Quote from: d5000 on January 21, 2017, 08:34:35 PM

Thank you @franky1, I'll read it. I don't consider LN an ideal solution for Bitcoin to scale, too. I think it's over-complicated, but for micropayments it may work (I consider it appropiate for payments where you, using fiat, would use a prepaid/gift card).

What I was interested in, however, was an explanation in layman's terms if the Segwit concept itself (not LN and also not the "soft fork" itself as you point out in your other posting) had some negative impact on security or usability of Bitcoin. I know the Bitcoin Core page about Costs and Risks, but it doesn't seem like this list is really a list of "disadvantages" as most of it could be applied to all larger "jumps" in software development. If you or another participant of this discussion have a link for me, I'd be grateful if you share it.

as you can see by the post i done with the couple images included up above.

say you made a transaction that used segwit keys.. P2WPKH...
if you relayed it to an old node. (while network is in current format) the old node see it as an anyonecanspend, which would get rejected and not relayed by an old node (while unconfirmed). but if a pool was to include it. even without segwit working. the pool could confirm it and then spend it.

this is why core need pool acceptance predominently and are willing to split off the opposing pools.
this is why core want segwit tx relays to be where segwit nodes are a separate layer to the old nodes. where segwit 'translates' the transactions into something old nodes cant mess with.

ok explaining the bit after the first picture in last post i made.

hopefully the color coding shows how core envision the node connections of the network will change. where the purple line is the "special white listed old nodes".. where the segwit nodes translate the block into standard 1mb block format

concentrating on left side below.. the red pool at the centre. going outwards

EDIT:gmaxwell buzzwords
downstream(old) <-> upstream(segwit) <-> pool

upstream(segwit) <-> pool<-> downstream(old)

however not many people will manually want to white list those old nodes and will think 'the pool or someone else can do it', which obviously will be the pools because of them being a segwit node, is able to whitelist some old nodes
and so
the image on the right is more so what the network would look like by adding in some context of human psychy .. bar maybe a couple purple lines that might go between the segwit nodes and the old nodes from some people who may make the extra effort

all because sending a segwit transaction unconfirmed to old nodes and pools not segwit ready has issues.
like i said they have not even got intention to release a wallet with segwit keys (p2wpkh p2wsh). and wont release it until the pools are ready and they have some segwit nodes to act as the gate keepers and translate the data to old nodes(in the left utopia)