Back in the summer of 2011 I had a bit a tricky conundrum. I needed to get from Bow in East London to Camden in North London avoiding the riots that were happening at that time. And ideally I needed to make the trip by bike. And a community had sprung up on Twitter around the tag #UKriots where all kinds of people were discussing the issue. “Cycle along the canal towpath – that’s fine” was the gist of the answer from complete strangers brought together by this online discussion. I did, it worked, I thanked the people, and now 12 years on I have no idea who those people were (and indeed because Twitter’s search is now so broken I cannot find the replies either)

I mention this story because I was reminded of it through this toot by Osma A on Mastodon. Osma writes: “The reason Twitter was so interesting was that by some happenstance combined with competitive pressure from Facebook, they managed to come up with “Interest Graph”, combining both [content from topics and people]. It wasn’t perfect, but it was unique.”

I think this pretty much nails my problems with all of my efforts to find substitutes for Twitter since stopping using it because of Musk’s takeover.

I can still find my communities of people on Mastodon (people I might not know offline, but who repeatedly write about the things that interest me – so I am fine to read more or less anything they write), on LinkedIn (people I know offline, and have some professional connection with), and, when I can bare it once in a while, on Facebook (mostly friends who are not on either LinkedIn or Mastodon).

The problem is that I am missing the topic bit. The thing that is important right now, and I could do with input from people outside of my normal networks to help me understand it. The sort of people who could help me solve the #UKriots conundrum.

Mastodon – because it shuns network-wide search and trying to find things across instances is damned complex – does not manage to fully fill the niche. I get why some people might not want their toots to be searchable, but it ought not be necessary to use an external service like tootfinder.ch to even make it half work. People who want their toots searchable and those who don’t need to be able to co-exist on a network – if we are to learn from what did and did not work on Twitter.

The same applies to a slew of other functions. I get that quote-tweets were sometimes used for take downs of others on Twitter, but they had their uses as well for swift context setting. A Mastodon that allows a user to set if they are OK with quote-toots of their toots would make sense. Same with threads – my #CrossBorderRail project was a story that developed over a day, and days into weeks – and Twitter threads made sense for that – and so toot-threads that can be enabled for individual users would be a good idea.

Users’ approaches to moderation also vary widely – and any centralised network like Facebook or Twitter is going to leave a lot of people unhappy, based on their political views. Here Mastodon – with its decentralised system of instances that all have their own moderators and different content rules – is probably doing as well as any network. Most content moderation is about rules that are non-legal – it is about making your community feel at ease, a network feel welcoming. Simply limiting what is illegal is not an adequate answer as I see it, but again others may take a different view.

And then we come to the thorny issue of algorithms. Sometimes I want to see everything, and have it all listed chronologically, but – once in a while – an algorithm that pre-sorts things for me, but that I know how it is sorting content for me, would make sense. Here at the moment I am stuck between extremes – nothing to help filter for me on Mastodon, but algorithms so all pervasive and impossible to understand on LinkedIn and Facebook that I am left annoyed by having my feeds so shaped by these corporates’ decisions. Bluesky’s idea to try to be able to apply different algorithms to feeds looks like an interesting route.

All of this then brings us to the degree of commercialism of a network (or not). To sustain the tech costs of a tool, advertising might be necessary – and an algorithm using data from a user’s feed might be a way to make those ads relevant. But a user needs an alternative to that – to be able to finance a network themselves through a subscription, or to use the source code to run their own instance. Ad-funded, subscription-funded and self-hosted services would ideally co-exist on the same network, and this would then also help avoid the Musk single point of failure problem of Twitter.

To sum all this up, I am looking for a network where it is possible to interact with people I both do and do not know offline, and to engage with topics as well as people on an ad hoc basis. I would like most content to be searchable (and I am fine for my content to be indexed), a network where quote-posts and threads are allowed (I’d have those on, but users could turn them off), a place where moderation is quite tough to maintain a civilised debate environment (but others may want it tougher or less tough than me – and could control that), a choice of different algorithms to filter content are available but can be switched off, and a tool where I can pay a subscription to use the network or run my own instance (although I am fine if others want an ad-supported option instead).

Importantly you reading this might have a different preference on each of those points, and that is fine. If we are to have really learned anything from Twitter’s demise we should be able to accommodate those differences of view and differences of technical function within the same network.

Mastodon offers some of what I am searching for. Bluesky has some promise, especially on the algorithm point. Something else built on top of ActivityPub could maybe offer some more of what I am seeking. But currently no tool is quite right – as a replacement for Twitter, and as a correction of some of Twitter’s worst traits.