Can't we see what Luke Jr. releases first?
I doubt that Luke Jr. is going to include hardfork language in the coding unless it appears that there is consensus for such... I thought that Luke Jr. already pretty much said that a hardfork would not be necessary in order to implement a 2 mg increase in the blocksize limit, so long as there is consensus for such protocol changes? And, even right now, there doesn't even seem to be any kind of consensus that an increase in the blocksize is actually needed prior to seg wit going live for some time.
luke JR was explaining that segwit has changed the parameter. there is no longer a MAX_BLOCK_SIZE variable in segwit
(well its there just renamed)
instead segwit is set as MAX_BLOCK_SERIALIZED_SIZE = 4000000 (all data tx and signatures)
but to not break consensus and cause a hard fork, segwit works by separating the tx and witness
this is done by
MAX_BLOCK_BASE_SIZE = 1000000
meaning traditional transactions(full tx of nonsegwit) and the non-witness(tx part of segwit) fits into MAX_BLOCK_BASE_SIZE
leaving 3mb spare for the signature area.
in short/layman MAX_BLOCK_BASE_SIZE is the same as MAX_BLOCK_SIZE
now to increase the capacity for traditional transactions is to increase the MAX_BLOCK_BASE_SIZE to 2000000
this will however need consensus because it is a hardfork(basically its the same as other imps increasing MAX_BLOCK_SIZE to 2000000)
for luke to forfil his agreement as a *cough*"independent core contributor"*cough* he has to release code for the MAX_BLOCK_BASE_SIZE=2000000
currently this (MAX_BLOCK_BASE_SIZE=1000000(1mb hard block))
https://github.com/bitcoin/bitcoin/blob/395521854efd5804433d57aaf69f46676e4b6efc/src/consensus/consensus.hneeds to change