Should we implement a "setFintroller" function in the HToken?

[PaulRBerg]

Since v1.8.0, the HToken contract expects a new argument in its constructor, which is the address of the Fintroller contract.

Would it be worth it to add a setFintroller function, callable only by the owner of the contract?

There's already a setBalanceSheet function .. though that hasn't been ever needed thus far.

[scorpion9979]

I think I would rather we remove setBalanceSheet than add setFintroller. Doing so would save 22 gas on any external function which has a signature that sorts after the one for setBalanceSheet in the function dispatch table. Of course, same result could be accomplished by renaming setBalanceSheet in order to push its function signature to the last position in the function dispatch table.

References:
https://medium.com/joyso/solidity-how-does-function-name-affect-gas-consumption-in-smart-contract-47d270d8ac92
https://ethereum.stackexchange.com/questions/68098/why-does-adding-a-view-function-make-write-functions-cost-more-gas

[PaulRBerg]

Having setBalanceSheet is really just an insurance policy against any potential bugs in the BalanceSheet contracts. Supposedly, we would patch those bugs ASAP, upgrade the proxy implementation, then call setBalanceSheet on the hToken.

Through this lens, 22 gas is a negligible cost.