Yeah, I wouldn’t try to use hashtags as communities. That’s not what Reddit did either. Hashtags were an interesting hack but too minimal to solve the communities problem we had on Twitter and have inherited with Bluesky.
Yeah, I wouldn’t try to use hashtags as communities. That’s not what Reddit did either. Hashtags were an interesting hack but too minimal to solve the communities problem we had on Twitter and have inherited with Bluesky.
The pieces for a Reddit-style community in bsky feel almost there? Hashtags + custom feeds + Discover-style "less like this" feedback + custom moderation. Issue right now is you have to cobble it together. Like, Blacksky is doing that but not sure who else is.
Also, creating a custom feed from hashtags and exact word matching feels like it should be built into the app itself and not something you need something like skyfeed for?
Yeah there is some kind of version of this already but it needs strong product support to flourish. Feeds could be enhanced to do it. They should probably play a role. But it’s probably just better to make an explicit thing to create the exact experience we need.
In app feeds could also create lockin if the controls / logic / definition are not portable. Even so, I do not think I can migrate my @graze.social to another provider They each provide a different set of knobs and ux, even though the viewing is consistent. Maybe a @lexicon.community "standard"?
#bb27
Curious... Given the constraints (e.g. 140 characters and SMS), how would you have solved the problem differently?
I think the Twitter product was good enough to compete until 2012 or so. And by then SMS wasn’t popular and they could have abandoned it and its limitations safely. I think namespaces e.g. sub-Twitters was the obvious and correct solution. For same reason subreddits made Reddit so sticky.
It appears that you're sidestepping the point of my question. Twitter launched "proper" communities in 2021 (they still exist). No one I know uses them. Why do you think these are the correct solution when they were tried and seem to have failed?
From your question it seemed like you were asking about the old days. Simple answer is that Twitter did a bad job with Communities. They didn't care or didn't understand what they were doing. It's easy to launch something that, on paper, sounds right but is way off. See: Google Video vs YouTube.
I agree with your point about copying community infra badly. -- This conversation is challenging because we're bouncing around in timeframes, constraints, and considerations. -- I originally addressed your critique of the hashtag and asked what you would have proposed in 2007.
I think the hashtag was a fun hack and did provide utility at the time. Lots of times these kinds of hacks/emergent behaviors are signs that a product is missing some good feature. Just like ad-hoc @ and RTs got turned into core features, so should have interest-based communities been productized.
Super interesting thread. My 2¢ - I'm not sure folk open BlueSky/Twitter for the same reason/vibe/exp that folk open Reddit (or FB). Porting features or trying to homogenise doesn't seem wise. Hashtags fit a microblog vibe: emergent, quick, easy. Communities are intentional + for elsewhere.
For clarity, what I'm trying to say is about 'intention' ... One joins a subreddit / FB Groups / forums – with subject-matter intention. The intention (for most, I suggest) on bsky/tw is more vicarious, curious, serendipitous. That's why Twitter communities failed. Horses for courses etc. 🖖
You could not be more wrong about this. That’s not what people wanted from Twitter. Nobody joined Twitter saying “I can’t wait to join a small community of individuals who are one particular interest.” Twitter was about braindumping whatever you were thinking about on any topic into a firehose.
I'm much less interested in what twitter "was" than what it could have been, had it kept evolving. Twitter completely failed to achieve strong retention and mainstream adoption, the way Reddit/YouTube/Instagram did. There were posts like the following for a many years without any meaningful fixes:
You seem to think users will tell you how to evolve a product. That's not how it works. Users identify problems, often in confusing ways, and then you (as the app developer) develop solutions. Users are busy, don't know all the options, lack the necessary context, and it's not their damn job.
Users will tell you how to evolve a product no matter what. Product design is the process of assimilating all of these requests and mashing them through constraints, while also maintaining an eye towards overall coherence and usability. As a product leader, I wouldn't outsource my role to users.
Wait no, but I'm talking about how *gestures around* feels like in 2025, things weren't like this in 2018
The issues were different/less crazy but the chaotic nature of Twitter was the same. No one really enjoys the disorganized chaos. Some just tolerate it better. One example: since Twitter launched users had been asking for a way to follow someone for only specific topics they were interested in.
your assumption that this is a problem that needs fixing as incorrect. It’s just the way things are. it’s not a product problem. simply an observation.
I partially agree with both of you 😉 The "everyone in one room" aspect of Twitter was indeed its strong point, and it's why I've been using Twitter until last year and why I like Bluesky now, and why I've only used Reddit for some periods of time in some specific topics (ehem, crypto)
But also Jake is right that often I only want to read some topics that someone is posting about and not others, but I can only follow them completely or not at all… so I think there is some space for some topic categorization if done right.
Could outline hashtags do it? 😉 Maybe, who knows? But also I don't want fully separate "community subreddits" like the Twitter Communities, because for a lot of people I do actually want to read everything they are posting on any topic, that's how you expand beyond your current interests
I'm working on building out this capability and building my own feeds. See my github.com/blebbit/at-m... project But first, "permissioned spaces" via my patches here: github.com/blebbit/atpr...