Ask HN: Article about how string comparison is not enough

I was digging through bookmarks looking for something I for sure found on HN sometime in 2018.

The article was essentially discussing how "unique" in the database is worthless in a lot of situations, namely:

- email addresses (with +sub@addressing)

- case sensitivity

- unicode folding (ᎻoⅿоglｙＰᎻ attack, puny code)

- a bunch of other cases around visually similar, but byte non-identical data

The article was discussing this in the context of "unique" or "equality" in terms of a database unique key, which is not enough to ensure uniqueness.

Popped back into my head today because of the unicode.style link that just landed on the frontpage (https://news.ycombinator.com/item?id=17752680)

Does anyone know the article I'm talking about?