11–20 of 106 posts
I'm somewhat surprised that this space feels relatively dormant compared to the more complex stuff out there.
(APIs suck)
Many moons ago Twitter used to have RSS (Atom?) feeds for each user so you could use any old news aggregator to keep up to date.
> cheaply as possible (put everything in S3 yikes.
feedback is very much needed
Cheapest is to stay away from anything related to the big clouds, especially Amazon with their "premium bandwidth" charges.
I'd probably spin up a storage box from Hetzner (or similar service for dedicated servers) as a first step, as you get unmetered connection. It should last you a pretty long while as most content will be fetched from caches.
Or what if I could ask: "Given Omer Shehata's Twitter history, formulate a phishing scam that he would be likely vulnerable to".
The problem I see with here is that there are far more bad actor use cases for identifiable user data than good. In my opinion the main reason most social networks have stopped doing public by default and now do private by default is because not doing so opens them up to Cambridge Analytica type scandals where people don't realise what they're signing up for.
Personally if you do this, I would be very clear with your users that by submitting their data it will be made available publicly in an identifiable form. And that even if they revoke their data from your service it's possible for their data will continue to be archived by others, possibly for malicious reasons.
[flagged]
[flagged]
In my experience it's really only those who go there for politics that get annoyed by it – perviously it was the right-wingers who were angry because it was too left-wing, and now it's left-wingers who think it's too right-wing.
Personally I've noticed practically no change.
The trend of shutting down / charging steeply for API access has fundamentally changed the internet.
Now the Web is expensive for some reason, and users have become so dependent on a small number of sites as a communications medium that the megacorporations running them feel like they have infinite leverage, and the social contract of user-generated content is completely forgotten.
Users should just get off this continued tragedy and API access wouldn't be an issue
A lot of users/communities are stuck because of network effects & the long history of posts (they still reference them, like living wikis) These pipeline tools can help people migrate if they want without losing that history & network (my prediction is once people see what's possible with an open API, that'll further motivate user migration, or for twitter to open up again)
As someone with a multitude of cat photo threads, 100x this.
While I'm around on Twitter, Bluesky and Mastodon, the "post history" aspect is my most pressing issue with Mastodon. Like, when you move your account, you cannot take your old posts to the new server - you're straight out of luck if your instance decides to call it quits for whatever reason. And even if the instance doesn't give up, being unable to view old(er) photos is something I very commonly notice on a bunch of Mastodon servers.
What is the game theory here? Twitter cooperates and OpenAI defects, and we call that a win?