The course prompts with a Chinese character and requires a pronunciation written in hanyu pinyin (roman alphabet with diacritics) as an answer. Previously I had no issues with this course, but it is no longer able to distinguish between roman letters with different diacritics e.g. mā, má and mǎ are all accepted as ‘correct’ answers when only mà should be accepted.
In addition to the four tone marks (ā, á, ǎ, à), the course is no longer able to distinguish between lu and lü (two different sounds in Mandarin). This essentially breaks the course.
H, I am afraid I can offer no solution, but I can confirm that Memrise seems to have an issue with unicode: E.g. superscript digits, which mark pitch pronunciation in Norwegian, are also not recognized. however not at he quizzing step but when entering them in to the database.
I’ve a similar issue with Hind. In Hindi Vowels are usually represented by some extra diacritic symbols, like accents. On Android any vowel is accepted; as long as the consonants are right, the answers are accepted. Hope both can be fixed in one go.
What’s most frustrating is that it never used to have this problem: using Memrise is a constant battle to ‘trick’ the website into providing the same functionality it had back in 2014.