ZegnatFor my implementation in php-mf2 I replaced all “textContent” requests in the spec with one and the same algo. The question being: was that right? For things like the very final step in u-* parsing, it might be better to stick with raw textContent.
