Feedback on 1Password v8

Yes, exactly.

I'm making a wild number of assumptions about how 1Password search works, spewing some potentially-related concepts, and suggesting that fixing search should be easy, all at once! :-) :-D

My assumption is that when Unicode support was added, the search algorithm was changed from a naive brute-force like %string% operation to a tokenized full-text system. I'm making that assumption because it's an obvious approach with a bunch of benefits, and it would explain the current behavior.

The downside of word-boundary tokens is that they don't work for arbitrary partial search.

My (semi-flippant) suggestion to add n-gram tokens is based on that assumption. That's one way to enable partial-string matches when using a full-text search system.

And as a customer, I'm always right. I can confidently state that this should be 1) https://en.wikipedia.org/wiki/Small_matter_of_programming and 2) top priority.