@AgentofCoin
Your description of the mechanism seems correct, but I still see zero value to the network for non-mining nodes that don't upgrade.
You could say that a non-upgraded miner creating blocks of non-sw tx adds to the network hashrate, but on the other hand its still SPV mining, which is something that should probably be discouraged.
For validation, the idea of 'well, its sort of compatible, sometimes, on certain transactions' makes "backwards compatibility" a misnomer.
IMO, the non-upgraded nodes (Level 2) still retain their security value for the network,
even though they are oblivious to the SegWit txs individually. They are still verifying
everything other than SegWit data. The SegWit proposal is only backwards compatible
as to that it does not break Level 2 nodes verifying nonsegwit txs, not that Level 2 nodes
can also verify SegWit, since that would be impossible.
One purpose of a SF is to preserve the old nodes, and in that sense the "new features"
are "backwards compatible" with Level 2, as that they do not violate Level 2 rules. That
is it basically.
To be really simple, SegWit nodes (Level 3) could be compared to a secondary layer
on top of the preexisting layer (Level 2)
**. Over time, as more and more verifying node
versions enter and leave the network, Level 2 nodes will naturally phase out as Level 3
nodes (and possibly higher levels) increase. The purpose of this SF method is the slow
natural transition of security from one Level to the next Level. Eventually at some point
in the future, our current Level 2 nodes will become fully extinct and then Level 3 Nodes
will become the new Level 2 nodes. So security is retained and features are added without
any loss (in theory). This is intentional, playing the long game, and the opposite of a HF,
which breaks and rejects all Level 2 immediately and mandates Level 3 full compliance
(which automatically becomes the new Level 2, since old Level 2 is banned).
The issue that Level 2 nodes not being fully validating as equally as Level 3 nodes is
semi correct, but is not the actual intention of this type of design. It is basically, "some
network validation for now (Level 2), as we transition into full validation of the new script
format later (Level 3)". This design mechanism allows for slow upgrade change transitions
without sacrificing the node network's "overall" security. As I stated above, eventually
Level 2 nodes will fade away just as the Satoshi clients and Bitcoin QT clients have
faded away over time as new things developed.
Hopefully this isn't confusing. This is my understanding.
**[ As a side note: Level 1 are the miner's themselves and prior to the centralization
of mining by ASIC and etc, Level 1 nodes were intended to be the only Level. Since
mining became exploitable in many different ways (SPV and etc) and that the cost of
competition for average people became too high too soon, the verification system naturally
split between mining nodes and non-mining (verifying) nodes. These non-mining nodes only
exist for the sake of double checking the miner's work and holding them in protocol
compliance. In the event of massive miner collusion or etc, the only remedy users have
are non-mining nodes alerting them to such. This is why Level 2 Nodes must remain
individual, decentralized and unregulatable, since without them, mining becomes truly
illusionary since they become trusted within a system intended to be based on trustlessness.
This is the contradiction. Due to this natural occurrence that Satoshi did not fully anticipate,
Level 2 Nodes manifested as a byproduct of this unfortunate failure. (Interestingly, it may
be that non-mining (verifying) nodes who are purely altruistic, properly balance out the
nature of the pure greed aspect of mining. The greed nodes vs the altruistic nodes may
have created a more balanced system than Satoshi original envisioned (pure greed)).
So, in theory the original Satoshi experiment (and Nakamoto Consensus) failed within
short time due to exploits, but a new answer immediately arose to plug the error and that
safety plug are Level 2 Nodes. This is one of the reasons for the current scaling debate.
Scaling, if done improperly, will prevent Level 2 Nodes from operating as this safety
plug, and may irrevocably remove them forever. In that future scenario, miners become
trusted parties and have free will to do anything they wish. Bitcoin then falls back into the
failure that Level 2 Nodes naturally arose to fix in 2010-2011. The Protocols will become
as illusionary as like the Financial World's Credit System, making Bitcoin a shadow
of its former self, absolutely worthless in monetary value, and IMO an insult to what
Satoshi was attempting to solve originally. ]
