So, I've been thinking about what would the ideal decentralized social network would be and I think I have the right idea.
Basically, it would just be a modernized RSS feed, just a simple file that can be hosted on any webserver, kind of like what twtxt is doing, but not focused on ultra-minimalism and instead be a JSON or XML format that can also include photos, videos, links, etc. Feeds can be accessed through clients just like RSS and there will be so called "aggregator servers" where people can access that will display aggregated content from various feeds submitted to it (similar to registries on twtxt).
I think there are some advantages to this:
* Feeds can be hosted on any webserver and can pretty much be implemented in almost anything from existing social media software, content management systems, to static-site generators.
* Self-hosting would be much easier.
* People are limited to interacting with the feeds and aggregator servers they follow. They can have a much more relaxed experience (if they choose) by only connecting to the feeds and aggregator servers that they trust.
* Someone can just have a simple feed and ignore people interacting with them. They could just have some feed that they either manually edit or have automatically generated with some script and not connect with anyone through a client.
Sure there will be some disadvantages and they would be the same as twtxt. For one, you will only be able to see replies if they are from feeds or aggregator servers that you connect to, but I think this would be perfect for those who just want to share stuff without getting into all the drama present in most social media. It can also be a viable replacement for RSS, something that should've caught on, but didn't due to the rise of social media (along with other reasons that I will explain later).
I will also like to take this as an opportunity to explain why RSS (and other feed formats) never really caught on outside of tech-savvy users because I hear a lot of people complain that RSS should've caught on instead of Twitter.
* The acronym makes it look like it's something that only web developers, web indexers, etc need to care about - just put the name "RSS" next to HTML, CSS, XML, etc and it sounds like it is meant for web crawlers or API programmers. The full name, "Really Simple Syndication" doesn't sound user friendly either. They should've called it "web feed", "e-feed", or "web subscription" and people who find the link to it will have a better idea what it is.
* Not many browsers supported it out of the box - The only one I remember doing it was Firefox with "live bookmarks" but that has been long discontinued. Apparently, everyone thought it was a better idea to have email clients support it, or just have services like Google Reader for reading RSS feeds instead.
* The negative stereotypes of bloggers - I don't know how many people remember this, but back in the mid to late-2000s bloggers were mocked and ridiculed as being a bunch of wannabe celebrities or failed journalists. Given that nobody liked bloggers, almost nobody bothered with RSS feeds.
* A lot of people were still editing pages manually - Not everyone was using blogging software or a CMS. A lot of people were either writing web pages by hand in a text/HTML editor, designing pages in a word processor/WYSIWYG editor, or using a custom backend they written themselves. With the former two, you had to manually edit the RSS file. The latter requires you to implement RSS yourself which many people were either too lazy to do or didn't know how to do.
I think my proposed replacement can address some of these issues. Having self-contained text and image posts and having servers that host it, in addition to self-hosting can allow anyone (not just webmasters and bloggers) have a feed. Getting major browsers to support it is difficult as it needs to become popular on it's own first. And a lot of people are using static-site generators, content management systems, and other software more than people edit pages manually these days, so it could probably work. The last issue is the name and it should sound as user-friendly as possible.