Many people underestimate the reliance of Plasma on the mainnet consensus, while others think of it as overly mysterious. In fact, Plasma's dependence on the mainnet is neither heavy nor light, but very precise.
#Plasma does not rely on the mainnet to execute transactions, nor does it rely on the mainnet to store data. Sidechains can be fast, chaotic, or even temporarily disordered, as long as these actions have not been 'convicted.' But it relies heavily on one thing from the mainnet: the irreversibility of consensus.
The role of the mainnet in the Plasma system is more like a 'time and adjudication machine.'
Every time a block header is submitted, it is essentially a timestamp on the mainnet: at this point in time, Plasma claims its state is this way. If the mainnet consensus can be rewritten, these commitments lose their meaning, and fraud proofs, exit windows, and challenge processes will all become invalid at the same time.
It is also for this reason that
@Plasma does not require the mainnet to participate in complex logic, only that it is slow enough, stable enough, and difficult to tamper with. The mainnet does not need to know who is right or wrong, as long as it guarantees that 'what happens first cannot be denied later.'
This is also why Plasma can theoretically be built on any consensus-stable settlement layer, not limited to Ethereum. It does not care whether you are
#POW or
#Pos , it only cares whether the finality is trustworthy.
So, Plasma does not outsource security to the mainnet, but locks the final adjudication power within the mainnet consensus.
Once this consensus is established, all games within Plasma become meaningful.
#plasma $XPL