Quote from: xyzzyx on August 13, 2014, 11:38:36 PM

Quote from: jl777 on August 13, 2014, 07:51:35 AM

This multisig is really fun. After taking all this time to get working, now I just cant stop trying many combinations. All looking quite stable and very solid, but of course it is using localhost, so no real networking errors.
UDP sometimes is pretty finicky, but that is another advantage of the shared secret multisig. It essentially acts as an error correcting code, well at least as long I dont mix in packets that are corrupted. If even one bit gets changed, it will not match the packet crc and so just get ignored, then the decryption will also fail, so I would say the odds of bad data getting processed is ignorable, even using UDP path.

Have you considered doing reed-solomon on the data to correct any corrupt packets? It would add some redundant overhead, but you'd gain reliability in noisy networks.

http://www.drdobbs.com/cpp/reed-solomon-error-correction/184410107

I think M of N has the same effect, though nowhere near as efficient as rs codes. I am not sure how big a problem it is, so if it does turn out to be a real issue, then I will certainly consider such solutions. For now, I have enough code to write so I want to finish what I know must be done first