op_ctv still has no technical objections

[Erik Aronesty]

It's been many years and there's been a lot of discussion about various covenants 

I think one of the biggest problems is everyone has to insist on their baby is the best baby. 

op_ctv is quite literally not the best at anything.  That's the whole point.  It's non-recursive, can't be used for strange or dangerous things, and can be used to emulate a lot of other opcodes. 

It's adequate.  And I don't think we want anything "better" than adequate the first time around. lnhance is more comprehensive.  but also it's so much harder to reason about three separate op codes and what the attack surface could be.

I don't think it's possible to optimize a series of covenants for all possible scenarios.  Easy to make them too powerful and now nodes are doing too much work and we're attracting the kind of network activity that nobody wants.  

Fortunately the risk of CTV is fairly low.  It's always possible to turn it off (no new tx)... if there's a game theory issue. 

I don't think there's any particular rush, but we could lose a lot of fees and support for miners if Bitcoin continues to do what it is doing now... scaling almost entirely in custodial systems.  That's also just not the Bitcoin that anyone loves.

At this point it feels like it's "perfect is the enemy of the good".  

We have an old and rather well tested pull request that is only a handful of lines of code that everyone has scrutinized a million ways. 

I don't think we're getting that for any other covenant opcode.  

[conduition]

As someone who has had a merely passive interest in covenants tech, I can confidently say that OP_CTV is probably the only covenants proposal whose effects I can confidently say I fully grasp. It's also easy to explain to others. Not saying i'm not in favor of more complex multi-pronged upgrades like LNHANCE, just saying I don't fully understand their opcode interplay enough to say yay/nay. Which is maybe an under-represented argument in favor of plain OP_CTV.

regards,

conduition

[/dev /fd0]

Hi Erik,

We can organize a meeting to discuss the activation parameters and build an activation client. If enough economic nodes and miners run the activation client it will be activated with no further politics or drama.

#ctv-csfs-activation IRC channel can be used for the meeting.

/dev/fd0
floppy disk guy

--
You received this message because you are subscribed to the Google Groups "Bitcoin Development Mailing List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to bitcoindev+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/bitcoindev/CAJowKg%2BcCoocSEYsTT3bLwte%3D-3Kbzo5k6YT--UnDwzoZPF1wQ%40mail.gmail.com.

[Erik Aronesty]

Hello, out of this meeting it was generally agreed that it's appropriate to propose a conservative activation mechanism, using BIP 9, with a long activation window and high miner threshold.  

Question:  Do we need a new BIP, or should we update BIP119 with the new activation parameters?

[moonsettler]

Hi All,

Just a small remark

> lnhance is more comprehensive. but also it's so much harder to reason about three separate op codes and what the attack surface could be.

It's 4 opcodes, but ofc it's safe to ignore INTERNALKEY when it comes to unexpected interactions.
We have spent basically a whole year on walking in circles with various opcode combos.

We came up with a set of threshold rules that make sense as an evaluation framework:
- Fine-grained introspection
- State-carrying covenants
- Bigint operations
- New arithmetic capabilities using lookup tables

These are key "ingredients" to exogenous asset protocols that are script interactible and novel bridge
constructions, that might interact badly with mining decentralization.

Many other proposals instantly violate some or all of them, not LNhance.
To this day I haven't seen anyone come up with anything remotely scary with CTV+CSFS+PC.

I would like to encourage people to take the time and try to come up with anything "nasty".

BR,
moonsettler


Sent with Proton Mail secure email.