Heh, Facebook popped up this "memory" today.

So a decade ago there was a bunch of postMessage handling code with a regex like /^(m|www).facebook…/, so I grabbed a domain that matched. I picked 'n' for the arbitrary letter, hoping to get a few free lucky bit flips, but then Baidubot hit me with these. Still not sure why. Seeding crawl by DNS logs would necessarily be a surprise, but... even if they did, I don't remember search-www.m.facebook… being a host name that was in use.