yea I didn't have a single word in the one I just did that I knew...
bens - plural of Ben
vrot - (South Africa) rotten
caas - third-person singular simple present indicative form of caa
cava - Alternative letter-case form of Cava, a Catalan white sparkling wine.
itch is giving me the red squiggly line under vrot, caas, and caa....
To be fair, "my personally curated word list" is the kind of thing a lot of crossword puzzle creators keep closely guarded, because it's hard to decide what words should or shouldn't "count", especially if you're planning to go through them one by one.
In theory, https://www.spreadthewordlist.com/ could help; it's an open-source curated word list. Since it's crossword focused, though, it probably does include things you'd accept as a crossword entry but not as a word in a puzzle that, like this one, doesn't use clues. (For instance, phrases like "UM NO", or uncommon proper names like "ELIA", which are things that you can write a clue for.) Also of course it's only in English.
Also potentially helpful, if you're a Pythonista, is the "wordfreq" package (https://github.com/rspeer/wordfreq), which lets you look up the frequency of words. If you have a too-long word list, it might be able to tell you that "vrot" and "caas" are rarer than you want your words to be. (Also has the advantage that it's hugely multilingual!)
Anyway--I'm leaving any comment at all because I love word games, and I love seeing more of them, and so I'm happy that you made this at all and would love to see it be as good as it can be! Keep up the good work!
Super interesting stuff. I am currently exploring a solution where I infer the word usage using the wikidictionary dump. Other solutions would be to make my own (or let user submitted a daily grid). I also want to try using some available word frequencies, but I need a solution that supports French, Spanish, German, etc.
wordfreq, if it's a thing you can use, definitely has information for French, Spanish, and German. (If you don't use Python, the data files are available in MessagePack format at https://github.com/rspeer/wordfreq/tree/master/wordfreq/data, which ought to be readable in the programming language of your choice.)
I know Elia, who wrote wordfreq, used wiktionary dumps for other projects, and I know there's useful data in there (I think among other things she was filtering out offensive words?), but I don't know much about it. Best of luck!
