[VaultBundles] force withdraw bundles

[QGarchery]

design

initial thread
follow up thread

[QGarchery]

We expect users to make the judgement of chosing between forceWithdrawLiquidV2 and forceWithdrawIlliquidV2. Wondering the consequences if the state flips between signing and inclusion. Calling liquid on an illiquid vault just reverts with gas costs, which is acceptable. However, calling Illiquid entry point on a liquid vault results in the forceDeallocate penalty which is avoidable. Is my understanding correct? Would it make sense even in the illiquid entry point to try first the liquid withdrawable without penalty and only the excess thru the forceDeallocate route?

yes this is documented "It may be the case that the vault became liquid, but calling this function still yields positions on the markets.", also see here. As discussed in the call the rationale is the following:

• the liquid case is similar to a normal vault v1 withdraw
• the illiquid case is supposed to be used for emergency scenarios. In those scenarios (where the curator/owner is adversarial), it is not expected that there is much liquidity anyway as other users would want to withdraw as well, without paying the penalty. Also we want to make sure that the emergency situations are as simple as possible, for safety reasons, and don't risk to revert

[MathisGD]

worth explaining "IKR"