...
Anyway, I noticed something interesting today. If you read the white paper.
1. There is nothing about block size. According to the white paper, any block of any size is valid if the contained transactions are valid.
2. The determination of validity is made by miners, expressed by mining on top of it.
As a corollary of #2, users who see a longest chain they believe to be invalid should simply be fixing/upgrading their software, not claiming that miners are misbehaving.
While perhaps a "good idea", the concept of a longest valid chain as distinct from the longest chain is not part of the design.
Whether this is a good idea is another question entirely, but it seems to me the small block side has skipped a step in their argument.
Very nice. Is it fair then to say:
"Bitcoin is the longest proof-of-work chain composed of valid transactions"
This is a subtle difference from the phrase that certain personas employ to label Bitcoin-XT and other larger-block clients as an "alt-coin":
"Bitcoin is the longest proof-of-work
valid chain"
According to the original design I believe so. Furthermore, I believe that a correct interoperation of the second phrase according to the original design is identical to the first, because
the determination of validity is made by miners.
I think your observation is quite important. Three arguments against changing the block size were:
(a) that is changes Bitcoin in a fundamental way;
(b) that changes to the consensus layer should never (or only extremely rarely) occur;
(c) that it would lead to centralization.
According to this new interpretation, a change to the block size doesn't represent a fundamental change to Bitcoin because it was always miners that determined the validity of a block by mining on top of it
anyways. Furthermore, this re-inforces the idea we were talking about yesterday that
the block size limit is not actually part of the consensus layer. The consensus layer should only define which
transactions are valid.
Point (c) may or may not be true. But by a group of developers limiting the block size and ignoring the will of the economic majority, Bitcoin would have already lost the battle against centralization.
I agree, although I will note that it is unclear how to determine the "will of the economic majority" objectively.
I also do not rule out that changes to the original design (such as validity being determined in some manner other than by miners building on top of blocks they consider to be valid) may be desirable or can be done in an appropriate manner. I do not believe that appropriate manner includes assuming such a change to have already been made.