I think the fediverse needs to come to terms with the fact that Instagram and Telegram are the biggest things we actually need to make solutions for, not Twitter.
1 Boost
0 Likes
@catgoat I think part of the problem here is that people have such different ideas of what Instagram *is* and what telegram provides. Like in Berlin telegram and Instagram are being used as event platforms more then anything else. And Instagram is used as an identity / address book more then its used as a photo sharing platform
@liaizon almost everyone in the US I know uses it as a messaging, short videos, and event/shopping platform. But the short form videos are what really keep people stuck in there. That + TikTok replaced Facebook as where to find local shops or restaurants, though.
@catgoat yeah exactly. It's this very weird issue that I don't think is probably discussed. Like you need all these different hooks combined in spesific ways. Like an event app needs to be combined into loops style content streams or something
lots of corporate platforms now help shape popular understanding. for the worse because they serve the interests of their billionaire owners and advertisers. that's what we should work to address. and via decentralization and the lack of algos and ads and with the power of human boosts, are perfectly positioned to do so. to provide an alternative ecosystem for the natural desire to know what's the next thing to pay attention to. blogs, podcasts, #peertube etc
I do not think the Pixelfed or Mastodon projects are at all moving in the direction able to take us there to build a base for the next million MAUs. For many reasons, but to start, for their choice of programming languages they are written in.
@liaizon I think the bigger obstacle for building viable ActivityPub replacements for the corporate photo sharing apps may be storage costs. It's a wall I keep expecting Pixelfed instances to hit.
@yhancik @liaizon I was thinking it might be cool to have an ActivityPub service that was all about lo-fi image posting, with conversion to <100kb format happening when you upload. Maybe it even provides a choice of "filters" so you can upload the pic with different types of compression, dithering, minimal color schemes, whatever. Could be a fork of Pixelfed, I suppose, but given Wake's point about its PHP foundation, it might be better to start from scratch.
@liaizon facebook ran a few dozen million MAUs on PHP just fine, on servers less powerful than my smartphone. Not to say that fedi software shouldn't be more efficient/higher quality/maintainable, but I think the choice of programming language is the smallest problem here.
@fionafokus I don't think it's that there are not ways to optimize PHP or Ruby but that for this to really work on the level we need it to work we need thousands of people excited to be writing code in those languages.
@liaizon I think I need clarification.. Do you want to say that it takes more developers to get these programming languages to perform, or do you want to say that more developers are needed to implement the features that are needed?
because in my experience governance issues are what keep people from contributing, not the programming language.
@liaizon @fionafokus people don't like Ruby? I don't do much backend, but I thought Ruby was still held in good standing. iirc Shopify is a Ruby shop.
What would the better + more popular option be these days, Go or Rust I'm guessing?
I think we need serious collective of people to build out a mission driven and ideologically motivated "startup" that lays out not just a single "product" attempts fostering an ecosystem that works together.
I look out at the tech collectives that do exist and don't know what exactly I think a replicatable model is. I think this problem is closer to building a web browser then it is to starting a company or designing an app.
Telegram is solved. Instagram partially (Pixelfed exists, but is inacceptable as a project).
@nik absolutely not
@liaizon telegram I think reasonably could have a federated replacement (even built out on xmpp or something) with better branding. Instagram, partly the addictive nature of reels/short videos is a huge driver for people using it. I’m not sure anyone should make a methadone version of Instagram