clarity
soft and hard is simply
soft: pool only vote
hard: nodes and pools vote
below these umbrella terms is what could happen.. in both hard and soft it can either continue as one chain. or bilateral split
softfork: consensus - >94% pools no banning/ignoring of minority. result: small 5% orphan drama then one chain. minority unsynced and dead
softfork: controversial - >50% pools no banning/ignoring of minority. result: long big% orphan drama then one chain. minority unsynced and dead
softfork: bilateral split - intentionally ignoring/banning opposing rules and not including them. result: 2 chains
hardfork: consensus - >94% nodes, then >94% pools no banning/ignoring of minority. result: 5% orphan drama then one chain. minority unsynced / dead
hardfork: controversial - >50% nodes, then >50% pools no banning/ignoring of minority. result: big% orphan drama then one chain. minority unsynced / dead
hardfork: bilateral split - intentionally ignoring/banning opposing rules and not including them. result: 2 chains
segwit doesnt offer 4mb of real usable space for everyone.
segwit doesnt offer 4mb of real usable space instantly when activated
the only way to get to 4mb is to get EVERYONE over to using segwit keys that also have extra bloated future features appended to the end of the tx in the future. such as segwit+confidential payment codes..
emphasis. its not just about getting all pools to be segwit, its not just about all nodes being segwit. its about ALL TRANSACTIONS using segwit keypairs (moving funds from old style addresses). before the benefits can be seen in full
yep if 100% of users used segwit+confidential payment codes. then a segwit block will get a weight of 4mb
yep if 100% of users used segwit without xtra features, then segwit bocks will gt a weight of ~2mb
emphasis needs100% of users using segwit to see all the promises met
but if people stay with native LEAN transactions (real data for data) you wont get as many transactions and wont have the bloat.
however
a hardfork consensus where the REAL block limit(not weight) is increased ALL transactions get extra room and more transaction flow
segwit has not solved attack vectors. its just disarmed innocent sheep
segwits 'features' which segwit has been over promising as fixes of mallicious things. can be bypassed very simply.. just use native transaction types. yep just not using a segwit key means that a person can still malleate and quadratic spam a block
also
although segwit nodes are programmed to be upstream of the network (centralised to the pools) to avoid an issue of segwit itself (anyonecanspend attack) a malicious user can MANUALLY copy an unconfirmed segwit transaction(from segwit) and paste it into an old node and mess with it... yep that is right. they can
within the first 15 minutes Johny of blockstream just ripped apart his own colleagues trust of lightning..(lol)
not verbatim 'because transactions are not on the blockchain confirmed you shouldnt trust them because pools can simply ignore them and not add them to a block'. making LN's close channel/settlement even in a blockstreamers eyes not trusted until its actually on the blockchain.
which as a blockstreamer knowing that things can be messed around with in the mempool (just ignored via malice and recent additions RBF, CPFP) he assumes its ok to not do anything to reduce the risk and thinks increasing the risk of tx's not confirming is ok
at an hour15min-1h 20min they were talking about mining.
firstly asic manufacturers sell gear so that they can use profits to sell more gear and also because profits give the manufacturer free gear. so there is profit in meaning asics. where by the blockstreamer suggested that mining chip manufacturing is not profitable with the lame excuse that manufacturers would refuse to sell gear if it was profitable to just internally mine with what they produce.
secondly the blockstreamer wanted to suggest that asic mining is centralised because there were only 4 manufacturers..
back in 2009-2011 it was either AMD or Intel
back in 2012-2013 it was either Ati or Intel
now its many more manufacturers..
bitmain
canaan creative
bitfury
and anyone can make an asic privately more easily than trying to make something comparable to the CPU/GPU requirement of 2009-2013
even the circuitry is out their in the open far more so than making a complicated CPU/GPU in 2009-2013
.
mining itself there are more POOLS now than 2012-14
those pools are not in just 20 physical locations. and the stratums are also distributed too
so mining is less centralised
as for blockstreamer johny. at 1:29:00-1:31:00 he actually argues that ethereum is very centralised because of proof-of-vitalik..
hmm sticking with that idea, but changing it slightly proof-of-gmaxwell!!!!
if only the question was asked to johny and any blockstreamer.
if gmaxwell moved over to BU and dropped core. how many blockstreamers would follow gmaxwell or stay with core.
