How do you enter these characters? With alt+num or directly?
From my previous hack of czech diacritics I believe it gets the codes based on the CP1250 codepage. I'll run that test for you right away.
Yup, it is. It's getting codes from the 0x80-0xFF range that correspond to my earlier research. I assume a codepage is enough for you and you don't need a list. If that's not so, let me know.
Thanks for looking into this. Back in January, before the UTF-8 support was complete I've made a somewhat hacked support of diacritics just by replacing the existing font. With the attached GRF, I could get 27/30 characters working as they should by just replacing the sprites. Two of the remaining mapped to somewhat similar characters, the last one isn't used too often. It worked fine, as long as the user was careful with those few odd characters - but I couldn't cause a crash with it.
You can see why it somewhat surprised me when it started crashing now.
Explanation of Czech used in comments:
male = small, velke = capital, hacek = caron, carka = acute, krouzek = ring above