I'm curious to hear clarification on most of Luke's non-activation related comments. 

> I would ideally like to see fully implemented BIPs for at least one of these

While that would be interesting, I think that's a heavy burden to be placed on this BIP. More in depth exploration would be helpful, but a fully implemented BIP I think is more than necessary.

> Why is it a problem for them to use an Eltoo-like protocol?

I think he was saying it is a problem *unless* its an eltoo-like protocol. Why I'm not sure. Maybe you can clarify Jeremy?

> It's not clear to me that this holds if OP_CAT or OP_SHA256STREAM get added.

Even were these opcodes to be implemented in bitcoin, a script writer could choose to not use them, making it still possible to use CTV to create covenant chains with a finite number of steps. 

>  w.r.t. the language cleanups... the legal definition of covenant ... I do think things like CLTV/CSV are covenants 

Maybe it would be useful to specify that these are "child covenants" or "inherited covenants" or something like that, since unlike things like CLTV, CTV and similar proposed opcodes place restrictions on the child output of the output containing the opcode call, which is the interesting unique behavior. Tho I don't think we need to be bound to the legal or dictionary definition in usage of the word covenant in the realm of bitcoin - its gonna have its own definition in this context anyway. 

Thank you Eric for pointing out the factual errors in LukeJr's mention and implications around BIP8. The fact is that the ST pull request was described as "BIP9-based". TBH BIP8 is also BIP9 based, and ST is its own thing that's neither BIP8 nor BIP9, so characterization one way or another is moot IMO. In any case, I also agree with Michael that this isn't the place to have a long discussion about activation method. That discussion should be kept separate. I'd go so far to say that BIPs should not advocate for any particular activation method, but should only go so far as to mention what types of activation methods are possible (if some types aren't possible for some reason). Separation of concerns would be very useful on that front to reduce noise in conversations.

Thanks,
BT


On Wed, Jan 19, 2022 at 6:37 AM Michael Folkson via bitcoin-dev <bitcoin-dev@lists.linuxfoundation.org> wrote:
Eric, Luke

Can I request that you don't discuss activation methods for future soft forks on a thread for CTV BIP review? I (and a number of others [0]) do not support an upcoming activation attempt of standalone OP_CTV. If you want to discuss activation methods for soft forks generally it would be much better if you set up a separate thread. OP_CTV is not the only current soft fork proposal and there will likely be more.

The activation discussion for Taproot was deliberately kept separate from the review of the Taproot BIPs and implementation. It only commenced once there was overwhelming community consensus for the soft fork to be activated (months after in fact). Though you are free to discuss whatever topics you wish (obviously) discussing soft fork activation methods on a OP_CTV thread might give the mistaken impression that OP_CTV is the next soft fork to be activated which is mere speculation at this point. In an ideal world the promoters of OP_CTV would follow the strong precedent set by the authors and contributors to the Taproot BIPs but regrettably that seems to have gone out the window at this point.

Thanks
Michael

[0]: https://gist.github.com/michaelfolkson/352a503f4f9fc5de89af528d86a1b718
--
Michael Folkson
Email: michaelfolkson at protonmail.com
Keybase: michaelfolkson
PGP: 43ED C999 9F85 1D40 EAF4 9835 92D6 0159 214C FEE3

‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐

On Tuesday, January 18th, 2022 at 11:00 PM, Eric Voskuil via bitcoin-dev <bitcoin-dev@lists.linuxfoundation.org> wrote:

> -----Original Message-----
>
> From: Luke Dashjr luke@dashjr.org
>
> Sent: Tuesday, January 18, 2022 2:10 PM
>
> To: eric@voskuil.org
>
> Cc: 'Bitcoin Protocol Discussion' bitcoin-dev@lists.linuxfoundation.org
>
> Subject: Re: [bitcoin-dev] CTV BIP review
>
> On Tuesday 18 January 2022 22:02:24 eric@voskuil.org wrote:
>
> > The only material distinction between BIP9 and BIP8 is that the latter
> >
> > may activate without signaled support of hash power enforcement.
> >
> > As unenforced soft forks are not "backward compatible" they produce a
> >
> > chain split.
>
> Enforcement of the Bitcoin consensus protocol is by users, not miners.

Given that I stated "hash power enforcement" it is quite clear that this is

in fact only produced by mining. You are misrepresenting my statement to

make an emotional appeal. Without "hash power enforcement", a soft fork is

NOT backward compatible.

"[enforcement of] consensus protocol" is of course by merchants, but that is

not the question at hand. The question is explicitly compatibility. Anyone

can activate a soft fork at any time, but without "hash power enforcement"

soft forks are NOT backward compatible.

> Softforks never produce a chain split. Miners can, and might try to do it

to cause disruption in retaliation, but the softfork itself does not.

Maybe you are trying to split hairs given the fact that blocks are produced

only by miners, so only miners can "cause" a split.

But through not intention ("disruption in retaliation") whatsoever by

mining, a soft fork will result in those activating the rule being split off

the original chain unless majority hash power enforces the rule. The fact

that doing nothing apart from deploying the rule will result in a split is

the very definition of NOT compatible.

I assume you will argue that the original chain is not "valid" and therefore

irrelevant (as if no chain split occurred). But again the point is about

compatibility. The appearance of multiple chains, which appear valid

according to either the previous or new rules, is obviously the

incompatibility.

I shouldn't have to point this out, but observed chain splits have occurred

in more the one large scale soft fork deployment. These splits have only

been resolved through hash power enforcement. In 2010 it took 51 blocks

before the current chain took the lead. In 2012 minority chains persisted

for months. The deployment of soft forks caused these splits, NOT the

actions of miners. And unless majority hash power eventually enforces it,

the soft fork branch necessarily dies.

> > It was for this reason alone that BIP8 never gained sufficient
> >
> > support.
>
> BIP 8 in fact achieved consensus for Taproot activation.

Please define "achieved consensus", because by any definition I can imagine,

this is simply untrue.

> > This is one of the most misleading statements I've seen here. It's not
> >
> > technically a lie, because it states what "should" happen. But it is
> >
> > clearly intended to lead people to believe that BIP8 was actually used
> >
> > ("again") - it was not. ST was some technical tweaks to BIP9.
>
> BIP 8 was used to activate Taproot.

No, it wasn't. I find it hard to imaging how you rationalize such grossly

misleading statements.

> > The outright deception around this one topic has led to significant
> >
> > unnecessary conflict in the community. Make your argument, but make it
> >
> > honestly.
>
> You are the one attempting to deceive here.

That is for others to decide. I appreciate your responses above, since they

certainly help clarify what is happening here.

e

bitcoin-dev mailing list

bitcoin-dev@lists.linuxfoundation.org

https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev
_______________________________________________
bitcoin-dev mailing list
bitcoin-dev@lists.linuxfoundation.org
https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev