Harden Tangle to Malicious Peers about tangle HOT 3 OPEN

Would you consider a simple majority consensus logic based on something light (e.g. some hash)?

from tangle.

Potentially misunderstanding the suggestions here, but it could also seat quite well with the solution to #7 (Dirty Flags) and a simple majority based on the content hashes of the deltas (which may require comparing a list of hashes if multiple changes are being synced)?

from tangle.

Yeah, hashing is likely the solution to detect and resync peers who have desynced.

The issue I want to prevent is a malicious peer sending bogus messages (or omitting messages) to just one peer in an effort to sabotage them. Imagine an evil peer (we'll call peer E) is in a room with a bunch of others and they want to sabotage peer A. Peer E could send A a bunch of bogus messages causing peer A to continuously desync.

Hash consensus could be used by peer A to detect that it's desynced (and likely something like that should exist anyways) but it will still give peer A a degraded experience, or at least give peer A increased latency. Is there a better way?

Another issue is when joining a room how can a peer know it's being sent the correct data? Likely a peer should receive a hash of the room state from a few peers and the full room state from one peer and then verify the hash itself when it receives the room state.

from tangle.