Has anyone noticed CM glitching when you try to add words to sets?
It didn’t used to do this, but started maybe 2 weeks or so ago. This is particularly annoying when trying to target prepositions and/or conjunctions, which are notoriously difficult as it is. It’s also annoying for people learning a language that uses declension, because CM has begun to treat combined letters as simply combined letters instead of as separate words. It also does it whenever I click on a word in a particular sentence to add it to a set. If that same word appears somewhere else in the sentence, it picks one of them at random, which is annoying since words mean different things in different places.
For example, the word дело in Russian includes the first three letters дел in all 12 of the different ways it can look because of declension. So there is a lot of overlap. But if you try to target the 12 different ways дело can appear by adding them all into a search for a Cloze Collection, any overlaps end up looking like {{дел}}ах instead of делах, {{дела}}м instead of делам, and {{дел}}е деле, etc. Russian is also a synthetic-fusional language with a high degree of inflection, so things like дел can also be affixed to form other words (e.g., отдел, раздел, неделе, разделение, etc.) This is dauntingly tedious to correct manually when you’re creating sets with thousands of sentences.
In short, please go back to treating words between brackets as specific stand-alone words instead of as parts of other words or the same word in a different usage.
All that said, I still love this site. Some update CM did probably just created this problem as an unintended consequence.