I just wonder why Gavin is so sure that "thousands of Classic nodes will appear". They didn't appear for XT. Maybe, he knows that "a little help from unknown friends" is sure to come...
I thought the post was pretty clear, it was " People are committing to spinning up thousands" and 'not thousands of people are committed to spinning up a node'-- it's a planned sybil attack-- and that is also what I've seen from this rise of "classic" "nodes". There are several less obvious node count measures that don't show the growth.
AFAICT, the latest strategy is to fake out the node counts with large numbers of sybils and then try to use that to pressure miners into adopting classic; which would then pressure actual users to go along with it. This isn't going to work, and most charitable way I can explain the strategies used by the people frantically pushing for a controversial hardfork is that the people involved in these forks keep thinking that everyone else in Bitcoin is stupid. How else can you explain the faux urgency-- that almost no one bought-- or the bait and switch policies for miners-- to the cheap characterization that Bitcoin Core is all blockstream and so on?
All these nonsense and attacks frustrate me-- they waste a tone of time and energy that could be used driving Bitcoin forward.
Why we cannot add code to filter non-core nodes from connecting to core nodes? Classic would have to update their code to look like a core node and by doing so classic nodes will not be detectable. It could be based on UA, version or protocol level, or set of other fields. Nodes can be configured to accept connections above a certain set and refuse everything else.
The classic will try to piggyback on your releases. So force them to adopt your code and they become invisible on the network.