock people out. Etc.
It's a weird solution to a problem that I agree exists.
Good!(?)
Someone who can read your DMs, even.
I just don't see it growing because of lack of discoverability. I'm betting most new users will hop on, not know who to follow or what to do, and never sign in again.
Just like now.
But: That someone might well be yourself! (shameless plug): https://seppo.app (German vaporware to date. Translate with e.g. deepl.com)
Also you generally join a server oriented around the subject(s) you like so you can check the local toot stream for people to follow.
There's a hidden misunderstanding here. Phrasing Mastodon as "the platform" implies you (probably subconscious) perceive it as a "classic" single, centralized, authoritative service managed by a single (legal) entity.
As you've learned: it's the exact opposite. It's not a service: it's just software.
Mastodon fits in between (self)hosted solutions like Matrix, Postfix, IRC, XMMP/Jabber. It would be the odd one out between Twitter, Slack or Discord.
> The main one being discoverability.
And that's a valid concern to have as a dealbreaker. It's also an advantage that's innate to centralized, authoritative services. Storing millions of users in a single telephone directory vastly lowers the bar for discovery compared to a distributed system of telephone directories.
Then again, it's also a chicken and egg discussion. Centralization is a key driver that prompted hundreds of millions of users to flock together. Aspects such as ease of use, convenience and discovery are key conditions as well...
... but let's not pretend this is a net positive: we also experience the many downsides of having millions of people on a single platform that provides common denominator interfaces and policies for governing interaction.
In a federated environment with many instances, agency over governance is delegated again to the countless of small communities and tribes that generally tend to emerge organically in any social context.
Remember, back in the early days, between BBS'es, Usenet, online fora, IRC communities,... the fragmented nature of online communities in itself wasn't seen as problematic. On the contrary, if anything, the freedom to be able to start your own IRC channel or host your own forum was seen as a boon.
In that regard, I don't think that discovery should be treated as a zero sum game where the only logical conclusion is a silver bullet solution e.g. either a single centralized authority or a distributed ledger that acts as a single source of truth. Discovery is massively complex and relies on innate heuristics people have used since forever to find and interact with others. It's never a "solved problem".
That doesn't mean throwing up arms not investing in discovery tools. But rather acknowledging that discovery isn't necessarily worth a trade off against other key aspects of complex interpersonal relationships such as having agency to build your own community space, and self-reliance to implement your own set of policies allowing for governance
I'm not sure but who really has lack of random contacts?
Maybe those vibrant meeting places could/should be a click away from my profile. In a walled garden that means in the same garden, behind the same wall. But in a free world it can be somewhere else. And as Ronald Reagan said: tear down this wall!
I found some account which creates curated lists for specific interests, but then you have to go through the list one by one to click follow... surreal.
Ideally, moderation policies and enforcement actions can be client-specific to some degree. Ultra woke or righty twitter can implement their own policies without losing out on the network effect of the greater network.
Ben Thompson has been crusading for this direction and it’s gain traction in free-speech circles
You can just ignore what's being said in "free speech circles", those people have no idea what they're talking about. If they were actually trying to solve these problems they wouldn't be pursuing a solution that we already know doesn't work.
The irony here is hilarious. This is the exact point people arguing against social media censorship are pointing out: if you don’t like it and the speech is protected under law, use your little block button and ignore it (rather than asking a social media company to ban it because it hurts your feelings).
I don’t believe Musk put those billions for the greater good… I won’t be surprised at all if Trump is welcomed sooner than later on Twitter again.
It's mainstream news outlets which amplified his every word.
They'll be doing the same thing if he runs for office again, no matter where he speaks.
- The very first instance I joined within a few weeks of launch stopped existing within a year
- Maintainers of instances got into weird internet fights and started banning entire instances
- DMs can be read by practically anyone
- in 2022, still zero discovery
- in 2022, fairly poor UX on a variety of web UIs, terrible UX on mobile clients
Mastodon really isn't worth your time unless you prioritize true decentralization. Unfortunately that appears to work against the concept of "community" and good software design.
Also, joining tech focused instances means the content you are reading is almost exclusively about tech. Twitter has mechanisms to filter the content you want, but it feels totally different and a little forced.
For example, my sidebar has entertainment news articles on Twitter. I don't want to see that crap on my page.
You don't. You join one - or host your own - and your mastodon instance is connected to the others. You can follow users on any other public mastodon instance.
You only have to make a login on one, and you can still talk to people on the others. It’s like email.
You don't have nor would be able to create an email address in all public email servers. (Without mentioning private/non-public email servers or personal/family ones of others)
One: https://joinmastodon.org/communities
Another: https://instances.social
I think your concern is highly valid. Mastodon is a protocol-based network, much as SMTP email, NNTP Usenet, HTML+HTTP Web, IRC-based chat, and others are and have been. It's susceptible to the same types of challenges which have either centralised or killed off those platforms.
You mention sheer size. That's only one option.
There's the old Triple-E: Extend, Embrace, Extinguish. Applied successfully by Microsoft to DOS, the Office Suite, and Web browsers. By Google to Web Search and Web Browsers, to email, and increasingly to Web design and hosting. By Slack to Web Messaging. By Facebook to social media.
There's the risk of small instances fading away (happened to me with Mastodon.cafe) or large ones being bought (Mastodon.cloud). There's the prospect of admins going rogue (witches.town). Or, from the Diaspora* world, of admins dying with no succession plan in place (pluspora.com).
There are the Four Horsement of the Infopocalypse: drug-dealers, money-launderers, terrorists, and pedophiles. That's largely what killed off Usenet: the inability to defend against either malicious use of networks or those who made hay of that use to advocate for the shutting down of those networks. (This was combined with the lack of a compelling business argument for keeping them operating.)
There may simply be the staling of the protocol and paradigm. Mastodon is fun and interesting, yes, but lacks robust search, filtering, and privacy-management tools (e.g., auto-expiring posts), as well as obvious issues in scaling. If, say, a manageable instance size is 10,000 users, and Mastodon is to scale to serve an appreciable fraction of the roughly 10 billion people worldwide --- each instance administrator needs to at least consider 10^(10-5) or 10^6 other server relationships. It's one thing to keep track of reputations of a few thousand other servers, it's another to deal with millions, absent some robust tools for management, reputation, appeal, and similar factors. (My suspicion is that some degree of hierarchy with intermediate "hub" servers would emerge at some point.)
That said: Mastodon has proved robust and resilient beyond the predictions of numerous early critics. Not without flaws and issues, but those have to date been surmounted.
Curious to see if this holds or if there's a deficit of ~30,000 in three months time...