You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Question:
Another point we need to discuss is what happens to chimera chains once the conditions dont hold anymore. Are the objects stuck there for ever? What if no one kills the chimera chain, but the conditions also have not been met for 1 months and hence there is no way to kill the chain?
Reply by Isaac:
Good question.
I suppose that, once the chimera chain has updated its quorums, users can see that its atomicity guarantee is now very bad (ex, atomicity is not guaranteed even if everyone is operating correctly).
However, users can still commit transactions to the chimera chain. What they lose is the guarantee that if they commit a "multi-state-machine" block, then if one state machine's transactions are committed, the other state machine's transactions are committed.
However, if they're only interested in committing transactions to one state machine, that would be ok.
Conveniently, the transactions necessary to move state back to a main chain are only on one state machine (the one corresponding to said main chain).
Therefore, even after atomicity guarantees are lost, we can move state back to main chains, or onto other chains with the same quorums.
Alternatively, if we had some uniform way of "moving state" from the chimera chain back to the main chain, we could automatically move any state back to the main chain as part of any quorum update that would make the chimera chain's atomicity guarantee worthless.
The text was updated successfully, but these errors were encountered:
Question:
Another point we need to discuss is what happens to chimera chains once the conditions dont hold anymore. Are the objects stuck there for ever? What if no one kills the chimera chain, but the conditions also have not been met for 1 months and hence there is no way to kill the chain?
Reply by Isaac:
Good question.
I suppose that, once the chimera chain has updated its quorums, users can see that its atomicity guarantee is now very bad (ex, atomicity is not guaranteed even if everyone is operating correctly).
However, users can still commit transactions to the chimera chain. What they lose is the guarantee that if they commit a "multi-state-machine" block, then if one state machine's transactions are committed, the other state machine's transactions are committed.
However, if they're only interested in committing transactions to one state machine, that would be ok.
Conveniently, the transactions necessary to move state back to a main chain are only on one state machine (the one corresponding to said main chain).
Therefore, even after atomicity guarantees are lost, we can move state back to main chains, or onto other chains with the same quorums.
Alternatively, if we had some uniform way of "moving state" from the chimera chain back to the main chain, we could automatically move any state back to the main chain as part of any quorum update that would make the chimera chain's atomicity guarantee worthless.
The text was updated successfully, but these errors were encountered: