I noticed that Phabricator is exposing lot of stuff to search engines like this:
I think that every single changed bit is not something we want to track. Also because often we strip-out email addresses or phone numbers and we do not want to be like Wikipedia and keep track of every single damn change. We are just here to do some co-working and we should not feed search engines with shit.
To do this I've edited the Apache virtualhost to expose a robots.txt:
User-agent: * Disallow: /transactions/
Alias /robots.txt /home/www-data/gitpull.it/www-stuff/robots.txt