ProgPoW Audit Delay Issue

Ethereum per se is a network of computers, maintained by a self-selected team of core developers. Our remit is the health of the network. The community is not a polity, but a loose and self-selected anarchy of people with some relationship to that network. What are called “votes” are not about democracy; they are about ensuring that network upgrades are not so contentious as to cause an unwanted fork.

The responsibility of the core developers to the community is ill-defined at best. They don’t elect us and they don’t pay us, so by ordinary standards they have no standing to tell us what to do. So I continue to believe we core devs can best serve the community by seeing to the health of the network as best we can. We are best served by the community when arguments are presented to in those terms.

4 Likes

I don’t get how you can claim that those ranked as twitter influencers for Ethereum have no clue what they’re talking about.

I’ve heard some core developers dismiss twitter influencers with a similar attitude. My theory is that some people hear “twitter influencer” and start imagining Instagram models.

Let’s actually take a look at who some of these twitter influencers who are against ProgPow are:

Hayden Adams (@haydenzadams), creator of Uniswap.
Martin Koppelmann (@koeppelmann), founder of Gnosis
Ameen Soleimani (@ameensol), founder of SpankChain and MolochDAO
Georgios Konstantopoulos (@gakonst) Layer 2 / plasma researcher
Philip Daian (@phildaian) Researcher
Jorge Izquierdo (@izqui9) Aragon cofounder
Dan Robinson (@danrobinson) Researcher, creator of Rainbow Network
Mariano Conti (@nanexcool) Head of Smart Contracts at Maker DAO
Scott Lewis (@scott_lew_is)
James Prestwich (@_prestwich)

What justification is there for dismissing what these highly technical and capable people think about ProgPow?

In the Ethereum community people become influential on Twitter in part by being extremely good technically, founding companies, and building the Ethereum ecosystem.

If you claim the people I listed haven’t spent as much time as others delving into the details of the ProgPow implementation, that’s irrelevant to why it’s being opposed. Unless this has changed in the past few months, the core developers have made almost no effort (as reflected in dev calls that I reviewed) to actually determine if ASIC resistance is desirable.

Examples?

Consider a proposal to tax 100 million dollars of wealth from citizens of the US, destroy half of it, and give the other half to the McDonalds corporation.

You could say the same thing when someone questioned the importance of a poll of people who worked at McDonalds. “The most relevant vote here is the McDonalds vote, because they are the most affected by this proposal and have the most to lose by its failure to be implemented.”

It is irrelevant whether a particular group of current miners have a lot to lose if ProgPow isn’t implemented. There is no reason to favor incumbents just because they’re incumbents. What matters is the effect on security of the Ethereum network.

1 Like

This is absolutely the point.

1 Like

You told me that I would not be approved to purchase your hardware. You also accused me of being a paid shill (of who I’m not sure) and deemed me a scammer(?).

Clearly you’re not having much luck with your ethash chips. The fpga’s are already exploiting the known weaknesses in ethash. The clock is ticking on both Ethereum and Ethereum Classic forking away from ethash. Why not make a deal with AMD, Nvidia or Intel for some GPU’s and build a 1.2x more efficient plug and play miner with standardized and interchangeable parts (like Obelisk). With ProgPoW both GPU and ASIC can mine happily side-by-side and that would also satisfy both sides of the gpu vs asic centralization argument.

:v:

1 Like

I respect all of these people, especially in their relevant domains of expertise, but unfortunately on this issue many of these voices do not have the relevant exposure to the details to make a real decision. I should know, because I was one of them months ago.

I changed my mind by reading the facts and understanding the nuance of the discussion, and arrived at the result that I was wrong, and this proposal is helpful to the security of the network, which is the end goal we’re seeking here. I really do not mean to offend anyone in saying this, but reasoning about hardware and the protocol layer and PoW algorithms and all the incentives at play is a very difficult and nuanced topic, and not something you can just pick up, even if you are very smart. Especially hardware design, which we have preciously few people in our community with that expertise, and they’re all here trying to explain to us why this is so important.

Out of the list you gave, perhaps only 2 or 3 have spent any amount of depth on the topic, and only one has written a long-form article on the subject. There are no academic papers to reference on the subject, and none of the claims made have been independently verified or disproven, making all of this harder. So, I don’t think appeals to authority (aka influencer’s opinions) is a good move here.


Besides the fact that it’s written in the Ethereum Yellowpaper (and thus a core design goal of Ethereum via Ethash), “ASIC resistance” is definitely a misnomer for something else. What we’re trying to achieve is a low barrier to entry to mining. GPUs are great for this (despite what Sonia is telling us) because the supply chains are pretty open, and I can obtain hardware through retail channels anywhere in the world. There is also an element of robustness, since those in countries like Venezuela and China, where the authorities confiscate mining equipment by force (on occasion), you can at least try the “but it’s an AI training rig!” defense.

The economies of scale of R&D and hardware production is hopefully such that, economically, it won’t make sense to produce an ASIC. This is intended to be validated by the audit, although I’m not sure how accurate we can get with this analysis.


Here’s a tweet thread walking through the motivation behind ProgPoW by @TrustlessState. I had not seen him support ProgPoW before, and he did a good job summarizing it. Knowledge is power!


I hear this sentiment argued a lot, I feel like the majority of dissenters hold this view. Regardless, it’s not about supporting the incumbents, it’s about ensuring our current security providers, which is a more politically decentralized group of miners with the ability to reuse or sell their hardware at less of a loss (GPU miners can be repurposed to other coins and use cases, right?) stay in power, making an easier transition to PoS.

Private hardware manufacturers are more politically organized and can move faster than our community can, which, coupled with a strong incentive to continue profiting off of selling specialized hardware for Ethereum mining, will oppose the move to PoS tooth and nail. One just needs to look how hard they are fighting to blackball this proposal to get a sense of the tactics that will be used in the coming years to oppose the transition to PoS.


Obviously, I’m not going to convince anyone who is already convinced of their point of view, but I hope a few take the opportunity to deeply explore the evidence, learn a little bit about the economics of hardware manufacturing and PoW mining, and arrive independently at their own understanding of what is best in this scenario: allowing specialized mining hardware to take over our network’s security, or maintaining the “status quo” which has been working for a while without too much of a problem.

2 Likes

Do you believe this proposal is contentious enough to cause a community split if enacted?

No strong opinion there, but I think it may also be worth thinking about whether it disenfranchises some application developers enough to get them to build on another network. This is obviously hard to quantify, but seems worth thinking through.

How would this proposal disenfranchise developers? Do you mean those that have the opinion it’s not helpful?

I think there are several well-known projects that have vocally opposed ProgPow (i.e. Gnosis, Aragon, SpankChain). If ProgPow gets implemented, obviously these people may feel like Ethereum is moving in a direction that isn’t aligned with their vision of it, and may feel somewhat disenfranchised from the community.

Is there writings from the Gnosis or Aragon teams or well-known members articulating why they are against this proposal? Would be helpful to add for reference. I’ve seen a few Tweets just saying “I’m against ProgPoW” and maybe linking to Phil’s post, but that’s about it.

Here are some direct quotes from twitter (couldn’t find any from Aragaon):



I think you’ll find in this discussion thread and many others that we are all ultimately in agreement on these principles:

  1. Protect against existential threats
  2. Ensure the network is technically stable and adheres to the social contract for all stakeholders
  3. Ensure the future technical roadmap of Ethereum is as easy to implement as possible

However, we seem to disagree on the particulars, and that is where the contention comes from.

  1. A failure to act (letting market forces take over) is an existential threat because it eventually trades a known group of stakeholders (GPU miners) for an unknown one (ASIC manufacturers). We cannot know how this new group of stakeholders will participate in our ecosystem or how it will change the game theory of network operation, but based on what we’ve seen so far I am not optimistic that they will participate in good faith. In this scenario, holding the “status quo” would be actively taking steps to ensure our current set of stakeholders remains profitable, and the balance is maintained.
  2. Transitioning to new hardware will cause instabilities both technically and socially as we adjust to the newer paradigm. There is also the issue of breaking the current social contract with miners (who are stakeholders) as the Yellowpaper states that Ethereum prefers GPU mining, at least until the transition to PoS. Our current miners understand this, perhaps more than most.
  3. As I stated above, the transition to a new set of stakeholders may put our transition to PoS in jeopardy. Our current group of miners understands this risk, but the new group of ASIC manufacturers are private, for-profit companies with much different incentives that are aligned with selling hardware for our ecosystem, and not ensuring our ecosystem thrives long-term. Like it or not, mining is pretty close to a zero-sum game, with large cost overheads from hardware and power, so the only ways to make money is either through economies of scale or speculation on the underlying token. However, GPU economies of scale are very difficult, so token speculation is the primary way Ethereum miners make real profits, which means their interests become aligned with the network. I’m not saying they should get special treatment as investors in our ecosystem or anything, but I feel as they should have just as much of a say as any other investor has. Some of these miners may eventually participate as Validators under PoS, so it’s not like we’re selling the future to pay for the present here.

Yes, this issue is contentious, and some prominent voices disagree, mostly on principles that are difficult to validate in reality. However, Ethereum should not be afraid of contentious change, because if we are it does not bode well for the long-term scalability of our platform. We should find a way to work together and find compromise that improves things for the better. That is what makes us better than other chains.


My compromise: ensure ProgPoW is technically and economically viable, and can be enabled quickly on all major clients. Maybe development other algorithms. Use the threat of an algorithm change as a deterrent to further ASIC development, and find metrics everyone can agree on for when such activity comprises an “existential threat”. Then, act on that threat (as we have done before) to ensure the long-term stability and future of our platform.

2 Likes

Also, a note on Phase 0 block reward reduction. Check out this thread on the development of the finality gadget, which enables us to make this reduction:

If Phase 0 goes live, it will probably be early 2020 when it does. We will need at least 6 months, if not longer, to validate that the mechanism is working correctly, and develop the technical upgrades necessary to compensate for the issuance reduction and resolve all outstanding issues with issuance in general (rectifying the fact that we have two assets). This means that the earliest we can expect the issuance reduction is late 2020/early 2021. That means there is almost 2 years until such proposals can even be considered as arguments, which is plenty of time to ship new hardware. ProgPoW is still relevant.

1 Like

One just needs to look how hard they are fighting to blackball this proposal to get a sense of the tactics that will be used in the coming years to oppose the transition to PoS.

“They” is me? We are the smallest and most insignificant group of all in this game, that’s why we can speak freely.
We are 5 people in Shenzhen, feel free to visit us. The tactics of team ProgPoW are tough, I have sympathy for the significant amount of time you invest in this and that you believe them.
It’s a nice campaign and attack pulled off by ProgPoW, and for sure they will never stop. The moment you hand them a mic, the story is back on.

For the record here is what we wrote on January 12th

We are happy with the aging.
After discussions and news in recent months, it’s a given for us now that ProgPoW is a corporate money-making attempt, brought to you by one mega-miner with special deals with Nvidia on the chip side, and Calvin Ayre/nchain on the customer side. GPU miners not part of the collusion between Nvidia and Core Scientific will suffer.

We are not in 2015 when Ethash was designed and audited. Today Ethereum pays out 100 mio USD / month for decentralized PoW security.
The days of “low barrier of entry to mining” are over unless Ethereum dramatically changes the incentives and PoW algorithm (not to ProgPoW, to something that actually does what it says).
The minimum size to be viable is 10 MW today.

Use the threat of an algorithm change as a deterrent to further ASIC development

Our Ethash ASIC development will continue.
We are a chipmaker. We don’t design chips for or against Ethereum, or for or against Ethereum Classic, Zilliqa, Bitcoin Cash, etc. I will happily work on long-form answers to serious questions such as what drives our chip decisions, what is hard and easy in chip design, what is costly and what is cheap, what the customers want, what about prepay, what about risk wafers, etc.
We are also learning, it’s painfully obvious that our time estimate back from the ETC Summit was wrong!

It would be far more interesting to discuss how an Ethash (or ProgPoW) asic could lead to chips that accelerate ETH2. The investment carries over, so even if there is zero market for Ethash or ProgPoW, then the technology will be used to accelerate ETH2.

No. How many ASIC miners are there?

What difference could this choice make to developers? The hash algorithm isn’t part of the developer interface.

1 Like

Less than 23% of the hashrate in all likelihood (77% voted yes in ProgPoW miner vote, 23% abstained)

I’ve lost track - are there any EthHash ASICs actually shipping yet?

Shipping? No. Shipped? Yes. I believe there are 2 models of Ethash ASICs (which have been discontinued due to lack of demand), and Linzhi is working on a 3rd. That’s all the public information I think we have.

Much of the information needed to answer that question is private. Logically however, there would be no reason for ASIC hashrate to vote for this proposal, although the abstaining is more likely to be simply not wanting to participate in the vote, so that at least gives us an upper bound.

More than what is advertised as available to public, I would be concerned with what fpga’s are currently capable of and what that means to future ASIC’s.

Anyone with an efficient bitstream or ASIC is incentivized to keep it secret.