Funny how people tend to quote Linus Torvalds on technical issues in order to prove a point. I hear the guy barely made it through college.
Actually, I lied about even that. I was thrown out of fourth grade because I couldn’t write my own name, and it’s been all downhill from there. I had to lie about getting into college just so that I’d have better chances of making a career here at McDonalds - if you have a college degree (or you lied about having one), they don’t make you scrape the burger pans.
heh, he had good sense of humour already back in 1999
there is a bridge, but some parts of the fediverse got very mad about it
https://github.com/snarfed/bridgy-fed/
https://fed.brid.gy/
so this is similar to LibRedirect?
https://github.com/libredirect/browser_extension
https://libredirect.github.io/
check out this list of privacy frontends and see if you can implement any of them in Predirect
Also, I see that you only handle one instance for redirect. Either the default instance or custom instance.
Consider implementing multiple instances.
Here is list of instances that LibRedirect uses:
https://raw.githubusercontent.com/libredirect/instances/main/data.json
Here is a list of instances that Farside.link uses:
https://raw.githubusercontent.com/benbusby/farside/master/services-full.json
Both LibRedirect and Farside still have to deal with the rate-limiting problem with Youtube and Reddit and I assume other sites as well.
Invidious instance or Teddit instance can be rate-limited, resulting in video not playing or post not showing up.
If your extension does not distribute user traffic across multiple instances, then I assume that your chosen instances will get hammered into ratelimit even faster than other instances.
So consider thinking about solution for this.
One of suggestions that I like is to allow user to provide an URL to a list of instances.
That way, someone or some automated system can periodically compose a list of instances and users can “subscribe” to that list.
Also some useful links:
https://github.com/digitalblossom/alternative-frontends
https://github.com/pluja/awesome-privacy#social-networks-and-platforms
https://github.com/mendel5/alternative-front-ends
https://matrix.to/#/#alternative-frontends:tchncs.de
doesn’t Steam Hardware Survey report way higher percentage of Linux users?
wouldn’t Valve’s numbers be more reliable?