Bug when the same cross-net message is included twice in the same subnet block?

[adlrocha]

We should include tests that check what happens if the same block of a subnet includes two cross-net messages that are equal (for instance because the address decided to inject the same amount of tokens to its own address in the subnet). In some demos that we've done we've experienced some cross-net messages being skipped and not being executed when they are equal to some other cross-net message included in the same block (we haven't been able to replicate it again, but this may a symptom of a nasty bug and we should double check it).

In any case, we should include some kind of nonce according to the source address (even if the message nonce assigned is the one to the meta) to prevent potential replay attacks