Just saying that we can talk in terms of weight alone after segwit. 8 mb weight is much more clear than 2 mb size to me. 2 mb size seems to obfuscate the actual new limit with the proposed hf, which simply 8 mb weight.

On 2 Apr 2017 12:03 pm, "Natanael" <natanael.l@gmail.com> wrote:

My point, if you missed it, is that there's a mathematical equivalence between using two limits (and calculating the ratio) vs using one limit and a ratio. The output is fully identical. The only difference is the order of operations. Saying there's no blocksize limit with this is pretty meaningless, because you're just saying you're using an abstraction that doesn't make the limit visible.