MIME-Version: 1.0
References: <Y0ZTtlRSBihNN9+v@erisian.com.au>
 <0hpdGx-1WbZdG31xaMXGHKTCjJ2-0eB5aIXUdsp3bqI1MlCx6TMZWROwpl1TVI5irrBqRN2-ydM6hmf3M5L-7ZQfazbx66oameiWTHayr6w=@wuille.net>
In-Reply-To: <0hpdGx-1WbZdG31xaMXGHKTCjJ2-0eB5aIXUdsp3bqI1MlCx6TMZWROwpl1TVI5irrBqRN2-ydM6hmf3M5L-7ZQfazbx66oameiWTHayr6w=@wuille.net>
From: Dario Sneidermanis <dario@muun.com>
Date: Wed, 12 Oct 2022 18:44:13 -0300
Message-ID: <CAKiPDnQqEQd+tcPo+PVDvyHW6kz+7Q2HikyzVtjnzpRFT4qa2Q@mail.gmail.com>
To: Pieter Wuille <bitcoin-dev@wuille.net>
Content-Type: multipart/alternative; boundary="000000000000dbaec805eadd4b93"
Cc: Bitcoin Protocol Discussion <bitcoin-dev@lists.linuxfoundation.org>,
 Anthony Towns <aj@erisian.com.au>
Subject: Re: [bitcoin-dev] [Opt-in full-RBF] Zero-conf apps in immediate
	danger
Precedence: list

--000000000000dbaec805eadd4b93
Content-Type: text/plain; charset="UTF-8"

Hello Pieter,

Thanks for taking the time to comment! I'll answer inline.

On Wed, Oct 12, 2022 at 2:51 PM Pieter Wuille <bitcoin-dev@wuille.net>
wrote:
> I certainly recognize that adding the flag is a likely step towards, over
> time, the full RBF policy becoming more widely adopted on the network.
That is
> presumably the reason why people are in favor of having the flag, even
default
> off - including me. I believe that policy's adoption is inevitable
eventually,
> but the speed at which that is achieved is certainly a function of
> availability and adopted of software which provides the option.

As stated in the original posting, I believe too that a full-RBF network is
not
only inevitable but also desirable. Miner incentives will eventually win,
so we
should address them before they fully kick in (ie. before transaction fees
become a meaningful portion of the block reward).

> So I have a hard time imagining how it would change anything
*immediately* on
> the network at large (without things like default on and/or preferential
> peering, ...), but I still believe it's an important step.

Notice that I'm not saying this changes anything immediately on the network
at
large. In fact, it is unlikely that the opt-in flag alone would be enough to
migrate the network at large to full-RBF.

There's a real possibility that, after deployment of the opt-in flag,
either no
meaningful hashing power adopts it or no connected component of
transaction-relaying nodes adopts it. If that's the case, the deployment
won't
help nodes participating in multi-party funded transactions protect against
the
class of attacks described in [1] (which was, as I understand, the original
intention of #25353).

If that's not the case, it means that at least some meaningful hashing power
adopted it and that there exist some connected components of
transaction-relaying nodes that adopted it. This is certainly far from
having
wide adoption of full-RBF in the network at large. However, once we reach
that
minimal level of adoption in the mining and relaying layers, any node on a
full-RBF connected component can send an on-chain payment to an application
and
then get a replacement mined. That is, applications that accept incoming
on-chain payments from untrusted parties can be immediately exposed to
full-RBF
transaction replacements, even if they didn't opt into full-RBF in their
nodes.

In an adversarial setting, such as the one for zero-conf applications (as
defined in the original posting), this increases the risk of an attack
substantially, making the entire strategy moot.

> In my view, it is just what I said: a step towards getting full RBF on the
> network, by allowing experimentation and socializing the notion that
> developers believe it is time.

Those are worthy goals. I believe we can design a deployment strategy for
full-RBF that takes them into account and, at the same time, gives a clear
timeline for any affected application to adapt.

This could be one such proposal:

1. We activate opt-in full-RBF on testnet now.
2. We commit now (in the code) to a block height in the future at which
opt-out
   full-RBF will activate on mainnet.

The first point will allow for experimentation and give a testing ground to
all
affected applications. The second point socializes the notion that
developers
believe it is time, giving a clear message and timeline for anyone affected
to
adapt. It also has the benefit that many more nodes will have upgraded by
the
time we reach the activation block height, making the transition to a
full-RBF
network much more predictable and easy to reason about.

There's an argument to be made that the miner incentive incompatibility
problem
of a non-full-RBF network gets measurably worse at the time of the next
halving.
To fix this, we could choose any block height before that, giving a clear
and
predictable transition timeline.

[1]
https://lists.linuxfoundation.org/pipermail/lightning-dev/2021-May/003033.html

On Wed, Oct 12, 2022 at 1:11 PM Pieter Wuille <bitcoin-dev@wuille.net>
wrote:

> On Wednesday, October 12th, 2022 at 1:42 AM, Anthony Towns <
> aj@erisian.com.au> wrote:
>
> > On Tue, Oct 11, 2022 at 04:18:10PM +0000, Pieter Wuille via bitcoin-dev
> wrote:
> >
> > > On Friday, October 7th, 2022 at 5:37 PM, Dario Sneidermanis via
> bitcoin-dev bitcoin-dev@lists.linuxfoundation.org wrote:
> > >
> > > > Thanks for the fast answer! It seems I missed the link to the PR,
> sorry for the
> > > > confusion. I'm referring to the opt-in flag for full-RBF from #25353
> > > > (https://github.com/bitcoin/bitcoin/pull/25353).
> > > > It is not clear to me why you believe the merging of this particular
> pull request poses an immediate risk to you.
> >
> >
> > Did you see the rest of Dario's reply, bottom-posted after the quoted
> > text? Namely:
>
> Oh, my mail client for some reason chose to hide all that. Dario, I'm
> sorry for missing this; I see now that you were certainly aware of what the
> PR under consideration did.
>
> Further comments inline.
>
> > On Fri, Oct 07, 2022 at 06:37:38PM -0300, Dario Sneidermanis via
>
> > > The question then is whether an opt-in flag for full-RBF will have
> enough
> > > adoption to get us from 1 to 2. If it isn't, then #25353 won't meet its
> > > objective of allowing nodes participating in multi-party funding
> protocols
> > > to assume that they can rely on full-RBF. If it is, then zero-conf
> applications
> > > will be at severe risk (per the logic in the initial email).
>
> >
> >
> > That logic seems reasonably sound to me:
> >
> > - if adding the option does nothing, then there's no point adding it,
> > and no harm in restricting it to test nets only
> >
> > - if adding the option does do something, then businesses using zero-conf
> > need to react immediately, or will go from approximately zero risk of
> > losing funds, to substantial risk
> >
> > (I guess having the option today may allow you to manually switch your
> > node over to supporting fullrbf in future when the majority of the
> network
> > supports it, without needing to do an additional upgrade in the meantime;
> > but that seems like a pretty weak benefit)
>
> I certainly recognize that adding the flag is a likely step towards, over
> time, the full RBF policy becoming more widely adopted on the network. That
> is presumably the reason why people are in favor of having the flag, even
> default off - including me. I believe that policy's adoption is inevitable
> eventually, but the speed at which that is achieved is certainly a function
> of availability and adopted of software which provides the option.
>
> That said, I think it's a bit of a jump to conclude that the only two
> options are that either the existence of the flag either has no effect at
> all, or poses an immediate threat to those relying on its absence. In my
> view, it is just what I said: a step towards getting full RBF on the
> network, by allowing experimentation and socializing the notion that
> developers believe it is time. So I have a hard time imagining how it would
> change anything *immediately* on the network at large (without things like
> default on and/or preferential peering, ...), but I still believe it's an
> important step.
>
> Cheers,
>
> --
> Pieter
>
>

--000000000000dbaec805eadd4b93
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">Hello Pieter,<br><br>Thanks for taking the time to comment=
! I&#39;ll answer inline.<br><br>On Wed, Oct 12, 2022 at 2:51 PM Pieter Wui=
lle &lt;<a href=3D"mailto:bitcoin-dev@wuille.net">bitcoin-dev@wuille.net</a=
>&gt; wrote:<br>&gt; I certainly recognize that adding the flag is a likely=
 step towards, over<br>&gt; time, the full RBF policy becoming more widely =
adopted on the network. That is<br>&gt; presumably the reason why people ar=
e in favor of having the flag, even default<br>&gt; off - including me. I b=
elieve that policy&#39;s adoption is inevitable eventually,<br>&gt; but the=
 speed at which that is achieved is certainly a function of<br>&gt; availab=
ility and adopted of software which provides the option.<br><br>As stated i=
n the original posting, I believe too that a full-RBF network is not<br>onl=
y inevitable but also desirable. Miner incentives will eventually win, so w=
e<br>should address them before they fully kick in (ie. before transaction =
fees<br>become a meaningful portion of the block reward).<br><br>&gt; So I =
have a hard time imagining how it would change anything *immediately* on<br=
>&gt; the network at large (without things like default on and/or preferent=
ial<br>&gt; peering, ...), but I still believe it&#39;s an important step.<=
br><br>Notice that I&#39;m not saying this changes anything immediately on =
the network at<br>large. In fact, it is unlikely that the opt-in flag alone=
 would be enough to<br>migrate the network at large to full-RBF.<br><br>The=
re&#39;s a real possibility that, after deployment of the opt-in flag, eith=
er no<br>meaningful hashing power adopts it or no connected component of<br=
>transaction-relaying nodes adopts it. If that&#39;s the case, the deployme=
nt won&#39;t<br>help nodes participating in multi-party funded transactions=
 protect against the<br>class of attacks described in [1] (which was, as I =
understand, the original<br>intention of #25353).<br><br>If that&#39;s not =
the case, it means that at least some meaningful hashing power<br>adopted i=
t and that there exist some connected components of<br>transaction-relaying=
 nodes that adopted it. This is certainly far from having<br>wide adoption =
of full-RBF in the network at large. However, once we reach that<br>minimal=
 level of adoption in the mining and relaying layers, any node on a<br>full=
-RBF connected component can send an on-chain payment to an application and=
<br>then get a replacement mined. That is, applications that accept incomin=
g<br>on-chain payments from untrusted parties can be immediately exposed to=
 full-RBF<br>transaction replacements, even if they didn&#39;t opt into ful=
l-RBF in their nodes.<br><br>In an adversarial setting, such as the one for=
 zero-conf applications (as<br>defined in the original posting), this incre=
ases the risk of an attack<br>substantially, making the entire strategy moo=
t.<br><br>&gt; In my view, it is just what I said: a step towards getting f=
ull RBF on the<br>&gt; network, by allowing experimentation and socializing=
 the notion that<br>&gt; developers believe it is time.<br><br>Those are wo=
rthy goals. I believe we can design a deployment strategy for<br>full-RBF t=
hat takes them into account and, at the same time, gives a clear<br>timelin=
e for any affected application to adapt.<br><br>This could be one such prop=
osal:<br><br>1. We activate opt-in full-RBF on testnet now.<br>2. We commit=
 now (in the code) to a block height in the future at which opt-out<br>=C2=
=A0 =C2=A0full-RBF will activate on mainnet.<br><br>The first point will al=
low for experimentation and give a testing ground to all<br>affected applic=
ations. The second point socializes the notion that developers<br>believe i=
t is time, giving a clear message and timeline for anyone affected to<br>ad=
apt. It also has the benefit that many more nodes will have upgraded by the=
<br>time we reach the activation block height, making the transition to a f=
ull-RBF<br>network much more predictable and easy to reason about.<br><br>T=
here&#39;s an argument to be made that the miner incentive incompatibility =
problem<br>of a non-full-RBF network gets measurably worse at the time of t=
he next halving.<br>To fix this, we could choose any block height before th=
at, giving a clear and<br>predictable transition timeline.<br><br>[1] <a hr=
ef=3D"https://lists.linuxfoundation.org/pipermail/lightning-dev/2021-May/00=
3033.html">https://lists.linuxfoundation.org/pipermail/lightning-dev/2021-M=
ay/003033.html</a><br></div><br><div class=3D"gmail_quote"><div dir=3D"ltr"=
 class=3D"gmail_attr">On Wed, Oct 12, 2022 at 1:11 PM Pieter Wuille &lt;<a =
href=3D"mailto:bitcoin-dev@wuille.net">bitcoin-dev@wuille.net</a>&gt; wrote=
:<br></div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.=
8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">On Wednesday, =
October 12th, 2022 at 1:42 AM, Anthony Towns &lt;<a href=3D"mailto:aj@erisi=
an.com.au" target=3D"_blank">aj@erisian.com.au</a>&gt; wrote:<br>
<br>
&gt; On Tue, Oct 11, 2022 at 04:18:10PM +0000, Pieter Wuille via bitcoin-de=
v wrote:<br>
&gt; <br>
&gt; &gt; On Friday, October 7th, 2022 at 5:37 PM, Dario Sneidermanis via b=
itcoin-dev <a href=3D"mailto:bitcoin-dev@lists.linuxfoundation.org" target=
=3D"_blank">bitcoin-dev@lists.linuxfoundation.org</a> wrote:<br>
&gt; &gt; <br>
&gt; &gt; &gt; Thanks for the fast answer! It seems I missed the link to th=
e PR, sorry for the<br>
&gt; &gt; &gt; confusion. I&#39;m referring to the opt-in flag for full-RBF=
 from #25353<br>
&gt; &gt; &gt; (<a href=3D"https://github.com/bitcoin/bitcoin/pull/25353" r=
el=3D"noreferrer" target=3D"_blank">https://github.com/bitcoin/bitcoin/pull=
/25353</a>).<br>
&gt; &gt; &gt; It is not clear to me why you believe the merging of this pa=
rticular pull request poses an immediate risk to you.<br>
&gt; <br>
&gt; <br>
&gt; Did you see the rest of Dario&#39;s reply, bottom-posted after the quo=
ted<br>
&gt; text? Namely:<br>
<br>
Oh, my mail client for some reason chose to hide all that. Dario, I&#39;m s=
orry for missing this; I see now that you were certainly aware of what the =
PR under consideration did.<br>
<br>
Further comments inline.<br>
<br>
&gt; On Fri, Oct 07, 2022 at 06:37:38PM -0300, Dario Sneidermanis via<br>
<br>
&gt; &gt; The question then is whether an opt-in flag for full-RBF will hav=
e enough<br>
&gt; &gt; adoption to get us from 1 to 2. If it isn&#39;t, then #25353 won&=
#39;t meet its<br>
&gt; &gt; objective of allowing nodes participating in multi-party funding =
protocols<br>
&gt; &gt; to assume that they can rely on full-RBF. If it is, then zero-con=
f applications<br>
&gt; &gt; will be at severe risk (per the logic in the initial email).<br>
<br>
&gt; <br>
&gt; <br>
&gt; That logic seems reasonably sound to me:<br>
&gt; <br>
&gt; - if adding the option does nothing, then there&#39;s no point adding =
it,<br>
&gt; and no harm in restricting it to test nets only<br>
&gt; <br>
&gt; - if adding the option does do something, then businesses using zero-c=
onf<br>
&gt; need to react immediately, or will go from approximately zero risk of<=
br>
&gt; losing funds, to substantial risk<br>
&gt; <br>
&gt; (I guess having the option today may allow you to manually switch your=
<br>
&gt; node over to supporting fullrbf in future when the majority of the net=
work<br>
&gt; supports it, without needing to do an additional upgrade in the meanti=
me;<br>
&gt; but that seems like a pretty weak benefit)<br>
<br>
I certainly recognize that adding the flag is a likely step towards, over t=
ime, the full RBF policy becoming more widely adopted on the network. That =
is presumably the reason why people are in favor of having the flag, even d=
efault off - including me. I believe that policy&#39;s adoption is inevitab=
le eventually, but the speed at which that is achieved is certainly a funct=
ion of availability and adopted of software which provides the option.<br>
<br>
That said, I think it&#39;s a bit of a jump to conclude that the only two o=
ptions are that either the existence of the flag either has no effect at al=
l, or poses an immediate threat to those relying on its absence. In my view=
, it is just what I said: a step towards getting full RBF on the network, b=
y allowing experimentation and socializing the notion that developers belie=
ve it is time. So I have a hard time imagining how it would change anything=
 *immediately* on the network at large (without things like default on and/=
or preferential peering, ...), but I still believe it&#39;s an important st=
ep.<br>
<br>
Cheers,<br>
<br>
-- <br>
Pieter<br>
<br>
</blockquote></div>

--000000000000dbaec805eadd4b93--