This issue has been raised for discussion by the DID WG based on feedback in Mozilla's Formal Objection to DID Core 1.0 (this is neither my or my organization's position -- I'm just the messenger):

https://lists.w3.org/Archives/Public/public-new-work/2021Sep/0000.html

Mozilla stated:

We must instead firmly oppose such proof-of-work technologies including to the best of our ability blocking them from being incorporated or enabled (even optionally) by any specifications we develop.

Per Mozilla's suggestion, should a future DID WG Charter that standardizes DID Methods contain language the specifically places proof-of-work DID Methods as "out of scope"?

@msporny wrote:

I would like to understand from @jandrieu @rxgrant and @csuwildcat if #17 (comment) is a valid, reasoned position.

In short, no. I do not. At length, I'll unpack why.

@bobwyman wrote

Whether or not existing PoW systems currently use electricity that would otherwise go unused, it is of some concern that in the absence of use by PoW systems, there would be greater market pressure to find other markets for that now unused electricity. The availability of cheap electricity is an indicator of a market failure that should be addressed. [emphasis mine]

@bobwyman starts with an arbitrary assertion that "the availability of cheap electricity is an indicator of a market failure that should be addressed." Yet, he provides no coherent argument for this. I recognize that he is involved in regulating these system and has unique expertise that I don't have. And yet, I find it an untenable position. Instead, I would argue that is a failure of regulators to incentivize enough local energy production in NYC and instead overproduced electricity in upper NY state. If energy is produced and used through a market mechanism, that is by definition a market success. When people can't buy or can't sell or can't find a mutually acceptable price, those are market failures. Cheap electricity doesn't meet that criteria. On the contrary, cheap electricity is in fact the holy grail for many sustainability efforts: getting renewable energy sources cheap enough to displace fossil fuels is the primary technical challenge of making the global energy system sustainable. As long as fossil fuels are cheaper than renewable options, those will continue to be exploited, with all of their attendance harms.

The rest of the comment veers into assertions that transmission and storage of electricity are more important than inexpensive production. Which is may be in New York. Many have different opinions on this when applied generally, because local energy production avoids storage and transmission costs, which are purely parasitic to the actual use. Thus, many argue that developing local production is, without reservation, the best option to fundamentally reduce the environmental impact of energy use. The real point here is that NY regulators have failed to internalize the externalities that Bob and others believe should be incorporated into the market price of electricity. I don't disagree. Yet, that is a regulatory failure, not a market one.

And it has nothing to do with whether or not PoW-using DID methods should be explicitly banned by the DID WG.

In a later comment, @bobwyman goes on to say

However, putting aside the question of the value of the end result, others would suggest that rather than simply comparing the cost of the means to the value of the ends, we should compare the costs of alternative means to the same ends.

Yes! I completely agree.

And Proof of Stake and other non-PoW approaches do not achieve the same ends. Full stop. They achieve approximate ends, and if your use case is focused on actual decentralization against bureaucratic overreach, there is no known solution that, in fact, represents an "alternative means to the same ends".

Since many of us in the community JOINED this community to build censorship-resistant identity solutions, PoW isn't just a nice idea, it is foundational to the entire DID gestalt. To propose that PoS is an equivalent is the kind of misunderstanding or ignorance that others have called out. And I mean ignorance in a functional way, not a pejorative one. Show me a system that achieves censorship-resistant global state that is more energy efficient than PoW, and I'll back you 100% to help bring that system to realization. To date, there are no know ways achieve what PoW does in a more efficient manner. And don't get me started with KERI...

However, the larger problem is that this argument has nothing to do with whether or not PoW systems are sustainable. Rather, the bulk of the comment is a political argument that storage and transmission are somehow more important problems than production. That's a legitimate political argument, but it has nothing to do with whether or not DIDs that rely on PoW systems are sustainable.

It has nothing to do with whether or not the DID Core spec should single out DID methods that utilize PoW networks for unique treatment.

So, for these reasons, @msporny I do not find @bobwyman's comment to be a valid reasoned opinion for why PoW-using DID Methods should be banned.

Second, on the merits of the argument of this issue itself. I'll restate the argument in hopes of clearly speaking to its individual components. And to clarify a mischaracterization of one of my points.

Sustainability matters
PoW systems are unsustainable
DID methods that rely on PoW systems are therefore unsustainable
DID methods that rely on PoW systems are so uniquely unsustainable that they should be explicitly banned or restricted in the WG Charter

I'll go through these, in turn.

Sustainability matters

I agree. I believe the only ethical Internet technologies are those that work when everyone on the planet chooses to use them. I wrote about this over a decade ago in a totally different context. https://blog.joeandrieu.com/2009/03/01/netizen-developer/

However, the current argument lacks a formal definition of sustainability. Even within the sustainability movement, there are competing definitions. Do we mean non-polluting? Do we mean carbon-neutral? Do we mean renewable (in the sense of 100% of inputs can be, and are, replaced/regrown/re-usable within a comparable timeframe of use)?

Without a rigorous definition, no arguments about "sustainability" can be considered scientifically meritorious.

Politically viable and socially activating? Yes. Scientifically meritorious? No.

PoW systems are unsustainable

This argument has great truthiness. It resonates easily because we hear the narrative about how PoW systems work—its decentralized nature meaning that anyone can use electricity and participant and be rewarded--and can immediately imagine the quantity of electricity used, as if that alone is a measure of unsustainability.

However, this is an illusion of perspective.

Although you can readily do some math based on current on-chain activity, it is fundamentally unknowable how much electricity is used and from where that electricity is sourced. That easy math contrasts with other systems in the world that are harder to measure, but are nevertheless accepted without question. Since the foundation of this argument is to establish "sustainability" as a differentiation for particular DID Methods, it is vital to understand sustainability in context, which begs a specific definition, which has yet to be provided.

In particular, the decentralized nature of permissionless PoW systems ensures that parties can participate or not, as their own interests align with the value of the network. Bitcoin is just as sustainable (in terms of sustainable continuity) as it is resilient to node operators joining or leaving. And it is extremely resilient to those changes, as we have seen with the recent ban on cryptocurrency mining in China. As long as there are parties willing to invest the resources in the network, the network is maintained. The activity, in fact, scales exactly with the value it creates. It scales dynamically and responsively in a way that, to me, defines a sustainable system.

More importantly, as certain sources of energy become unsustainable because of cost or regulation, the network will automatically shift to energy sources that are more sustainable. To those parties who want PoW systems to be "sustainable", the solution is easy: simply make unsustainable energy sources too expensive.

Again, I think this is a regulatory failure for electricity production at best and has nothing to do with the value or sustainability of PoW-using DID Methods.

In contrast, what I feel is unsustainable are systems that rely on massive military infrastructure and dollar-denominated oil sales to maintain a fiat currency controlled by a small group of people. The US Dollar is the world's reserve currency largely because international oil sales are priced in USD. Not only does this link the USD to a non-sustainable resource—making intermediations about that resource intimately entangled with economic consequences above and beyond just the use of that resource, it also forces the economic policy decisions of a tiny minority of the world population on the rest of the world. That is a recipe for exploitation and instability at a global scale. I don't find that sustainable. And I find it ridiculous that environmental advocates prefer the waste the USD system to the relatively clean use of electricity, networking, and cryptography to achieve a similar end.

Again, without a definition of "sustainable" we can't even talk about evaluations of sustainability, making assertions about any system as sustainable or not nearly meaningless.

@bobwyman misunderstood me when he summaries my argument thus:

@jandrieu argues that currencies where governments can increase the money supply cause unsustainable behavior. This roughly agrees with folks who oppose Bitcoin because fixed- and decreasing-supply monetary systems suppress economic activity. It certainly seems like periods of decreasing economic activity—recessions and depressions—cause more harm to average and vulnerable people than periods of expansion, but it's plausible that Joe could produce evidence of the opposite.

My point is that "sustainability" has many meanings and gave an example of different interpretations where bitcoin is in fact, the most sustainable monetary system ever invented.

You may not agree with me on the merits of my argument, but the fact that "sustainability" is underdefined for the question of this issue should be obvious to even the most cynical reader.

PoW-using DID Methods are unsustainable

This is like saying bicycles are unsustainable because they use roads built for cars and trucks.

Yes, there could be new PoW networks created solely for a novel DID Method, but none of the PoW-using DID Methods I know of do that, and I believe this objection is about those DID methods, not theoretical possible future methods. Such methods use an existing PoW network to anchor their Verifiable Data Registry. BTC is going to continue whether or not BTCR or did:ion exist. Ethereum is going to continue whether or not did:ethr and did:jolo exist. Is the marginal use of those DID Methods going to have significant impact on the electricity used by those systems? Not any time soon, and IMO, probably never.

Again, without a definition of "sustainable" we can't really discuss this except in the abstract such as whether or not the sins of the father should be laid upon the children. IMO, they should not. An opinion backed up by thousands of years of moral, religious, and legal precedent.

Therefore, ban PoW-using DID Methods

This is where the current proposal truly jumps the shark.

Even if we accept "sustainability" as a useful term, and if we accept some sense of how "unsustainable" PoW is, and even further, we accept some notion of how "unsustainable" PoW-using methods are, we still have not established that such PoW-using methods are any more or less sustainable than other methods. In short, even accepting these points, we still have not established that such methods are so unsustainable as to be called out for special treatment.

In particular, any argument that did:btcr or did:ion should be treated as unsustainable would require that did:web be considered unsustainable and presumably also out of scope for the DID WG. Estimates on both sides are problematic (as all estimates are), but the Web has been estimated to use as much as 3x as much electricity as bitcoin. However, IMO, neither did:web or did:ion will fundamentally change the energy footprint of the underlying substrate.

So why are PoW-using methods called out for banning while web-using methods are not? No argument is even presented here.

Instead, we have a simple one-sided assertion attacking PoW-using systems without making a distinction with those other methods that would be allowed.

This layer of the argument undermines itself, revealing a motivation to attack PoW systems rather than actually advance the cause of sustainability. Which is why this objection has been called a witch-hunt rather than a legitimate attempt to improve the sustainability of DID Methods.

I use that term for two reasons, one: I'm commenting on the term that was actually used and, two: the synonyms that come up such as inquisition, pogram, reign-of-terror, mcarthysm, ALL have their own cultural baggage. Personally, I think the functional term is a disinformation campaign to attack cryptocurrencies, but that doesn't capture the extremely pointed and unscientific nature of the attacks. While I will respectfully refrain from the term for the rest of this thread, as requested, it does capture the absolutely unscientific, hostile, targeted, and baseless nature of the objection.

Again, if we had a good definition of sustainability, we could refute or otherwise engage to resolve these gaps in the argument. Just as the witches that were actually hunted could have better refuted their accusers if "witch" had a legitimate definition.

Third, as to your assertion @msporny that

Continuing to insist, or seeming like you're insisting, that the other side of the debate is without merit is not a good strategy.

I encourage you to choose your own words, I'll choose mine.

I find the arguments about PoW are not scientifically meritorious. Others do.

Reasonable people can reasonably disagree. This is how science works. I went to great lengths to acknowledge that different people can reasonably have different opinions. That doesn't mean I have to accept unscientific arguments as scientific.

As an example, let's talk about another arbitrary topic which happens to have much greater weight of both scientific and political discourse: climate change. I happen to accept the science on the matter and recognize that human impact is the leading drive of catastrophic changes to our ecosystem. Others do not.

However, many of the arguments on the other side of the climate change debate are not scientifically meritorious. Some are good points, especially early in the exploration when the causal links were not as clearly understood, but as the decades moved on, the arguments against became more and more vitriolic and disconnected from scientific facts. There is no moral, ethical, or scientific argument that anyone has to accept arguments that fail to pass scientific muster.

There are moral, ethical, and scientific arguments that anyone should be able to be heard in their arguments. After all, if we don't allow the airing of such arguments, how could we possibly evaluate them?

But that doesn't mean those arguments are meritorious.

In fact, the reason we have an obligation to hear those arguments is so that we can make a determination about their merit.

In other words, the whole point of this conversation is to evaluate the merit of arguments for or against restricting DID Methods that rely on Proof-of-Work registries.

I don't find the case made so far to have any merit at all.

If you're still reading this…

First, thanks for wading through an insanely long discussion.

Second, if you still believe that PoW-using DID Methods should be banned because they are not "sustainable", please provide a rigorous definition of "sustainable" so we can all understand what you actually thing merits pariah status for PoW-using DID Methods.

For bonus points, include a comparative analysis of PoW-using systems in contrast to did:web as well as literally any other globally resolvable DID method that supports rotation with complete auditable provenance without reliance on any trusted third parties.

@jandrieu wrote:

@bobwyman misunderstood me when he summaries my argument thus: @jandrieu argues that currencies where governments can increase the money supply cause unsustainable behavior. [etc....]

That was written by @jyasskin, not me. You're responding to the wrong person.

If you're still reading this… First, thanks for wading through an insanely long discussion.

I read the whole thing through... twice. Thank you for taking the time to write it, I really appreciate it.

It's discourse like that, and that put forward by @bobwyman, that will help us advance this conversation.

I particularly found the way the argument to ban PoW-based DID Methods was deconstructed and walked through helpful.

I hope that @jyasskin, @tantek, and others will be able to see the depth at which this question has been analyzed over the years... none of this is new, but @jandrieu's summary of the thought process will hopefully be helpful for those that haven't been steeped in this discussion for years. We should lift that out into a separate document so it doesn't get drowned in what is shaping up to be a very long comment thread.

@jandrieu wrote:

@bobwyman starts with an arbitrary assertion that "the availability of cheap electricity is an indicator of a market failure that should be addressed." Yet, he provides no coherent argument for this.

Market Failure is defined in Wikipedia, and similarly elsewhere, as:

In neoclassical economics, market failure is a situation in which the allocation of goods and services by a free market is not Pareto efficient, often leading to a net loss of economic value. Market failures can be viewed as scenarios where individuals' pursuit of pure self-interest leads to results that are not efficient – that can be improved upon from the societal point of view.

This is a much more constrained definition than that which @jandrieu appears to use. It is, however, the definition that I had in mind when I made my comment.

In New York, NYISO (New York Independent System Operator) operates our state-wide electricity market with a charter to address the needs of the state as a whole. Ideally, the price of the electricity commodity would be largely the same everywhere in the state, with any differences in price reflecting only differences due to transmission line loss -- one should expect that costs will increase in proportion to the distance that the point of use is from the point of generation. (Note: I address here only the commodity cost of electricity, not the "delivery" cost which is driven by local utilities' need to recovery the costs of their investments in physical distribution assets. (i.e. depreciation of the "ratebase") )

However, the price of electricity varies greatly within New York for a variety of reasons. One of several reasons is that "congestion" often arises when transmitting electricity between various parts of the state. Even if some areas of the state have "excess" generation capacity, which could satisfy demand in some other part of the state, they can't actually export that electricity unless there is sufficient congestion-free transmission capacity.

The price of electricity in congested areas tends to fall, even if it remains high elsewhere. Additionally, falling prices will tend to cause curtailment of generation within congested areas and thus we might simultaneously curtail relatively clean generation in a congested area while increasing generation from dirtier sources in some other areas. In that case, not only will areas blocked by congestion have higher than ideally efficient commodity costs, residents of all areas will suffer higher externality costs due to the increased average pollutant and GHG emission rates. (An excellent example of this occurs regularly in some parts of Texas where, due to limited transmission capacity serving wind farms, off-peak electricity is actually "free" even while coal plants are used to generate electricity for other parts of the state...)

Clearly, congestion can be avoided by increasing the inter-region transmission capacity and thus the ability to import or export electricity. Also, storage could be used to allow nonexportable electricity to be buffered within a region or to allow maximization of transmission capacity utilization. Historically, when congestion remains a problem long enough or grows great enough, regulators, transmission operators, etc. have regularly proven that they will work together to increase transmission capacity. Recently, the same actors have begun to study and deploy storage systems.

Although I've glossed over a few market failures, another market failure occurs whenever a rational actor, such as a Bitcoin miner, notices that electricity costs are lower in a congested area and, in pursuit of rational self-interest, decides to relocate to that area only in order to exploit the low prices. The result of the miner's new demand within the congested area will probably be a reduction in the likelihood that either storage or transmission upgrades will be pursued. Congestion will have been "relieved" simply by consuming the otherwise excess electricity in the local market. Thus, generation capacity which could otherwise have served the interests of the entire state, and would have enabled not only a reduction of statewide variance in electricity prices but also average emission rates, would be effectively captured for the largely exclusive benefit of a single customer.

We should also consider that Bitcoin miners have regularly demonstrated their ability to start-up, tear-down, and relocate mining facilities very rapidly and with little notice. We know that the economics of mining are constantly changing and there is always some new area that offers lower electricity prices. We know that mining becomes progressively harder and that, eventually, the maximum number of coins will have been mined. This means that, unlike most other large sources of new electricity demand, which once installed tend to remain in place for a very long time, a Bitcoin miner's commitment to any specific area is unusually tentative even though that miner's consumption of local nonexportable electricity may have distorted or caused the cancellation of long-term grid planning processes for transmission, storage, etc. Regulators and grid-planners are left in a quandary. Should they assume that a miner is merely a transitory load and thus potentially waste resources by planning as though they weren't there, or should they assume that they will remain in place for the long-term?

Hopefully, this illuminates what I meant by "market failure" as well as some of things that make it hard to accept without some skepticism even earnestly asserted claims that Bitcoin mining is an unalloyed benefit to the electric grid. As is the case with many such issues, the reality is a bit more complex than advocates proclaim.

Note: Nothing I've said above, or in any previous message, indicates my position on the registration of DID methods based on proof-of-work systems. As one who is not a member of W3C, I have no vote. Nonetheless, my personal opinion is that any DID method which satisfies the rubric, and is likely to be used, should be registered. On the other hand, if it is necessary or appropriate for the W3C to state a preference for any DID methods (even if only as examples in explanatory documentation), I believe that they should prefer methods which can be implemented simply and without relying on the complexities and controversies presented by systems which require PoW. You may note that as the proposer of the did:tag method, my personal preferences are expressed in its definition.

Nonetheless, my personal opinion is that any DID method which satisfies the rubric, and is likely to be used, should be registered. On the other hand, if it is necessary or appropriate for the W3C to state a preference for any DID methods (even if only as examples in explanatory documentation), I believe that they should prefer methods which can be implemented simply and without relying on the complexities and controversies presented by systems which require PoW. You may note that as the proposer of the did:tag method, my personal preferences are expressed in its definition.

I would from mesur.io's perspective offer an adjustment to this comment and and propose these potential principles for the WG:

The DID WG should focus on interop and data model concerns, along with notes on DID method evaluation, registration, and implementation, not on standardizing specific DID methods
DID methods that are broadly essential for interop testing, adoption within existing web infrastructure, and that are not tied to specific commercial interests, such as did:web and did:key might be worth considering for standards definition at the WG level
ANY DID method that is an implementation of DID Core should be allowed to be added to the registry - we should not as a WG be in the business of picking winners
The Rubric should provide the ability to evaluate any DID method by a series of criteria, including ethical and environmental criteria, in addition to the more obvious concerns such as level of decentralized control, privacy, and security.
The Rubic criteria, should as much as possible, prefer criteria in a given area of concern that is authored and informed by subject matter experts in that area of concern

These (and possibly other) base principles would allow us to evaluate and assess proposals such as stated in the OP.

@bobwyman

I have no vote

Bob, as someone who has taken the effort to be informed and to engage, your thoughts on the issues matter and will be read by everyone here as well as the most dilligent members of the W3C who are voting. Every good point of discussion matters and affects our shared understanding.

[...] propose these potential principles for the WG:

I mostly agree.

The Rubic criteria, should as much as possible, prefer criteria in a given area of concern that is authored and informed by subject matter experts in that area of concern

I think every idea has to fend for itself, and enshrining topical authority is a vehicle for centralization. I'd propose that each criteria in the rubric gets its own "permanent record" of discussion as a separate github issue, where subject matter experts can offer the best references to support the best arguments.

I think every idea has to fend for itself, and enshrining topical authority is a vehicle for centralization. I'd propose that each criteria in the rubric gets its own "permanent record" of discussion as a separate github issue, where subject matter experts can offer the best references to support the best arguments.

Excellent idea! think that is a great way of balancing and tracking history on the "informed consensus" so to speak

Hopefully, this illuminates what I meant by "market failure" as well as some of things that make it hard to accept without some skepticism even earnestly asserted claims that Bitcoin mining is an unalloyed benefit to the electric grid. As is the case with many such issues, the reality is a bit more complex than advocates proclaim.

Excellent. I believe we are in complete agreement here. I don't believe anyone suggested that "bitcoin mining is an unalloyed benefit". That would be an untenable assertion. My point is the same as yours: reality is more complex than advocates of this issue have acknowledged. You can't simply estimate a ballpark of the total electicity usage of bitcoin and blindly claiming that PoW is unsustainable. Well, you can, it just wouldn't be scientifically valid.

Nonetheless, my personal opinion is that any DID method which satisfies the rubric, and is likely to be used, should be registered.

Please note that "satisfying the rubric" is not a thing.

Even the crappiest method can be evaluated. Evaluations don't result in a canonical yes or no, valid or invalid, satisfied or unsatisfied.

People use the rubric to select the criteria they care about to evaluate the methods they care about. The result is a bespoke evaluation from a subset of criteria representing one evaluators judgment for those criteria. When aggregated across methods, it provides a way to compare DID Methods on an apples-to-apples basis using the criteria that matter most to the party doing that aggregation.

It is fundamentally a subjective process that represents the judgment of a particular person or team at a particular point in time for particular use cases. It does NOT result in an objective anything. Any attempt to do so would undermine the very decentrality we are here supporting. What matters to one person may or may not matter to another. What matters in one use case may not matter in another. The Rubric does not stand in as an arbiter of what makes a "good" method.

@jandrieu writes:

"Proof of Stake and other non-PoW approaches do not achieve the same ends."
"there is no known solution [to actual decentralization against bureaucratic overreach] that, in fact, represents an 'alternative means to the same ends'."
"PoW ... is foundational to the entire DID gestalt."
"To propose that PoS is an equivalent [to PoW] is the kind of misunderstanding or ignorance that others have called out."

While repetition is a useful indicator of the author's strength of conviction, such repetition does little to illuminate the reasons why PoS, or any other alternative to PoW, would not "achieve the same ends" as PoW.

The fact that alternatives to PoW have been and are being proposed indicates that a belief in the absolute superiority of PoW is somewhat controversial. It also appears that some systems that began using PoW, such as Ethereum, are now in the process of converting to PoS.

Given that controversy exists, I would appreciate it very much if someone could identify at least the essential attributes of PoS, or other alternatives, that cause it to appear so deficient to the advocates of PoW.

bob wyman

I would appreciate it very much if someone could identify at least the essential attributes of PoS, or other alternatives, that cause it to appear so deficient to the advocates of PoW.

The very short answer is that it costs energy to build the castle to protect staked coins from all possible attacks, and if the cost of the attack is less than the payoff, it's rational to put further resources into protecting the stake. The original declaration of this principle was outlined in Nothing is Cheaper than Proof of Work, an essay from 2015 by Paul Sztorc.

Similar security problems affect any organization with large holdings, even under the proof-of-work cryptocurrencies, but any successful attack there would not be a systemic risk to the rest of the blockchain's operation (although hashrate does drop for a while when the market reacts to disasters).

Fungibility in effort was empirically observable during the launch of Chia earlier this year, an "eco-friendly" proof-of-space cryptocurrency, which ended up consuming hard drive stock and manufacturing capacity, causing a hard drive shortage.

P.S. Off-topic but of interest to the thread in general: Sztorc wrote an essay earlier this month on why hashrate (and thus energy usage) will fall as the subsidy halves again and again, if no other mechanism to pay for security is developed.

[...] essential attributes of PoS, or other alternatives, that cause it to appear so deficient to the advocates of PoW.

Oh, the other reason is that Bitcoin's proof-of-work distribution function was the only fair way to do it that was available on day one, and they're sticking to it.

This fairness issue is more relevant to money than reliable DID updates. However, people want to express their values on these matters, so they don't want to ignore distribution fairness any more than they want to ignore effects on the environment.

If you, the founder, are starting a proof-of-stake blockchain, you have to ask how you're distributing the coins...

Do you just sell them? But you must keep some for your development team, which you control, or no one will believe the coins will maintain any value... To this day Ethereum issued more coins in its ICO than have ever been mined, if I recall correctly. Cardano laundered its ICO through Japan.

Did a cryptocurrency start with proof-of-work but now professes to be moving off? The miners will work to delay that (as they have for Ethereum, going on seven years now).

Do you skip coins and pick some companies to be functionaries? Several VDRs in this space were created by companies that have partnered up with similarly interested companies in order to be decentralized. This federation model is very close to proof-of-stake (validate updates with signatures, sometimes in exchange for an operator reward), and I have no problem with it. But none of them have achieved more than a few dozen node operators enforcing their DID Method rules, which means that they're subject to only a few local jurisdictions (and sometimes that's stable enough, but it's a different threat model). So if a DID Method design team wants their method to be completely decentralized, then the best available design space is highly incentivized global blockchains.

Even if our hypothetical founder does, somehow, initially distribute proof-of-stake coins fairly, staking has structural inequality. The more coins one has, the more staking rewards one gets. Capital helps miners in proof-of-work, too, but if you look at the volatile history of top miners, and especially the steep price that BitMain paid for its mistakes in the blocksize war, you can see that there's a lot more to it.

Proof-of-work in Bitcoin levels the playing field. Even leveling the playing field doesn't directly address prior unjust inequality, but it does reward curiosity, innovation, and hard work over long periods of time.

P.S. The original paper on why proof-of-stake doesn't work is On Stake and Consensus, by Andrew Polestra, last revised in 2015. The main point is that there is no external source of truth corroborating the checkpoints that you first see. The argument is hard to apply to recent proof-of-stake proposals because they have become so complicated, so I started out with shorter answers.

Glad to see the issue is progressing.

Even the crappiest method can be evaluated [ / registered ].

I think this is the key topic for us to discuss regarding registries at W3C.

Specifically, there are registries with values baked in, and ones which strive not to have values baked in.

On a side note, calling methods "crappy or dog****" is a way of trying to inject values into a registry, and I think its a terrible idea.... there are lots of registries with various quality and quantity of entry.... and complaining about the quality of items in a list eventually extends to complaining about the list itself. Have a look at this registry:

https://www.iana.org/assignments/cbor-tags/cbor-tags.xhtml

Are we going to say there are a lot of "crappy or dog****" cbor tags? we should discourage the use of such language, I don't think the folks using it intend to cause harm, but there is real potential for it if we continue to use this kind of tone.

The rubric has at least some values baked in.

One of those values is that "any method can be evaluated"...

The other part of @jandrieu's answer is also correct imo:

Evaluations don't result in a canonical yes or no, valid or invalid, satisfied or unsatisfied.

I think folks often seek this kind of thing without realizing how hard it is to get in a fair way for everyone, and how likely we are to end up in a situation that is unfair and filled with bias....

Let's consider DID Core v1 Conformance a rubric criteria, and ask:

Should did methods that have a privacy and security considerations section that is almost empty evaluated?

IMO, the answer is yes... however, the confidence of everyone in any evaluation for such a method is low... after all, there was not a lot to evaluate.

Now let's imagine 5 years from now, W3C has a new mandatory section called "Environmental, Ethical and Social" considerations.... DID Core could require method specs to have such a section.

Some methods might put a lot of time and effort into that section.... they might invent fantastic atomic swap / proof of stake / carbon credit awesomeness.

Others might say electricity is required by most processing systems, and electricity should be derived from renewable sources whenever possible.

When performing a rubric evaluation, these sections of the method spec can be relied on.... but I don't think the rubric, or the did core v1 spec is the right place to tell a did method author what they are or are not allowed to use to build a did method, nor is either the right place to tell them if they "considered the environment enough"... its not possible to achieve that.... we can always do better.

In summary, DID WG members (and W3C generally) should avoid taking the position that we can alter culture through normative statements in specs.

We can say things like: address these 5 questions... but the second we start grading answers from different political perspectives as "passing or failing" we are in deep trouble... privacy, sustainability, security.... all of these are political issues today.... we don't want to encourage folks to play games with addressing them, by punishing or privileging attempts to speak to values that overlap with our own... we want to encourage authors to address them to the best of their ability...

If a method author thinks carbon credits and proof of stake are a better solution than proof of work and wind energy, let them spend their time arguing for that in their method... The current normative requirements for did methods do not require them to do that, but they remain free to exceed those requirements if they are passionate enough.

The rubric is meant to give folks questions and tools to evaluate did methods... we are not here to make the bar easier or harder to meet for specific technologies, or bias the bars towards certain evaluation outcomes.

The did spec registries is meant to be a list of pointers to specs, not a place to pick winners, or pseudo-evaluate did methods.

The did core v1 spec has normative requirements for a conforming did method spec, those requirements don't enable us to evaluate if a required section was "bad", "ok" or "excellent"... and probably experts won't agree on some sections (such as privacy or security benefits of public ledger systems)...

@jandrieu wrote:

Please note that "satisfying the rubric" is not a thing.

Please forgive my imprecise writing. By "satisfy the rubric" I meant only that the rubric had been responded to; that the DID method authors had made an apparently good faith effort to consider and respond to the rubric. I do not suggest that any metric for a "correct" or even "satisfactory" response should be imposed.

I believe that the registry should be nothing more than a mere public listing that identifies, and provides access to descriptions of, DID methods associated with DIDs that my code may come across. The mere fact that a DID method is registered should not imply that it has been standardized, endorsed, or even evaluated by the W3C or any of its working groups, community groups, or whatever. (i.e. If a Bitcoin-specific DID method is "standardized," it should be standardized by whatever community defines Bitcoin standards, but, whatever the Bitcoin folk come up with, and no matter how much others may object to or support what they do, if others are likely to be exposed to their DIDs, then their DID method should be registered.)

Some DID methods may be deficient in meeting requirements, such as those for security or privacy, that are widely accepted by the community. Some DID methods might rely on truly bizarre or onerous technical requirements. But none of that should influence what is or is not registered. If a DID method is in use, and I discover a DID which uses a DID method with which my code is unfamiliar, I should have a means to discover the intended methods associated with that DID. Of course, on inspecting the DID method's definition and rubric answers, I might decide that I have no intention of supporting such DIDs. On the other hand, I might decide to invest in whatever development is needed to support them. However, without a registry, I have no opportunity to make these decisions. I should have the freedom to choose.

While I don't think that the W3C should be involved in the "standardization" of DID methods, I do think it would make sense for the W3C to non-exclusively highlight one or a very small number of DID methods that do, in fact, satisfy some broadly accepted set of requirements and that can be implemented with a minimum number of technical dependencies. Thus, I would support the W3C's non-exclusive highlighting of DID methods such as the now registered did:web method, or the did:tag method that I have proposed, but not registered. The non-exclusive endorsement of such easily implemented and broadly useful DID methods should be intended only to highlight suggested "defaults" for those who need DIDs but who don't have any requirements that would cause them to need to rely on some other DID method or to go to the trouble of defining their own unique DID methods.

bob wyman

Beginning standardization of DID Methods was added to the charter in #20 Which methods to choose, if any, will be determined by the WG.

w3c / did-wg-charter

Do not standardize DID Methods that are not environmentally sustainable #17

Sustainability matters

PoW systems are unsustainable

PoW-using DID Methods are unsustainable

Therefore, ban PoW-using DID Methods