>"For example, none of the sequences starting with a packet sent by the remote peer triggers blocking , confirming that censorship is likely not symmetric wit

>"For example, none of the sequences starting with a packet sent by the remote peer triggers blocking , confirming that censorship is likely not symmetric with respect to inside and outside Russia. Such behavior makes remote measurement extremely challenging. We also note that sequences that contain a remote-sent SYN packet can “un-block” a connection despite the presence of a blocking trigger." If the above is accurate -- then that would mean that outbound connections (from an arbitrary IP and/or associated domain) are apparently not blocked, or can open up a block if one exists -- but incoming connections (to an arbitrary IP and/or associated domain) could very well be blocked... In other words, if we assume the article to be 100% correct and wholly true -- then a potentially important asymmetry (inbound vs. outbound connection) would indeed seem to exist there, as the article notes...