Mastodon: The Wild Wild Neolithic West

Its been a long while since my last social media post: https://blog.mathoffthegrid.com/2017/08/how-i-use-twitter.html  and everything is all of a sudden in huge flux.  With all the turmoil on Twitter I’ve been exploring Mastodon on the math focused server a friend runs: mathstodon.xyz.   However, a new platform means starting over again

• You have to rebuild your network of follows and followers.  This is huge and discovering people has made my previous two attempts at using Mastodon unsatisfying.   But this time is different due to the chaos at Twitter. Large enough groups of people I know have migrated that I could start with a core group of folks and participate in enough conversations to have fun while finding new people.

Why things are vital this time - The extraordinary growth of the network

• Peculiarities.   Mastodon isn’t twitter and has a few quirks that you have to get used. No quote tweets due to a fear of harassment (which seems overly paranoid to me - since you’re just a screenshot away from the same effect)  and a general stance that makes discoverability harder.  You have to use hashtags since full text search doesn’t exist across the fediverse. And sometimes the other instance’s data is an extra hop away from your server . For instance, you’re surfing a profile on another server and sometime have to click to it to get full info.  Crucially, since each instance only stores the posts that users on it follows - there is a **deep effect on search **even for hashtags. You can only see what you know about or someone else on your instance knows about.  As a consequence, the larger the server grows the more useful it becomes if you’re interested in finding things . There’s also an overly precious stance on content warnings that doesn’t fit my theory of action. But I can live with that. 

• *Usability.  **Once you have enough people in your network Mastodon is quite usable despite the large discoverability issues. Can it be your only microblogging platform? That remains an open question for me.  The overall network is *much smaller than twitter. I don’t need most of the twitterverse though just the parts I read.  And on that front - the missing piece is probably government and media accounts.  For now there are bridge sites like birdsite.wilde.cloud that will publish tweets to toots.  But they aren’t quite realtime or completely reliable.  But the network is growing very rapidly (See above) so the situation could very easily shift in the upcoming months.

• Trust.  This is a huge general issue. I went with mathstodon because I knew the admins and could implicitly assume they would operate in good faith.  But how are millions of people going to make that leap?   I feel like there needs to be some type of  vetting process or change in structure  for many of the instances to answer this question.  Perhaps there will be multitudes of small sites where everyone personally knows someone involved - perhaps companies will enter the space and you’ll pay for some additional guarantees of stability/security?  This also remains to be seen and I expect new developments as growth continues.  But that’s the key as well - the network is growing despite this issue so its not an adoption blocker yet.

• Scalability.  In addition to the question of trust across large networks Mastodon is untested technically at twitter-like scales.  What’s going to happen when instead of 50k posts per hour there are 1M?  I could easily imagine a rewrite in its future off Ruby for instance. But even if you make technical fixes when everyone runs there own server you can’t easily distribute them. The same goes for scaling up infrastructure when your not running in a single companies datafarm.  Its a huge distributed process.

• Maintainability.   I don’t see how longer term, all the infrastructure can continue to be supplied by volunteers.  Keeping servers up and running at scale is real wor and  I’m not sure if enough new people are going to step up to provide sites for a 10x increase in users.  Practically, it all costs $$$ to run larger sites. Perhaps this is going to be solved with donations but that will be a long term issue. As will governance and long term succession plans.  How will each site organize to spread out serveradmin tasks, make sure if someone retires there is plan going forward, make crucial moderation decisions etc. 

But nevertheless, I’m having fun and experimenting.  I think people are going to tackle most of the issues I raise out of necessity.  So despite all the open questions I’m in for the ride for now and we’ll see where things go …   You can fine me @benleis@mathstodon.xyz