On Tue, Jul 23, 2019 at 12:15 PM Jason Cobb <jason.e.c...@gmail.com> wrote: > A player generally CAN become a party to an existing contract by > announcement. However, if the contract explicitly limits the > persons who can become party to itself, any person not > fulfilling those restrictions CANNOT become a party to the > contract. Before the creation of a contract, if a person could > not, in the hypothetical where the contract already exists, > become party to the contract, e is not counted as consenting to > the agreement for the purposes of the previous paragraph, even > if e has agreed to be party to the contract. > > [Comment: The goal is to resolve the bug that G. recently showed (with > the contract that states that it is impossible to join). This would > prevent such a contract by ensuring that it could never reach the two > parties required to create it. This also gives force to clauses that > purport to limit the set of parties.]
I'd be sort of interested in the CFJ interpretation of the current rules. For example, you could argue that if the players are making an "agreement among themselves" based on a text that doesn't let them join, they are in fact agreeing that they are not making the contract, which is a no-op. Also, the situation may boil down to the straightforward paradox "We agree to join the following contract: {this is not a contract}". Under your wording, a "hypothetical contract" stating that it is not a contract is as paradoxical as a non-hypothetical example. There's also a broader issue: any time you let players create texts freely that regulate CANs and CANNOTs (as opposed to dealing exclusively in SHALLs) it's pretty easy to come up with paradoxical contracts of many different sorts. We may need some broader protections: this particular version of Contracts was originally written assuming that only SHALLs were being put into contracts, the additions that made asset transfers and act-on-behalfs POSSIBLE but regulated by the contract text were bolted on without looking at the issues more deeply (like publishing requirements and how to error-trap indeterminate stuff). -G.