I think you mean language filter. Internalization often means something else
Mastodon tries to implement a language filter too, but it didn’t go very well from what I heard – false negative and false positive happen both too often.
I wouldn’t be against this feature though. IMO there should be a message confirming if the language is recognized correctly to refine the language model, but that might be diverting developers’ effort from more helpful features.
Everything with underscore behaves weirdly on Lemmy
The client API can be used to make it possible. There is a mirror bot on lotide, btw
I think these comments would mention commenters of all parent comments, which could be annoying for Lemmy users.
I agree, there is too much bad smells from this instance, starting from the arrogant title.
There are reasons to oppose covid passport other than antivax, such as concerns about privacy and surveillance.
Because the filter is hard coded, you must fork and remove that part in the source code, and then recompile it (Rust compilation is costly in terms of time and memory btw). Then you would have to manually apply patches from upstream whenever there is an update.
I think the ones without a word separator should be gone, because it’s harder to read