One Mammoth of a Job: An Interview with Eugen Rochko of Mastodon
Mastodon’s creator sheds light on maintaining one of the most popular projects in the fediverse
Eugen Rochko, known online by his handle Gargron, maintains and develops Mastodon full-time. His employment is entirely sustained through thousands of donations to an ongoing Patreon campaign.
Mastodon has exploded with popularity over the course of the past year, and has brought in nearly one million new people to the fediverse, an ad-hoc supernetwork comprised of hundreds of independent servers.
The project has accrued many volunteer developers, contributors, translators, and evangelists, who work tirelessly to make Mastodon better for everyone.
Thanks for joining today, Eugen! Tell us a little bit about yourself. How did you get involved in the federated web space?
When I was in high school, I had some online programmer friends who were into Identi.ca which led to me checking it out. I was also running a custom-coded forum at a time, and I remember trying to make a federated revamp of that, which is how I acquired the zeonfederated.com domain which remains my personal homepage to this day.
Of course at that time I didn’t have any success as I didn’t have the required skills to see it through. Even though I left the space for many years, becoming a heavy Twitter user, I was always sold on the idea a service like that ought not to be a commercial property.
What about decentralization / the federated web appeals to you?
I’m actually quite conservative when it comes to my routines, I hate switching apps and services, and I hate having friends on a dozen different chat applications, as was the case a couple years ago (nowadays it more or less became “everyone is on WhatsApp and Facebook”).
“At some point I was like, come on, how many more times do I have to go through this?! We need something that’ll work forever.”
I spent years advocating for XMPP to everyone I talked to. I’d much rather have an app that stays around as long as I want it to because I’m hosting it, and talk to anyone I want to without barriers.
As a teenager I was on MySpace, I was on Friendfeed, I was on SchülerVZ (German version of Facebook for students), I helped convince my schoolmates to switch to Facebook, I jumped on Twitter in 2008, I was on ICQ, AIM, gTalk, Viber, Telegram…
At some point I was like, come on, how many more times do I have to go through this?! We need something that’ll work forever.
Coordinating the development of a popular Open Source project often comes with a lot of challenges. What do you do to keep things organized and moving in the right direction?
I’m not a stellar example of being well-organized, I feel a lot more like Bob Ross, looking at the software and seeing that I can put a tree here, a cloud there. I use Mastodon every day, whenever I’m online, and so when I notice a rough edge, I go to fix it. There’s an overall “core idea” of Mastodon, too, and sometimes it feels like I’ve got the contours down and I’m just coloring it in.
When I show Mastodon to a stranger or friend for the first time, I listen to their feedback, or imagine approaching it from their perspective, and sometimes that leads me to discovering unpolished parts of it.
For example when I invited one of my artist friends, and she had these vertical paintings that got completely butchered by the thumbnails, that gave me the idea to add focal points to Mastodon. Oddly enough nobody had requested that feature before, even though there were other artists before her, too.
Feedback from the community plays a role, too. Mastodon receives a lot of feature requests. Sometimes they’re really fitting, sometimes they’re too niche. It involves me rejecting a lot, because Mastodon can’t be everything for everyone, and dealing with the disappointed people is one of the most unpleasant parts of the job.
At the very beginning I had no roadmap at all except the idea in my head of what it should be, which evolved as I was using Mastodon and seeing other people use it. But rather soon after Mastodon had its break-through, I started writing down a rough roadmap for a few months in advance so that people who were financial backers of the project would have some idea of where we’re at. Once that roadmap was done, I’d sit down and compose the next one.
Realistically, I had a lot of detours all the time, implementing cool things I hadn’t thought of in advance.
“I’m not a stellar example of being well-organized…”
Currently I’m working down a roadmap for the next minor version of Mastodon, and I have maybe 1–2 grand items beyond that, however, in terms of features I believe that Mastodon has reached its optimum a few months ago: everything I’m working on nowadays is either polish, bug fixes or general maintenance.
Do you have any thoughts or insights on Mastodon’s popularity?
I’m very glad that Mastodon (and by extension, the fediverse) is as popular as it is. When I started, I had not expected this. All I thought I would accomplish is conquer the niche of people who were already using GNU social by creating a more polished product, but instead Mastodon managed to attract new blood, even non-technical people; it completely overshot and overshadowed its predecessors and entered the fringes of mainstream perception.
Of course, the curse of ambition is that goals change as you succeed, and once Mastodon had its November 2016 success my goals shifted to actually rivaling Twitter for real. We’re making small dents in it, but we’re not where I’d want us to be yet. Network effects are getting stronger over time: 8 years ago, everyone had 20 social media profiles without batting an eye, but nowadays people are legitimately hesitant to use just one more platform in addition to Twitter and Instagram.
On top of that, a decentralized system presents unique discovery challenges. It’s one thing to get someone to try out Mastodon, but for them to keep using it, they need to acquire the happy medium of content and personal connections to have a reason to come back.
“People would use a social network based on smoke signals if everybody else was using it.”
You were instrumental in working with Chris Webber to build the first ActivityPub implementation used at scale. What was that experience like?
Scary. Being the first to implement a specification without any prior examples leaves you with a lot of choices about how to do things. I mean, sitting in on the meetings and brainstorming with Chris about how some part should be changed was cool, it’s really just the huge weight of the responsibility that’s terrifying.
In the end, I am happy with how everything turned out, and I think some of the feedback I’ve given was instrumental to ActivityPub becoming a practical protocol.
There have been a handful of clashes within the fediverse between different platforms, developers, and groups. As a project head, how do you handle infighting and hostility?
When it comes to developers, I think we’re all on friendly terms, be it GNU social, Pleroma, Peertube, Pixelfed or Misskey.
As for general conflicts within the project, I had to block some people from our GitHub repository for breaking the code of conduct a few times.
What does the future have in store for Mastodon?
The best things that could happen to Mastodon are social, not technical in nature. We need more people, more creators, who would in turn attract others to join; we need more organizations with recognized names hosting servers that people can trust.
“…we want more performance, we want improved account migrations, and we want to investigate better group communications on Mastodon.”
The past couple weeks I have re-focused my attention on public relations work, like running our social media accounts and writing blog posts, instead of pure development work. I also invested in an animated video that explains Mastodon a few months ago.
From a technical standpoint, Mastodon is better than its competitors, it’s a million times better than it was in April 2017 when it went viral for the first time. What we need is no longer more features, what we need is for people to know about it. But in terms of engineering, we want more performance, we want improved account migrations, and we want to investigate better group communications on Mastodon.
Out of all the lessons that you’ve had in developing Mastodon, what is one lesson that stood out to you the most?
You could spend an insane amount of time on engineering to bring out a new release with cool features and only get a lukewarm response. Then some celebrity will randomly plug Mastodon when you’re not expecting it and it’ll go viral. There is no way to control when success comes, only to lay a foundation for it.
However, features actually play a much lesser role than you would think. People would use a social network based on smoke signals if everybody else was using it.
What’s your overall impression of new ActivityPub implementations coming into the fediverse?
I think it’s starting to click for people how cool the idea of the fediverse is once they see what you can do with it. For many people, especially those without a technical background, they still have this commercial silo mindset.
But when you show them PeerTube and Plume, Pixelfed and Misskey, it becomes more apparent that they’re not just betting on Mastodon, that it’s really this big, future-proof endeavour that anyone can contribute to.
Is there anything you’d like to say to the people reading this interview?
Tell your friends about Mastodon. Create invite links and put them on your profiles on other websites. Word of mouth beats everything in terms of getting people on board, and getting people on board is good for all of us — as they say, a network is only as good as the amount of people you can talk to through it.
Thanks for reading this interview! If you’d like to know more about Mastodon, how to support it, or where to get involved, check out the links below.