MIME-Version: 1.0
In-Reply-To: <F5F02B94-E094-4C16-80B6-8B0876E423E4@toom.im>
References: <CAAS2fgR84898xD0nyq7ykJnB7qkdoCJYnFg6z5WZEUu0+-=mMA@mail.gmail.com>
	<20170406023123.GA1071@savin.petertodd.org>
	<CA+KqGkqSxeAUZFVFqM_QkEWcGFHgZXwGuOE==7HpXp1+D_Tj3Q@mail.gmail.com>
	<20170406024910.GA1271@savin.petertodd.org>
	<CAFVRnyrqiNY_JOqhv2ysm2WsBMYsU3tTAASAtHzMbA68_9Yx8g@mail.gmail.com>
	<F5F02B94-E094-4C16-80B6-8B0876E423E4@toom.im>
From: David Vorick <david.vorick@gmail.com>
Date: Thu, 6 Apr 2017 08:04:16 -0400
Message-ID: <CAFVRnyrmWvCJ2B9LjLwL6GoqkVJNaMpxD+3Nv9JMCXW3DRE-tg@mail.gmail.com>
To: Jonathan Toomim <j@toom.im>, 
	Bitcoin Protocol Discussion <bitcoin-dev@lists.linuxfoundation.org>
Content-Type: multipart/alternative; boundary=001a11442ae0e72917054c7e4e22
Subject: Re: [bitcoin-dev] BIP proposal: Inhibiting a covert attack on the
 Bitcoin POW function
Precedence: list

--001a11442ae0e72917054c7e4e22
Content-Type: text/plain; charset=UTF-8

On Thu, Apr 6, 2017 at 2:24 AM, Jonathan Toomim via bitcoin-dev <
bitcoin-dev@lists.linuxfoundation.org> wrote:

> Ethically, this situation has some similarities to the DAO fork. We have
> an entity who closely examined the code, found an unintended characteristic
> of that code, and made use of that characteristic in order to gain tens of
> millions of dollars. Now that developers are aware of it, they want to
> modify the code in order to negate as much of the gains as possible.
>
> There are differences, too, of course: the DAO attacker was explicitly
> malicious and stole Ether from others, whereas Bitmain is just optimizing
> their hardware better than anyone else and better than some of us think
> they should be allowed to.
>
> In both cases, developers are proposing that the developers and a majority
> of users collude to reduce the wealth of a single entity by altering the
> blockchain rules.
>
> In the case of the DAO fork, users were stealing back stolen funds, but
> that justification doesn't apply in this case. On the other hand, in this
> case we're talking about causing someone a loss by reducing the value of
> hardware investments rather than forcibly taking back their coins, which is
> less direct and maybe more justifiable.
>
> While I don't like patented mining algorithms, I also don't like the idea
> of playing Calvin Ball on the blockchain. Rule changes should not be
> employed as a means of disempowering and empoverishing particular entities
> without very good reason. Whether patenting a mining optimization qualifies
> as good reason is questionable.
>

Bitmain is blocking protocol upgrades to preserve their mining advantage.
This is quite distinct from someone taking advantage of a visibly broken
and highly toxic smart contract to net themselves tens of millions of
dollars. Further, Bitmain is performing a patented hardware optimization.
The patents mean that other miners are unable to capitalize on these
optimizations. These optimizations are to the tune of 30%. If you give one
player in the mining industry a permanent 30% cost advantage they will
eventually own everything. It's an industry where margins tend towards zero.

The asicboost patent is a direct threat to the health of the Bitcoin
ecosystem, and now we have visible proof. The war against segwit and the
strife with Bitcoin Unlimited was very damaging to the ecosystem, damaging
to the price, and holding back significant improvements and upgrades to the
Bitcoin protocol. I interpret this as a direct attack on the Bitcoin
ecosystem.

I don't know if changing the rules to nullify asicboost is the right move.
I'm sure this won't be the last patent that causes damage to the ecosystem.
But you need to recognize that the issue is not that Bitmain ran a hardware
optimization. It's that hardware optimizations exist which directly inhibit
upgrading the protocol. And it's that hardware optimizations exist
encumbered by patents enough to give one party a decisive advantage in
mining, decisive enough for them to build a single, centralized monopoly.

Each problem is separate, and each problem is significant, and each problem
is fundamental. The DAO attack was a one-time bout of stupidity that
threatened a fixed amount of money. asicboost is an ongoing status that
directly damages Bitcoin's ability to upgrade, and directly damage
Bitcoin's ability to retain any modicum of decentralization in the
hashrate. The DAO issue did neither of these things for ethereum.

--001a11442ae0e72917054c7e4e22
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">On Thu, Apr 6, 2017 at 2:24 AM, Jonathan Toomim via bitcoi=
n-dev <span dir=3D"ltr">&lt;<a target=3D"_blank" href=3D"mailto:bitcoin-dev=
@lists.linuxfoundation.org">bitcoin-dev@lists.linuxfoundation.org</a>&gt;</=
span> wrote:<br><div class=3D"gmail_extra"><div class=3D"gmail_quote"><bloc=
kquote style=3D"margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,=
204);padding-left:1ex" class=3D"gmail_quote">Ethically, this situation has =
some similarities to the DAO fork. We have an entity who closely examined t=
he code, found an unintended characteristic of that code, and made use of t=
hat characteristic in order to gain tens of millions of dollars. Now that d=
evelopers are aware of it, they want to modify the code in order to negate =
as much of the gains as possible.<br>
<br>
There are differences, too, of course: the DAO attacker was explicitly mali=
cious and stole Ether from others, whereas Bitmain is just optimizing their=
 hardware better than anyone else and better than some of us think they sho=
uld be allowed to.<br>
<br>
In both cases, developers are proposing that the developers and a majority =
of users collude to reduce the wealth of a single entity by altering the bl=
ockchain rules.<br>
<br>
In the case of the DAO fork, users were stealing back stolen funds, but tha=
t justification doesn&#39;t apply in this case. On the other hand, in this =
case we&#39;re talking about causing someone a loss by reducing the value o=
f hardware investments rather than forcibly taking back their coins, which =
is less direct and maybe more justifiable.<br>
<br>
While I don&#39;t like patented mining algorithms, I also don&#39;t like th=
e idea of playing Calvin Ball on the blockchain. Rule changes should not be=
 employed as a means of disempowering and empoverishing particular entities=
 without very good reason. Whether patenting a mining optimization qualifie=
s as good reason is questionable.<br></blockquote><div><br>Bitmain is block=
ing protocol upgrades to=20
preserve their mining advantage. This is quite distinct from someone=20
taking advantage of a visibly broken and highly toxic smart contract to=20
net themselves tens of millions of dollars. Further, Bitmain is=20
performing a patented hardware optimization. The patents mean that other
 miners are unable to capitalize on these optimizations. These=20
optimizations are to the tune of 30%. If you give one player in the=20
mining industry a permanent 30% cost advantage they will eventually own=20
everything. It&#39;s an industry where margins tend towards zero.<br><br>Th=
e
 asicboost patent is a direct threat to the health of the Bitcoin=20
ecosystem, and now we have visible proof. The war against segwit and the
 strife with Bitcoin Unlimited was very damaging to the ecosystem,=20
damaging to the price, and holding back significant improvements and=20
upgrades to the Bitcoin protocol. I interpret this as a direct attack on
 the Bitcoin ecosystem.<br><br>I don&#39;t know if changing the rules=20
to nullify asicboost is the right move. I&#39;m sure this won&#39;t be the =
last=20
patent that causes damage to the ecosystem. But you need to recognize=20
that the issue is not that Bitmain ran a hardware optimization. It&#39;s th=
at hardware optimizations exist which directly inhibit upgrading the protoc=
ol. And it&#39;s that hardware optimizations exist encumbered by patents en=
ough to give one party a decisive advantage in mining, decisive enough for =
them to build a single, centralized monopoly.<br><br></div><div>Each proble=
m is separate, and each problem is significant, and each problem is fundame=
ntal. The DAO attack was a one-time bout of stupidity that threatened a fix=
ed amount of money. asicboost is an ongoing status that directly damages Bi=
tcoin&#39;s ability to upgrade, and directly damage Bitcoin&#39;s ability t=
o retain any modicum of decentralization in the hashrate. The DAO issue did=
 neither of these things for ethereum.<br></div></div></div></div>

--001a11442ae0e72917054c7e4e22--