Thanks for your feedback, everyone. Some updates from our team.
Those long verb words at the top are essentially what we want.
Unfortunately we don’t plan to add back the old splitter at the time, but we will continue to work on the new splitter and improve it.
@sentionaut thanks for letting me know I tried it, did re-split with ai and guess what happened, the whole title sentence is now blue 
as you suspected and @zoran confirms the only function that makes Japanese usable on LingQ has been removed!
I was going to try and be positive about this but then my latest uploaded lesson shows up like this even after being “re-split” and it has two horrors on the very first line. A third (unmarked in the second) So yeah, this is a HUGE disappointment.
It is not only completely unusable from today but the study time I will lose from now until this ‘new parser’ (which I think is just the old one) improves or the feature is re-implemented cannot be recovered. This is a setback for all Japanese users.
Someone on the development team obviously think they know better than everybody else, and despite being shown example after example will still maintain that those tokens are “what they want”.
It doesn’t matter if from a grammatical standpoint they’re essentially noise that has to be filtered out by the user, doesn’t matter if no user submitted definitions show up proving that nobody has ever used Lingq this way, doesn’t matter if multiple users come forward and explain that it just doesn’t work.
I’m done using Lingq for japanese until this actually get fixed, I’ve got a few study goals in other languages that I can use my subscription for while I move my reading setup somewhere else but I’m done wasting my time trying to get this issue sorted.
