Comment by wruza

HN actively erases unicode regions to prevent emoji abuse and other zalgoing. Sites and apps do it nowadays, just not with emoji. It's the other side of your point – unicode can do too much and it's not a regular text, so you can't search within that sort of bold, validate, etc. So people choose to work with a subset, which may still leak: https://news.ycombinator.com/item?id=42231608