I find it interesting that they didn't randomize a couple of long strings and tried to resolve those instead like the article mentioned has been done in the past