benbusby/farside

twitter redirect biased to nitter.net

1268 opened this issue · 1 comments

1268 commented

I did this 9 times and got this 8 times:

> curl -I -A iPad https://farside.link/twitter.com/mikko/status/1692503249595584526
HTTP/1.1 302 Found
location: https://nitter.net/mikko/status/1692503249595584526

and this once:

> curl -I -A iPad https://farside.link/twitter.com/mikko/status/1692503249595584526
HTTP/1.1 302 Found
location: https://nitter.it/mikko/status/1692503249595584526

at this point I might as well use nitter.net directly.

It's not biased, it just so happens that nitter.net is typically the first to roll out fixes to its scraping algorithm, and seems to have fewer issues than other instances. Although, running your same test didn't return nitter.net for me at any point. There isn't any code to prefer one instance over another though, it just selects a random one from a pool of working instances and then prefers that instance for a short period of time afterwards.