A tale of two prototypes

Join the discussion on infosec.exchange, lemmy.blahaj.zone, and Bluesky!

Mastodon, two years later discusses two ways of looking at Mastodon and its forks like Glitch and Hometown: a Twitter alternative, or a platform for networked communities. It's a long enough post that I decided to serialize it, and this installment continues that discussion.

Contents

• Intro
• Mastodon 2017 and Glitch 2017
• A BDFL gets to do what he wants
• Flash forward seven years ...
• Seven years later, is Mastodon significantly closer to being a good Twitter alternative?

"Fringe development is my term for the development work (research; design; coding) taking place on forks and individual instances of Mastodon, without the intention of sending that work upstream, and without mainstream acknowledgment or recognition. Increasingly, as Mastodon as a project grows, I believe that it is this development work that happens on the fringes that will shape the future of the software."

– Margeret Kibi, Fringe Mastodev: The Beginnings, 2018

It's not surprising that most people think of "Mastodon" as the official Mastodon project run by Mastodon founder and Benevolent Dictator for Life (BDFL) Eugen Rochko (aka Gargron), CEO of Mastodon gGmbH. Rochko gets almost all of the attention in the tech press, has more followers than any other techie on Mastodon, and is often portrayed as "the man who built Mastodon."

But the reality's a lot more complex.

Even though Rochko has done more work on Mastodon than anybody else over the years, most of the key innovations came from others. Rochko frequently hasn't recognized or acknowledged others for their contributions, and sometimes has treated them quite shabbily. And many of the Fediverse's communities are on instances running Mastodon forks (variants) like Glitch and Hometown, which include some key fringe development contributions that Rochko has blocked from the official Mastodon code.

Further complicating the narrative, most of fringe development work has been done by trans, queer, and non-binary people, whose vision for the software prioritizes community and safety – especially for marginalized people.⁶ By contrast, Rochko's vision of Mastodon is more as a Twitter alternative, and he has a track record of not prioritizing community and safety – especially for marginalized people.⁷ Of course, a Twitter alternative could also prioritize community and safety, especially for marginalized people; Rudy Fraser's three-part series Blacksky: Expressing the Black Everyday in a New Digital Space has some great perspectives on this. But that's not Rochko's vision.

Mastodon 2017 and Glitch 2017

From that perspective, it's interesting look back at Mastodon 2017 and Glitch 2017 as prototypes of two different visions.

• as a prototype of a platform to support networked communities, Glitch 2017 was a huge success. Local-only posts and other Glitch 2017 functionality have been adopted widely by other forks – and other Fediverse software. Some of Glitch instances that started back in the day are still around and doing well; many instances since then have also used (and improved) Glitch and related forks. Mastodon 2017 was less of a success, because it was missing some key features from Glitch, but it certainly wasn't a failure: some of the Mastodon instances that started back in the day are also still around and doing well.
• as a prototype of a Twitter alternative, Mastodon 2017 and Glitch 2017 were inconclusive. They had made a huge amount of progress in its first year, and even though (like most prototypes) they clearly wasn't ready for broad use yet, a lot of people already found them useful.
  
  On the other hand, Mastodon didn't have a plan to address other major challenges like usability and onboarding that had limited previous attempts at decentralized Twitter alternatives such as StatusNet and Gnu Social. Not only that, the Dogpiling, weaponized content warning discourse, and fig leaf for mundane white supremacy had driven away most Black users – and a Twitter-wannabe without Black Twitter isn't really a Twitter alternative.

Based on these initial results, Mastodon could have changed course and focused on networked communities ... but that wasn't Rochko's vision. And there were reasons for optimism that Mastodon could become a good Twitter alternative. Community innovation was in high gear. Mastodon was certainly more usable than previous decentralized Twitter alternatives, and the welcome modal was a big step forward on onboarding ... perhaps that progress would continue.⁸ So it's not surprising that Mastodon continued to focus on trying to become a Twitter alternative.⁹

Of course, Mastodon could have expanded its scope to also be a platform for networked communities, while still keeping its focus on being a Twitter alternative. It would have been easy enough for Mastodon to integrate the already-written code for local-only posts (just as Mastodon integrated some of Glitch's other improvements). Oh well.

A BDFL gets to do what he wants

"Eugen Rochko is the CEO of Mastodon — the open-source decentralized competitor to Twitter. Now, if you are like me, you hear the words “open source” and “decentralized” and then the word “CEO” and think, wait, why does the decentralized open standard have a CEO? The whole point is that no single person or company is in charge, right? Well, welcome to the wild world of open-source governance. It’s a riot, my friends. You’re going to hear me and Eugen say the phrase “benevolent dictator for life” in dead seriousness because that’s how a lot of these projects are run."

– Nilay Patel, Can Mastodon seize the moment from Twitter?, March 2023

Well, actually, nobody's the "CEO of Mastodon" – Mastodon is a complex ecosystem, and nobody's in charge. Still, Rochko's CEO and owner of Mastodon gGmbH, the German company that runs mastodon.social (by far the largest Mastodon instance), provides the official software release and mediocre official mobile apps, controls the joinmastodon.org landing page, and owns the Mastodon copyright. And also he's the BDFL. So he's the closest thing there is to a "CEO of Mastodon".

It's worth highlighting that not everybody agrees that a BDFL model is a good one.

"Diverse models of usership, shared between servers, range from venture-capital-backed alt-right platforms to Japanese imageboard systems, anarcho-communist collectives, political factions, live-coding algoravers, ‘safe spaces’ for sex workers, gardening forums, personal blogs, and self-hosting cooperatives.... [T]he rise of this new kind of usership meant a new questioning of the archetypal models of governance of FLOSS projects, such as the benevolent dictator."

– Aymeric Mansoux and Roel Roscam Abbing, Seven Theses on the Fediverse and the Becoming of F/LOSS (2020)

And not everybody agrees that Rochko's a good BDFL for Mastodon. But that's the thing about dictators: they don't have to care all that much about whether people want a dictatorship or whether people think they're a good dictator.

Even as BDFL and CEO of Mastodon gGmbH, Rochko's power isn't absolute – there have been times when loud enough pressure from users and instance admins has gotten him to change his mind about something. Still, as BDFL he can keep functionality that clashes with his vision out of the code base. In Patel's interview, Rochko points to this as a reason why he thinks the BDFL model is a good one for open-source projects:

"A good product needs long-term vision and it needs cohesive vision. That’s something that a committee cannot give. When you have a lot of people who have pet issues and one thing that they care about, it kind of ends up being a patchwork.... Sometimes, you need to make executive decisions about changing stuff in a serious way, which might not be popular with what most people in a committee would want."

As I wrote in Fork it! It’s time for a Mastodon hard fork (April 2024)

"So if Rochko thinks that functionality that lets people protect themselves against harassment (and is ideal for instances with local communities) shouldn't be available in the official software ... too bad for people whose pet issues are privacy and community. For people whose pet issue is accessibility, if Rochko says a design decision that make the software harder for use with low vision get as a "feature not a bug.", it's not available in the official software.... And these are just a few examples."

One more example: the ability for instance admins to easily change the maximum post length is another good example. After cassolotl suggested this in 2017, it was quickly implemented by lambadalamda and integrated into glitch-soc. But Rochko refused to support it in vanilla Mastodon:

"I feel like the UX is designed, visually and behaviourally, around a certain number of characters, and deviations have a negative impact on it."

Since then, similar changes been proposed – and rejected – repeatedly: in 2018, 2019, 2022, 2023, and quite possibly others I missed in my quick search of Mastodon issues. So on instances where people don't think 500 characters are enough, instance admin whose pet issue is trying to give people what they want are out of luck – unless they've got the skills, time, energy, and desire to install the Mastodon server software, change the code, reapply the changes every time there's a new release, sysadmin the Linux machine and database, and learn about Sidekiq queues.

Oh well. It is what it is. If Rochko doesn't care whether Mastodon's a good platform for networked communities and instead wants to keep focusing on Mastodon as a Twitter alternative ... well, he's the BDFL, he gets to do what he wants.

Flash forward seven years ...

"Thankfully (for Mastodon, and myself), even as I left the stage, others pushed forward with the fringe Mastodev agenda.  Most notable among these is Darius Kazemi, who has developed and expanded upon my original thesis in an encouraging (and ultimately ((although to what extent being a well‐known and male‐passing internet personality and (now ex–)Mozilla fellow has helped him in this is anyoneʼs guess)) more effective) way through his guide, Run your own social, and accompanying fork, Hometown."

– Margaret Kibi, Progressing Fringe Mastodev (2019)

Glitch, Hometown and other forks have continued to improve as a platform for networked communities – despite only getting a tiny sliver of the resources that have gone to Mastodon gGmbH.¹⁰ Today, the experience on a well-run instance running a fork with local-only posts, formatted posts, and a longer maximum post length than vanilla Mastodon is as good as anything on the Fediverse (at least for people using apps that are better than the mediocre official Mastodon versions). Some forks even have quote-boosts!

Of course even for instances using forks, there's still the endemic whiteness and sexism of Mastodon and the broader Fediverse to deal with (and don't worry, I'll discuss that in an upcoming installment)¹¹. Still, local-only posts along with a good instance blocklist provide some insulation, and allow-list federation can provide even more. And as I'll discuss later in the post, a networked communities approach could provide a path to make progress on these long-standing problems as well.

Mastodon has adopted many Glitch and Hometown improvements; in fact, earlier this year, Emelia Smith reported that Glitch maintainer Claire is now leading Mastodon development. Mastodon also benefits hugely from a handful of experienced volunteers (including Smith) who provide valuable contributions release after release – as well as dozens of people who have made more occasional (but still valuable!) contributions.

But Mastodon still hasn't adopted local-only posts. As Kazemi says on the Hometown wiki

"Being able to have conversations with people on your server that don't federate is a hugely liberating thing. It allows inside jokes to develop. It allows people the freedom to complain about things that they wouldn't necessarily feel comfortable leaving a trusted server (cops, employers, etc). It also lets us do things like have a server-wide movie night where we flood the local timeline with posts about the movie, and it doesn't pollute the rest of the Fediverse."

So Mastodon's still not anywhere near as good a platform for networked communities as forks like Glitch and Hometown.

Seven years later, is Mastodon significantly closer to being a good Twitter alternative?

And despite all the improvements over the last seven years, at least to me it's not clear that Mastodon has made all significant progress addressing its challenges as a Twitter alternative.

• From a safety perspective, Mastodon still lacks basic (and vitally important) Twitter functionality like private profiles and reply controls
• No matter how many times Mastodon loyalists say "it's easy, just like email" ... it's not easy.
• Onboarding, usability, and discoverabiilty are still huge problems as well. The challenges Erin Kissane discusses in July 2023's Mastodon Is Easy and Fun Except When It Isn’t are all challenges that were broadly discussed back in 2017.
• So is anti-Blackness. And even though his position has now evolved, Rochko's 2018 perspective that quoting "inevitably adds toxicity to people's behaviours"¹² means that Mastodon still doesn't support functionality André Brock notes in Distributed Blackness is a core aspect of Black digital practices).¹³

And while Mastodon's adoption of the ActivityPub protocol in September 2017 has had some benefits,¹⁴ ActivityPub's approach to federation where people only see a portion of the global conversation is very different from Twitter's flat model.¹⁵

Of course, these are all addressable problems. Still, at this point, Bluesky's a better decentralized Twitter alternative than Mastodon; Threads is much easier for people who already have Instagram accounts to try out. Both have much bigger budgets than Mastodon, so it'll be a challenge to catch up.

Fortunately, there's a lot more to social networking than Twitter alternatives.

To be continued!

The next installment, A faux “Eternal September” turns into flatness, looks in more detail at the challengers that newcomers in Mastodon's November 2022 wave faced. Here's the contents:

• It's not "just like email"
• Usability and gatekeeping weren't the only challenges newcomers faced
• The first complicated high-stakes decision is even before you sign up
• Why not help people choose an instance that's a good fit?
• But no

And there's more to come after that! To see new installments as they're published, subscribe to the Nexus of Privacy newsletter, follow @nexusofprivacy@infosec.exchange (or on Bluesky, via Bridgy Fed), and/or join the Nexus of Privacy community on lemmy.blahaj.zone.

Or just come back and visit Mastodon, two years later in a few days!

Sneak previews of upcoming installments

As always, these are from draft versions of upcoming posts, and the final versions may be different.

....

Of course it's over-simplifying to look at the history in terms of just two prototypes. Mastodon and its forks have also prototyped

• A trans-led fork with a trans, queer, and non-binary-centric community
• A consent-based culture that (while imperfect and intermittent) points to a very different path than surveillance capitalism
• A real-life testbed for all the complexities of decentralized moderation and federated diplomacy
• A protocol-based platform for (somewhat) interoperaable social media operations

....

I talked a lot about Mastodon's long-standing problems with race in Mastodon: a partial history, and things haven't improved since then – 5 things white people can do to start making the fediverse less toxic for Black people has plenty of newer examples and links. And as Dr. Johnathan Flowers points out in The Whiteness of Mastodon, Mastodon is a metonym for the whiteness of the broader Fediverse....

And while the original partial history talked a lot about anti-Blackness and the complicated interactions between Mastodon and LGBTQIA2S+ people, I didn't really delve into sexism and misogyny – which are also big issues. And here too, Mastodon is a metonym for the broader Fediverse.

....

One or more hard forks still seems like the mostly likely outcome to me, perhaps combining a focus on networked communities (and a less friendly attitude towards Meta) with some of the ideas in Fork it! It’s time for a Mastodon hard fork and/or Ro's A Case for Community (both April 2024). Hard forks that start with Glitch or Hometown as a base would have significant advantages over Mastodon if it can get support from hosting companies, and if it also improves allow-list federation and adds support for federations of instances it could be a good fit for The Website League and other emerging projects. Then again, there's been discussion of Mastodon hard forks for years, and it's not as easy as it sounds, so maybe it won't happen.

Whether or not there's a hard fork, the same opportunities are there for the official Mastodon project if its direction and leadership structure evolves. The official project has some big advantages – including funding – so let's hope that happens. And it's not an either-or situation; forks don't have to be hostile, and there are lots of potential synergies if they can work together effectively (as well as challenges).

Notes

⁶ Rochko's far from the only fediverse influencer who doesn't prioritize community and safety, especially for marginalized people. The (annotated) case for “Big Fedi” (December 2023) looks at cluster of ideas that Evan Prodromou characterizes as "Big Fedi". Prodromou, who describes himself as "the father of the Fediverse", says that he's mostly as a Big Fedi person", and these perspectives are in my experience the norm among cis male ActivityPub Fediverse influencers.

⁷ Fork it! It’s time for a Mastodon hard fork (April 2024) goes into detail on Rochko not prioritizing safety, especially for marginalized people. The low-hanging fruit discussed in Safety is an especially good area to focus on is all stuff that Rochko could have implemented in Mastodon, but hasn't. And mastodon.social, run by Rochko's Mastodon gGmbH, has a well-deserved reputation for relatively lax moderation against transphobia, Islamophobia, and racism.

⁸ Although, as Ana Valens describes Mastodon is crumbling — and many blame its creator (2019), Rochko had harshly criticized Shel Raphen's original design of the welcome modal and made significant changes that "broke the pedagogy and curriculum built into the design of the onboarding modals," which had been reviewed and approved by various other contributors. So the impact was less than it could have been – and it was a breaking point for the queer community that Mastodon has never really recovered from, so destroyed the momentum of community innovation as well. It's not surprising Rochko didn't see it that way, but as Lessons (so far) from Mastodon (2017) highlights, it really was obvious at the time.

⁹

Back in April 2017, Rochko said

"Twitter changed the reply system, which everybody told them they shouldn’t do, and then removed the iconic egg avatar for new users, and suddenly all of my work of telling people that one day Twitter would do something they didn’t like and they’d need a viable alternative paid off."

In April 2023's Can Mastodon seize the moment from Twitter?, Rochko noted

"I was a pretty heavy Twitter user back then — I think I started using it in 2008 or so, when I was a teenager — and it quickly became a very important part of my life for talking to friends or finding out what was happening in the world. Around 2016, I felt fed up with how Twitter was being run as a company, where it was heading, the community that was on there, the harassment, and so on. I started looking into alternatives, and after viewing the landscape, I decided to build a product of my own — and tried to make it good."

So more accurately, Rochko's vision of Mastodon is an alternative to Twitter as he experienced it from 2008-2016. Which means it's also not surprising that Rochko didn't realize that a Twitter-wannabe without Black Twitter isn't really a Twitter alternative. I put this in a footnote to emphasize the point that whiteness – on Mastodon and in the broader Fediverse – often manifests itself by not even considering Black perspectives. As Dr. Johnathan Flowers says in The Whiteness of Mastodon (2022)

"Mastodon is a very white platform, and insofar as I'm saying it's a very white platform, here I'm drawing on the work of Sara Ahmed in her piece A Phenomenology of Whiteness, who argues that spaces inherit the orientations of the people within them. So if you have a space that is predominantly populated by white persons regardless of their other identities, if you are in a space primarily populated by white persons, the norms, the habits, the very structure of that space will take on a likeness to whiteness by virtue of how the majority of people participate in that space."

¹⁰ It's not like Mastodon itself has ever been flush with cash; as Emelia Smith points out, "Mastodon has usually only had one developer, maybe two, working on it full-time at any given point over the past 8 years." But forks typically have almost no cash, and rarely have even one full-time developer for more than a few months. And Mastodon gGmbH also has contract developers working on mobile apps, hired a design consultancy, and runs mastodon.social (by far the largest Mastodon instance) and mastodon.online ... so it has more resources than just "one or two" developers. One way of looking at this is that Mastodon gGmbH does a lot with relatively minimal resources. Another way of looking at it is that Mastodon gGmbH has gone down a path of trying to do a lot more than it has resources for. There are other mobile apps and web interfaces out there that are better than the official ones; there are other instances that are better than mastodon.social and mastodon.online. Newer projects like GoToSocial are focusing much more narrowly, not doing their own UI and avoiding a flagship instance. It would be a big change but Mastodon could also go that route and devote more of its limited resources to where they're most impactful.

¹¹ And in the footnotes of this one!

¹² Reality begs to differ; see Hilda Bastian's Quote Tweeting: Over 30 Studies Dispel Some Myths (January 2023) for an excellent and very readable research survey. Also, as Black Twitter, quoting, and white views of toxicity on Mastodon discusses, there are plenty of positive uses of quote tweets. So Rochko's 2018 position on quoting was doubly incorrect.

¹³ So Rochko's 2018 position on quoting was racist as well as doubly incorrect. As I say, Rochko's views have evolved, and that's good, but as Black Twitter, quoting, and white views of toxicity on Mastodon also discusses, the discourse around quote boosts reinforced Mastodon's image for racism – and still does to some extent. As futurebird said on GitHub in December 2022

"I don't really think mastodon needs quote-boost to thrive. It would be better, helpful, positive, but users are creative and will work around it. We will find new ways to communicate. What I'm concerned about now is what this restriction has come to represent.

Whose concerns matter?
Who is welcomed?
Who is supported?
Who isn't?"

¹⁴ Thanks to ActivityPub, Mastodon's now at least somewhat compatible with dozens of other federated software platforms, which is good thing all around. Also, since Mastodon's historically been the big elephant in the Fediverse, other implementations have to conform to its often-quirky interpretations of the loosely written standard, and Mastdon's proprietary API has much more adoption than the ActivityPub C2S standard. If you're thinking that sounds like a classic embrace-and-extend strategy, you're right, and along with ActivityPub's complexity (a barrier to new entrants) it helped Mastodon keep its dominant position for several years despite its relative lack of innovation. Of course, while Mastodon has benefited, this embrace-and-extend has held the Fediverse as a whole back ... once again, oh well.

¹⁵ Mastodon's upcoming Fediverse Discovery Providers project, funded by NLNet, is an interesting attempt to work around this limitation, and it'll be interesting to see whether it uses ActivityPub ... but as I write this, it's still in the very early stages.