[peterwwillis]

This seems like a fun new opportunity to find new ways to avoid censors.

One way would be to seed messages or websites with pre-loaded meanings, and use new, unrelated comments to refer to them. Example of a social media feed:

8/15/2017: "Today at noon in the square."

9/15/2015: "I think it's evil and corrupt and we can not stand for it!"

10/22/2017: "The government has taken a new decision today that affects us all."

11/10/2017: "We must protest this injustice!"

2/27/2018: "10-22-2017 + 9-15-2017 + 11-10-2017 + 8/15/2017"

Obviously this is clunky, but in theory you could chain together random bits of other people's messages to post a new one, and a censorship bot would have to be written to merge them all together to filter them; if some of these were behind logins on other sites, this could be very difficult. You could also make them less literal and more made up of memes and concepts with hidden meanings.