Improve Option::None size validation

[danielbate]

In #2322 we removed the expected size validation for nested options. That should be reintroduced.

[arboleya]

Is this still a thing?

[danielbate]

It is yes, as we are not correctly validating option length. Not a massive deal as Option::Some is validated correctly at the child level and None is empty bytes. But with the correct design we can solve for this. However improved length validation in #1795 would also solve this.

[arboleya]

Two more questions:

1. Can this be a blocker for mainnet?
2. Is there any relationship between this issue and this PR?
   • chore: add tests to EnumCoder #2515

[danielbate]

1. No it should not. At some level, an option will need to have a native type, at which point it will be validated correctly
2. No, there is a single line missing coverage in the EnumCoder that this will resolve.

[danielbate]

Going to un-assign my self as this could be picked up by anyone and is not on my immediate radar.

[danielbate]

At some level, an option will need to have a native type, at which point it will be validated correctly

For the above reason, I think we can deprioritise this issue. Additionally this code has now been audited in a few ways and it remains sound. There is enough validation in other coders and around the empty option that we can have confidence.

IMO the real value will come from improved coder wide length validation in #1795.