I read all the info published. Without multisig, I don't see a way of forceful situation in a p2p system. There's no restrictions to the middle nodes, you can't simply "expect" it to behave correctly
It is not the problem if we need to know how middle nodes work, it is the problem that if I put a fake node with same APIs but doing different things (e.g. steal coins or send coins to different places), how the system can prevent that? You can't simply expect people use the software you provide, otherwise that's too easy, there will be no cheaters, period.
Cloak is in the middle of implementing a "block escrow" system that may share some similarities to a multi sig solution but is infact unique.
Once in place there will be nothing that a "bad/fake/evil" node can do to act malicious during a send from A to B. This is because of the order of the posa tx, combined with the blockchain, will not allow corruption.
PoSA will soon be entirely anon + trustless as we set out to achieve from day 1.
Having any kind of risk that a corrupt node could interfere with a transaction was a genuine concern (of the Cloak team and others) and as such it's been dealt with accordingly.
Good to know, and btw thanks to come to this thread and explain. Looking forward to see your "block escrow" system. But the multisig tech is there, of course you don't have to use it, but implementing some thing similar from scratch sounds stupid to me. This is like to say I don't want to use C++'s class, but I want to implement some structs to mimic the classes
Node corruption or any kind of error handling in the middle of the transaction is a complex problem, I am looking forward to Supercoin's dev to publish their detailed algo, with their trustless system. But usually this has to be handled depending on the stage of the transaction. Again I see the power of 2-of-3 transactions (or in general m-of-n transaction), as this can be handled easily.
