[Bitcoin-development] BIP70: why Google Protocol Buffers for encoding?

[Isidor Zeuner]

> That was essentially what we did in the end, we replaced the network
> identifier ("main"/"test") with the genesis block hash. The result is
> never going to accidentally work with Bitcoin Core (nor vice-versa), but
> is readily extensible to any other altcoins that want to use the
> specification without requiring any sort of central registry.
>

Interesting approach, and I also think that requiring a central
registry would be potentially harmful.

However, I think it might not be adequate to think of the network
identifier as being congruent with the genesis block hash. In the
theoretical case of the blockchain being continued on two forked
chains (with two communities which prefer one of the chains each),
clients would not be prevented from interpreting messages on the wrong
chain.

Best regards,

Isidor