Rollup for mass exit

[boolafish]

Origin idea from PG: https://plasma.build/t/rollup-plasma-for-mass-exits-complex-disputes/90 And some notes from Plasma call 21.

This uses "roll up" but nothing with snark. That is, use call data to put all the exit and save only merkle root of all the exit data. Challenge period still needed. The exit data would need your basic output data + the block it should be included (don't need inclusion proof while start the exit, could be challenged later and respond with the proof)

Seems to me that this can increase our max UTXOs in plasma MVP. However, one issue for mass exit remains is that how can people coordinates. IMO, it is probably only safe to assume people would mass exit all their UTXOs at once. With this assumption, we can claim our system can support N users where each user have at most K outputs.

N is the limit to start mass rollup exit in our exit period and K is the limit how many exits we can batch exiting in rollup.

[slavamirovsky]

Any updates on Max Exits here?

[boolafish]

probably Piotr's idea is better: https://github.com/omisego/research/pull/106