Agreed. I am very wary of efforts to lock people into a “roadmap” when we can barely see what’s in front of us. It’s very unclear to me what is possible and desire-able longer term, but my personal preference is to bite the bullet and build/use something from the ground up that people can actually reason about.
That’s clearly a long way out and there’s no assurance that anything will pan out for practical and reasonable reasons, so thinking shorter term improvements is the best we can do, whether or not we decide to adopt the most mature proposals.
Caveat here being it only emulates a form of SIGHASH_ALL
style APO. This is somewhat limiting and comes with downsides / additional roadblocks to adoption as a slot in replacement. e.g., it will heavily rely on CPFP instead of single tx exo fees which is a bit cheaper and reliably from relay PoV.
It depends on a number factors such as timing with other updates(one update of commitment tx is easier than two), difficulty of swapping out parts, and synergies with other changes to protocols.
I do think it would accelerate updating to PTLCs, f.e., even if symmetry per-se is never adopted. Re-bindable signatures just makes life so much less of a headache when stacking protocols…
I would caution on an overall effort to gate a softfork on a single group’s public declaration of using it promptly. If they’re good, powerful, modular primitives that seem to improve many existing constructions (with proper integration testing!), I think that’s the signal we’re looking for for “utility”.
It’s still in development by multiple companies from my understanding, so it’s hard to judge whether or not it will lead to usage, or what the end usage will be for. I am not interesting in jpegs, but I find trust-minimized bridges to a Shielded CSV-like system quite attractive.
What I’d like on this claim is concrete numbers on the size of improvement. CSFS doesn’t change the security/trust model of BitVM, when something like TXHASH could. And the efficiency claims would have to be weighed against alternative improvements e.g., bignum support alone supposedly brings down the disprove tx from 4MWU to <400kWU.
Unfortunately this is not an effective strategy to gather industry feedback. It is not incumbent on all reviewers to chase down motivations and validation for changes.
Overall, I think CTV+CSFS is an incrementalist improvement, but it is an improvement, built on modular pieces that likely won’t become wholly deprecated in the future if and when we rewrite script or decide we’re not worried about more MEVil. In other words, ignoring the social/technical costs of deployments which are very non-trivial, I believe it would be a positive outcome with limited maintenance burden.