"You can just do things!" Well, not really. Core has a tremendous amount of latent power for at least two reasons:

(i) every business has basically hardcoded the use of that implementation with levels of risk aversion surpassing "never got fired for buying IBM," creating an amazing de facto monopoly on economic nodes, and

(ii) bitcoin is so complicated tha t any non-developer inherently must delegate to some technical authority they trust, and Core winds up (for most sane people) being the target of that delegation.

This is a tremendous amount of sticky power. To not acknowledge that is naive.

Yes, in theory a large part of the ecosystem could decide to run an alternate implementation - but the switching/coordination/incidental-risk costs there are huge. It would take a 5-10x reward:risk proposition for most businesses to switch.

And meanwhile, the Core line is so often "well running another implementation would be dangerous! think of the possible consensus incompatibilities!" Convenient.

This single chokepoint is going to be a big problem for bitcoin, even if the filters issue isn't per se an epitome of it."

Reply to this note

Please Login to reply.

Discussion

Nodes are run by the people, and simply not upgrading to Core v30 will have an impact; moving to Knots is not required (although that’s what I am now running).

Not upgrading is not enough in the long run

If everyone moved to Knots today the miners would have no choice but to comply by publishing blocks that are compliant with the rules of the network.

But… agreed. Still, running Knots now is a worthy signal for what needs to come. This is a wakeup call for everyone to put aside their assumptions that Bitcoin is perfectly indestructible, and to start asking questions about the node software(s) that makes Bitcoin run.