Planning to make a new micro blogging platform :)
I am in LOVE with Bear blogs, it is just so simple and convenient to use! I hardly post anything in here, but I love reading other people's post.
Nowdays social media like X or reddit is full of shit and are filled with idiotic, nonsensical people.
I had a idea while browsing through bearblogs, why shouldn't I make my own blogging platform? Though I did not pay any heed to the suggestion that I should make it algorithmic driven, such as twitter or the million other social media.
It should be simple to use, minimal, decentralized, privacy focused, and only worked in the terminal like twtxt. Speaking of twtxt, it was such a wonderful project that had a lot of potential, but all of that is now useless. So i'm planning to make my micro blogging platform like Twtxt. Let's call my micro blogging platform "Curtain" for now. Curtain seems like a seems like a nice name.
So, this is how I imagine how it should work: Every user has three files - their following.toml, about.toml and a directory containing posts-year-month. Their about.toml contains their metadata such as their username, the day they joined Curtain etc.
And the directory labeled posts which contains their posts AND their timestamps, separated out by the month. At first, I thought to just make a single file containing all their posts, but over time the file would become gigantic and it would be much simpler to organize as well as parse.
Oh, speaking of time taking to parse it, I am planning to make this in Rust (btw) :)
So, basically the idea is that every users hosts their own files, with their own rules on how it is stored. You could theoretically self host this on a old machine, on a static server, or on github pages (the easiest method).
So you could "follow" people to get their "feeds" which is basically their posts.
This is where following.toml comes in.
You could basically add the links to the website where other people's curtain files are hosted, and then whenever you start it, Curtain fetches their latest updates and gives you.
For now, I don't have any idea on how to get people's update that you have NOT seen. My english is kind of bad, but what I am trying to say is that you should be able to see the new things that others have posted and not get their entire files whenever you run this. I have thought about storing their last post, and then after getting the new, updated file, comparing the last saved post vs the newly fetched file, and only parsing the post that have occured since then. This saves time while parsing the .json files, but it would mean redownloading their file again and again everytime you run this. This wouldn't cost that much bandwidth, but I'd still like some advice that could tell me how to solve this problem WITHOUT creating a server or a backend. I do not want to absolutely do that.
Anyways, sorry for my rambling and incorrect english, I apologize for any mistakes. I hope you got what I was trying to communicate with you. If you have any idea on how to solve the problems, please be sure to hit me up at cdrlsw9n@anonaddy.me (can't share my real @, sorry!) . Until then, see you next time!