On Tue, Dec 22, 2020 at 12:22:37AM +0000, Pieter Wuille via bitcoin-dev wrote:
> 
> Re-reading your proposed text, I'm wondering if the "consensus-only validation" extension is intended to replace the inconclusive-due-to-consensus-and-standardness-differ state. If so, I don't think it does, and regardless it doesn't seem very useful.
> 
> What I'm suggestion could be specified this way:
> * If validator understands the script:
>   * If signature is consensus valid (as far as the validator knows):
>     * If signature is not known to trigger standardness rules intended for future extension (well-defined set of rules listed in BIP, and revisable): return valid
>     * Otherwise: return inconclusive
>   * Otherwise: return invalid
> * Otherwise: return inconclusive
> 
> Or in other words: every signature has a well-defined result (valid, invalid, inconclusive) + validators may choose to report inconclusive for anything they don't understand.
> 
> This has the property that as long as new consensus rules only change things that were covered under for-future-extension standardness rules, no two validators will ever claim valid and invalid for the same signature. Only valid+inconclusive or invalid+inconclusive.
>

I like it!

My thinking regarding standardness vs consensus rules was essentially that
I wanted to enforce the included standardness rules for anti-malleability
reasons, i.e. the hope that for "normal scripts" we would get strong signatures,
which may be important for anti-DoS reasons. (What I mean by this is that
if you can easily create mutations of signatures, it may confuse software
in similar ways to the Gox-era malleability attacks on wallet software of
the time.) But conversely, it is hard to enforce these rules as an
implementor, because libbitcoinconsensus does not expose them. So allowing
both forms of validation, to me, was an attempt to encourage adoption
rather than anything principled.

I didn't even consider the idea that validators should be able to signal "this
signature appears to use future consensus rules", although I should have been
clued in by your "upgradeable rules" language that this was your goal. Now that
you say this, it's obvious that this is desireable, and also obvious that using
the "inconclusive" state is an elegant way to achieve this.

I also agree that "confirming validators should never disagree on valid vs
invalid" is a good design goal and we should make that explicit.


I'll add a commit to my PR at https://github.com/bitcoin/bips/pull/1048 which
adds these thoughts.

-- 
Andrew Poelstra
Director of Research, Blockstream
Email: apoelstra at wpsoftware.net
Web:   https://www.wpsoftware.net/andrew

The sun is always shining in space
    -Justin Lewis-Webster