- From: Will Abramson <will@legreq.com>
- Date: Fri, 18 Jul 2025 09:25:52 +0100
- To: Melvin Carvalho <melvincarvalho@gmail.com>
- Cc: Filip Kolarik <filip26@gmail.com>, Steve Capell <steve.capell@gmail.com>, Benjamin Young <byoung@digitalbazaar.com>, Adrian Gropper <agropper@healthurl.com>, Credentials Community Group <public-credentials@w3.org>
- Message-ID: <CAPJWd2SpQZeChE4Won2Dw48aJghTuy_R2xhiQfsj_wnEDrArvA@mail.gmail.com>
Apologies, I have not fully digested this thread. But I just wanted to say I find "pubkey as name" to be a pretty suboptimal solution. People lose of want to change their keys regularly. I mean who all here has lost there house keys right. Thing is I lose my house keys, I don't lose my house. Same here, I should be able to change or lose my keys without losing my name. That is a big part of what DIDs are all about. Thanks, Will On Fri, Jul 18, 2025, 07:22 Melvin Carvalho <melvincarvalho@gmail.com> wrote: > > > čt 17. 7. 2025 v 23:35 odesílatel Filip Kolarik <filip26@gmail.com> > napsal: > >> On Thu, Jul 17, 2025 at 11:23 PM Steve Capell <steve.capell@gmail.com> >> wrote: >> >>> I don’t see how dns is centralised. It’s a massively distributed lookup >>> system technically. In a governance sense it empowers any beating heart to >>> pick a domain name that isn’t already taken >>> >> >> Technically, DNS is distributed, but governance is centralized. TLDs are >> controlled by a small number of registries under government jurisdiction. >> Recent domain bans and seizures (e.g. in Russia, Turkey, and India) show >> how easily access can be revoked at the top. So yes, you can pick a name, >> but you're still playing in someone else’s namespace. >> > > There are open alternatives to DNS. > > The simplest way is to have a keypair. The user holds a private key, and > then the public key becomes their "name" on the internet. > > Short names are another class of problems, and for that the key property > is that it's a non-proprietary level playing field. > > There are some believers in the "zooko triangle" unproven thesis that > short names are impossible, but in fact, all you need is a fair tie-breaker > for two people that want the same short name. What did Zooko get wrong? > He forgot about time, let people choose a name in time fairly, and then > when two people choose the same name, the tie-breaker is which was earlier. > > I will implement these 2 strategies (pubkey as name, shortnames with > tiebreaker) in addition to DNS, which I think gives users the choice and > the best of all worlds. > > >> >> Best, >> Filip, https://github.com/filip26 >> >> >> >>> >>> I must be missing something. >>> >>> On the other hand I’m deeply suspicious of anything that even smells >>> like a blockchain. Private ledgers are tech vendor snake oil. Public >>> ledgers are money laundering Ponzi schemes. Can’t see how they are >>> anything but that. >>> >>> Steven Capell >>> Mob: 0410 437854 >>> >>> On 17 Jul 2025, at 11:12 pm, Benjamin Young <byoung@digitalbazaar.com> >>> wrote: >>> >>> >>> On Thu, Jul 17, 2025, 5:00 PM Steve Capell <steve.capell@gmail.com> >>> wrote: >>> >>>> Anytime I hear anyone say anything like “Bitcoin is a good thing” it >>>> makes me shudder and want to vomit. As far as I can tell It’s a monstrous >>>> Ponzi scheme that is good for money laundering and not much else >>>> >>>> Why do we perceive did:web (or its improved variants like did:webvh) as >>>> “centralised”? What could be more decentralised than the web? Certainly not >>>> any distributed ledger >>>> >>> >>> DNS (as deployed) is the centralizing component of what most people call >>> "the Web". An HTML-based ecosystem that (de)references things with >>> universal identifiers (URIs) and locators (URLs) doesn't necessarily have >>> that same constraint. >>> >>> In so far as did:web and did:webvh also have a strong dependence on >>> DNS...they would sadly be centralized. >>> >>> However, if the are protocol (beyond HTTP) and/or naming (beyond DNS) >>> agnostic, then they would still have some level of decentralization. >>> >>> But...like the Web...their dominant "expression" would likely be >>> centralized (or at least entangled with a centralized system). >>> >>> (Obviously ignoring mDNS, /etc/hosts, and other means of local naming or >>> DNS overriding) >>> >>> That's my understanding, anyway. >>> >>> Cheers, >>> Benjamin >>> >>>> >>>> >>>> Steven Capell >>>> Mob: 0410 437854 >>>> >>>> On 17 Jul 2025, at 10:41 pm, Melvin Carvalho <melvincarvalho@gmail.com> >>>> wrote: >>>> >>>> >>>> >>>> >>>> čt 17. 7. 2025 v 22:24 odesílatel Adrian Gropper < >>>> agropper@healthurl.com> napsal: >>>> >>>>> Nostr might be a good start for de-platforming social media on the >>>>> basis of pseudonymity and relay-based discovery, but unless >>>>> the architecture also supports untraceable payment the major surveillance >>>>> platforms will persist. >>>>> >>>> >>>> Nostr is tied to any payment system. But it is largely built by people >>>> in the bitcoin community, so there have been some integrations with bitcoin >>>> technologies, such as the lightning network. >>>> >>>> Innovation continues in this area. I think that integration with >>>> Blockstream's Liquid [1] would be a good start. >>>> >>>> [1] https://blockstream.com/liquid/ >>>> >>>> >>>>> >>>>> Adrian >>>>> >>>>> On Thu, Jul 17, 2025 at 3:58 PM Melvin Carvalho < >>>>> melvincarvalho@gmail.com> wrote: >>>>> >>>>>> >>>>>> >>>>>> čt 17. 7. 2025 v 21:38 odesílatel Adrian Gropper < >>>>>> agropper@healthurl.com> napsal: >>>>>> >>>>>>> It's clearly time for a new architecture. One that benefits from our >>>>>>> experience with SSI as an anti-pattern that is too easily inverted or >>>>>>> ignored. >>>>>>> >>>>>>> I would suggest an architecture that sees platforms for payment and >>>>>>> social media as the problem instead of focusing on identity. An >>>>>>> architecture that, like cash and geocaches, defaults to anonymity by design. >>>>>>> >>>>>>> I would also suggest an architecture that ignores licensed >>>>>>> professionals and things. With the benefit of hindsight, the premise that >>>>>>> identity standards must span licensing and supply chains seems inane. >>>>>>> >>>>>> >>>>>> We have a fairly advanced ecosystem working on all these problems >>>>>> over at Nostr, with several million users, and several thousand DAU. >>>>>> >>>>>> We also have a W3C Nostr Community Group [1] and have already begun >>>>>> work on a did:nostr spec. >>>>>> >>>>>> [1] https://www.w3.org/community/nostr/ >>>>>> >>>>>> >>>>>>> >>>>>>> Sorry, >>>>>>> - Adrian >>>>>>> >>>>>>> >>>>>>> On Wed, Jul 16, 2025 at 3:59 AM Christopher Allen < >>>>>>> ChristopherA@lifewithalacrity.com> wrote: >>>>>>> >>>>>>>> I have occasionally posted a link to one of my blog articles to >>>>>>>> this group, but I thought this article deserved a broader discussion by our >>>>>>>> CCG community, so I'm sharing here. >>>>>>>> >>>>>>>> The original article is at >>>>>>>> https://www.blockchaincommons.com/musings/gdc25/ >>>>>>>> >>>>>>>> -- Christopher Allen >>>>>>>> >>>>>>>> Musings of a Trust Architect: When Technical Standards Meet >>>>>>>> Geopolitical Reality >>>>>>>> Digital Identity, Sovereignty, and the Erosion of Foundational >>>>>>>> Principles >>>>>>>> By Christopher Allen <ChristopherA@LifeWithAlacrity.com> >>>>>>>> 2025-07-15 >>>>>>>> >>>>>>>> *Reflections on recent conversations about digital identity, >>>>>>>> sovereignty, and the erosion of foundational principles* >>>>>>>> >>>>>>>> Echoes from Geneva >>>>>>>> >>>>>>>> I wasn't present at the [Global Digital Collaboration]( >>>>>>>> https://globaldigitalcollaboration.org/) conference (GDC25), but >>>>>>>> the observations shared by colleagues who attended have crystallized some >>>>>>>> issues I've been wrestling with for years. I should note there's a >>>>>>>> selection bias here: I'm the author of the [10 principles of self-sovereign >>>>>>>> identity]( >>>>>>>> https://github.com/WebOfTrustInfo/self-sovereign-identity/blob/master/self-sovereign-identity-principles.md), >>>>>>>> so my community tends to have strong opinions about digital identity. >>>>>>>> Still, when multiple trusted voices independently report similar concerns, >>>>>>>> patterns emerge that are worth examining. And these weren't casual >>>>>>>> observers sharing these concerns. They were seasoned practitioners who've >>>>>>>> spent decades building identity infrastructure. Their collective unease >>>>>>>> speaks to something deeper than technical disagreements. >>>>>>>> >>>>>>>> It's hard to boil the problems at GDC25 down to a single issue, >>>>>>>> because they were so encompassing. For example, there was a pattern of >>>>>>>> scheduling issues that undercut the community co-organizing goal of the >>>>>>>> conference and seemed to particularly impact decentralized talks. One >>>>>>>> session ended up in a small, hot room on the top floor that was hard to >>>>>>>> find. (It was packed anyway!) Generally, the decentralized-centric talks >>>>>>>> were in bad locations, they were short, they had restricted topics, or they >>>>>>>> were shared with other panelists. >>>>>>>> >>>>>>>> I think that logistical shuffling of events may point out one of >>>>>>>> the biggest issues: decentralized systems weren't given much respect. This >>>>>>>> may be true generally. There may be lip service to decentralized systems, >>>>>>>> but not deeper commitments. Its value isn't appreciated, so we're losing >>>>>>>> its principles. Worse, I see the intent of decentralization being inverted: >>>>>>>> where our goal is to give individuals independence and power by reducing >>>>>>>> the control of centralized entities, we're often doing the opposite — >>>>>>>> still in the name of decentralization. >>>>>>>> >>>>>>>> The Echo Chamber Paradox >>>>>>>> >>>>>>>> The problems at GDC25 remind me of Rebooting the Web of Trust >>>>>>>> (RWOT) community discussions I've been following, which reiterate that this >>>>>>>> is a larger issue. We debate the finer points of zero-knowledge proofs and >>>>>>>> DID conformance while missing the forest for the trees. Case in point: the >>>>>>>> recent emergence of "[`did:genuineid`]( >>>>>>>> https://genuinein.com/DIDMethod)" — a centralized identifier >>>>>>>> system that fundamentally contradicts the "D" in DID. >>>>>>>> >>>>>>>> Obviously, decentralization is a threat to those who currently hold >>>>>>>> power (whether they be governments, corporations, billionaires, or others >>>>>>>> who hold any sort of power), because it tries to remove their >>>>>>>> centralization (and therefore their power), to instead empower the >>>>>>>> individual. But if we can't even maintain the semantic integrity of >>>>>>>> "decentralized" within our own technical community, devoted to the ideal, >>>>>>>> how can we fight for it in the larger world? >>>>>>>> >>>>>>>> The Corpocratic Complication >>>>>>>> >>>>>>>> GDC25 was held in Geneva, Switzerland. 30+ standards organizations >>>>>>>> convened to discuss the future of digital identity. Participants spanned >>>>>>>> the world from the United States to China. There was the opportunity that >>>>>>>> GDC25 was going to be a truly international conference. Indeed, Swiss >>>>>>>> presenters were there, and they spoke of privacy, democratic involvement, >>>>>>>> and achieving public buy-in. It was exactly the themes that we as >>>>>>>> decentralized technologists wanted to hear. >>>>>>>> >>>>>>>> But from what I've heard, things quickly degraded from that ideal. >>>>>>>> Take the United States. The sole representative of the country as a whole >>>>>>>> attended via teleconference. (He was the only presenter who did so!) His >>>>>>>> talk was all about Real ID, framed as a response to 9/11 and rooted in the >>>>>>>> Patriot Act. It lay somewhere between security-theatre and >>>>>>>> identity-as-surveillance, and that's definitely not what we wanted to hear. >>>>>>>> (The contrast between the US and Swiss presentations was apparently >>>>>>>> jarring.) >>>>>>>> >>>>>>>> And with that representative only attending remotely, the United >>>>>>>> State's real representatives ended up being Google and Apple, each >>>>>>>> advancing their own corpocratic interests, not the interests of the people >>>>>>>> we try to empower with decentralized identities. >>>>>>>> >>>>>>>> This isn't just an American problem. It's a symptom of a deeper >>>>>>>> issue happening across our digital infrastructure. It's likely the heart of >>>>>>>> the inversions of decentralized goals that we're seeing — and likely >>>>>>>> why those logistical reshufflings occurred: to please the gold sponsors. In >>>>>>>> fact, the conference sponsors tell the story: Google, Visa, Mastercard, and >>>>>>>> Huawei were positioned as "leading organizations supporting the advancement >>>>>>>> of wallets, credentials and trusted infrastructure in a manner of global >>>>>>>> collaboration." >>>>>>>> >>>>>>>> While Huawei's presence demonstrates international diversity — a >>>>>>>> Swiss conference bringing together Europe and Asia — it also raised >>>>>>>> questions about whose vision of "trust" would ultimately prevail. When >>>>>>>> payment platforms and surveillance-capable tech giants frame the future of >>>>>>>> identity infrastructure, we shouldn't be surprised when the architecture >>>>>>>> serves their interests first. >>>>>>>> >>>>>>>> This echoes my concerns from ["Has SSI Become Morally Bankrupt?"]( >>>>>>>> https://www.blockchaincommons.com/musings/musings-ssi-bankruptcy/). >>>>>>>> We've allowed the narrative of self-sovereignty to be co-opted by the very >>>>>>>> platforms it was meant to challenge. The technical standards exist, but >>>>>>>> they're being implemented in ways that invert their original purpose. Even >>>>>>>> [UNECE sessions acknowledged]( >>>>>>>> https://unece.org/trade/events/global-digital-collaboration-conference-international-trade-identity-across-borders) >>>>>>>> the risk of "diluting the autonomy and decentralization that SSI is meant >>>>>>>> to provide." >>>>>>>> >>>>>>>> The Sovereignty Shell Game >>>>>>>> >>>>>>>> Google was partnered with German Sparkasse on ZKP technology and >>>>>>>> that revealed a specific example of this co-opting. >>>>>>>> >>>>>>>> Google's open-sourcing of its Zero-Knowledge Proof libraries, >>>>>>>> announced July 3rd in partnership with Germany's network of public savings >>>>>>>> banks, was positioned as supporting privacy in age verification. Yet as >>>>>>>> [Carsten Stöcker pointed out]( >>>>>>>> https://www.linkedin.com/posts/dr-carsten-st%C3%B6cker-1145871_opening-up-zero-knowledge-proof-technology-activity-7348195852085067776-nKDB), >>>>>>>> zero-knowledge doesn't mean zero-tracking when the entire stack runs >>>>>>>> through platform intermediaries. Carsten noted that Google has "extensive >>>>>>>> tracking practices across mobile devices, web platforms and advertising >>>>>>>> infrastructure." Meanwhile, the Google Play API makes no promises that the >>>>>>>> operations are protected from the rest of the OS. >>>>>>>> >>>>>>>> The Google ZKP libraries ("longfellow-sk") could be a great >>>>>>>> [building block]( >>>>>>>> https://news.dyne.org/longfellow-zero-knowledge-google-zk/) for >>>>>>>> truly user-centric systems, as they link Zero-Knowledge Proofs to legacy >>>>>>>> cryptographic signature systems that are still mandatory for some hardware. >>>>>>>> But they'd have to be detached from the rest of Google's technology stack. >>>>>>>> Without that, there are too many questions. Could Google access some of the >>>>>>>> knowledge supposedly protected by ZKPs? Could they link it to other data? >>>>>>>> We have no idea. >>>>>>>> >>>>>>>> The European Union's eIDAS Regulation, set to take effect in 2026, >>>>>>>> encourages Member States to integrate privacy-enhancing technologies like >>>>>>>> ZKP into the European Digital Identity Wallet, but integration at the >>>>>>>> platform level offers similar dangers and could again invert the very >>>>>>>> privacy guarantees ZKP promises. >>>>>>>> >>>>>>>> Historical Echoes, Modern Inversions >>>>>>>> >>>>>>>> Identity technology's goals being inverted, so that identity >>>>>>>> becomes a threat rather than a boon, isn't a new problem. In ["Echoes of >>>>>>>> History"]( >>>>>>>> https://www.blockchaincommons.com/articles/echoes-history/), I >>>>>>>> examined how the contrasting approaches of Lentz and Carmille during WWII >>>>>>>> demonstrate the life-or-death importance of data minimization. Lentz's >>>>>>>> comprehensive Dutch identity system enabled the Holocaust's efficiency; >>>>>>>> Carmille's deliberate exclusion of religious data from French records saved >>>>>>>> lives. Even when they're decentralized, today's digital identity systems >>>>>>>> face the same fundamental questions: what data should we collect, what >>>>>>>> should we reveal, and what should we refuse to record entirely? >>>>>>>> >>>>>>>> But we're adding a new layer of complexity. Not only must we >>>>>>>> consider what data to collect, but who controls the infrastructure that >>>>>>>> processes it. When Google partners with Sparkasse on "privacy-preserving" >>>>>>>> age verification, when eIDAS mandates integration at the operating system >>>>>>>> level, we're not just risking data collection: we're embedding it within >>>>>>>> platforms whose business models depend on surveillance. Even if the data is >>>>>>>> theoretically self-sovereign, the threat of data collected is still data >>>>>>>> revealed — just as happened with Lentz's records. >>>>>>>> >>>>>>>> The European eIDAS framework, which I analyzed in a [follow-up >>>>>>>> piece to "Echoes from History"]( >>>>>>>> https://www.blockchaincommons.com/articles/eidas/), shows how even >>>>>>>> well-intentioned regulatory efforts can accelerate platform capture when >>>>>>>> they mandate integration at the operating system level. As I wrote at the >>>>>>>> time, a history of problematic EU legislation that had the best of >>>>>>>> intentions but resulted in unintended consequences has laid the groundwork, >>>>>>>> and now identity is straight in that crosshairs. One of the first, and most >>>>>>>> obvious problems with eIDAS is the mandate "that web browsers accept >>>>>>>> security certificates from individual member states and the EU can refuse >>>>>>>> to revoke them even if they’re dangerous." There are many more — and >>>>>>>> I'm not [the only voice]( >>>>>>>> https://news.dyne.org/the-problems-of-european-digital-identity/) >>>>>>>> on eIDAS and EUDI issues. >>>>>>>> >>>>>>>> Supposedly self-sovereign certificates phoning home whenever >>>>>>>> they're accessed is another recent threat that demonstrates best intentions >>>>>>>> gone awry. This not only violates privacy, but it undercuts some of our >>>>>>>> best arguments for self-sovereign control of credentials by returning >>>>>>>> liability for data leaks to the issuer. The [No Phone Home]( >>>>>>>> https://www.blockchaincommons.com/news/No-Phone-Home/) initiative >>>>>>>> that Blockchain Commons joined last month represents one attempt to push >>>>>>>> back on that, but it feels like plugging holes in a dam that's already >>>>>>>> cracking. It all does. >>>>>>>> >>>>>>>> The Builder's Dilemma >>>>>>>> >>>>>>>> What troubles me most is the split I see in our community. On one >>>>>>>> side, technology purists build increasingly sophisticated protocols in >>>>>>>> isolation from policy reality. On the other, pragmatists make compromise >>>>>>>> after compromise until nothing remains of the original vision. >>>>>>>> >>>>>>>> The recent debates about [`did:web` conformance]( >>>>>>>> https://github.com/w3c-ccg/did-method-web) illustrate this >>>>>>>> perfectly. Joe Andrieu correctly notes that `did:web` can't distinguish >>>>>>>> between deactivation and non-existence — a fundamental security >>>>>>>> boundary. Yet `did:web` remains essential to many implementation strategies >>>>>>>> because it bridges the gap between ideals and adoption. It provides >>>>>>>> developers and users with experience with DIDs, but in doing so undercut >>>>>>>> decentralized ideals for those users. We're caught between philosophical >>>>>>>> purity and practical irrelevance. >>>>>>>> >>>>>>>> In my recent writings on [Values in Design]( >>>>>>>> https://www.blockchaincommons.com/musings/ValuesDesign/) and the >>>>>>>> [Right to Transact]( >>>>>>>> https://www.blockchaincommons.com/musings/RightToTransact/), I've >>>>>>>> tried to articulate what we're fighting for. But values without >>>>>>>> implementation are just philosophy, and implementation without values is >>>>>>>> just surrender. >>>>>>>> >>>>>>>> The Global Digital Collaboration highlighted this tension >>>>>>>> perfectly. International progress on digital identity proceeds apace: >>>>>>>> Europe, Singapore, and China all advance their frameworks, but there are >>>>>>>> still essential issues that invert our fundamental goals in designing >>>>>>>> self-sovereign systems. Meanwhile, the U.S. remains even more stalled, its >>>>>>>> position represented only by the platforms that benefit from the status >>>>>>>> quo. Alongside this, technical standards discussions proceed in isolation >>>>>>>> from the policy, regulatory, and social frameworks that will determine >>>>>>>> their real-world impact. >>>>>>>> >>>>>>>> Where Do We Go From Here? >>>>>>>> >>>>>>>> I find myself returning to first principles. When we designed [TLS >>>>>>>> 1.0](https://datatracker.ietf.org/doc/html/rfc2246), we understood >>>>>>>> that technical protocols encode power relationships. When we established >>>>>>>> the [principles of self-sovereign identity]( >>>>>>>> https://github.com/WebOfTrustInfo/self-sovereign-identity/blob/master/self-sovereign-identity-principles.md), >>>>>>>> we knew that architecture was politics. Ongoing battles, such as those >>>>>>>> between Verifiable Credentials and ISO mDLs, between DIDComm and OpenID4VC, >>>>>>>> demonstrate disagreements over these power relationships made visible in >>>>>>>> technological discussions. >>>>>>>> >>>>>>>> The question now is whether we can reclaim our ideals before >>>>>>>> they're completely inverted by the side of centralized power and controlled >>>>>>>> architecture. >>>>>>>> >>>>>>>> The path forward requires bridging the gaps Geneva revealed: >>>>>>>> >>>>>>>> - Between corporate platform dominance and global digital >>>>>>>> sovereignty >>>>>>>> - Between the promise of decentralization and the reality of >>>>>>>> recentralization >>>>>>>> - Between technical standards and policy reality >>>>>>>> - Between privacy absolutism and implementation pragmatism >>>>>>>> >>>>>>>> A Personal Note >>>>>>>> >>>>>>>> After three decades of building internet infrastructure, I've >>>>>>>> learned that the most dangerous moment isn't when systems fail, it's when >>>>>>>> they succeed in ways that invert their purpose. We built protocols for >>>>>>>> human autonomy and watched them become instruments of platform control. We >>>>>>>> created standards for decentralization and saw them twisted into new forms >>>>>>>> of centralization. >>>>>>>> >>>>>>>> This conversation continues in private Signal groups, in conference >>>>>>>> hallways, in the space between what we built and what we've become. The >>>>>>>> [Atlantic Council warns]( >>>>>>>> https://dfrlab.org/2024/10/01/analysis-a-brave-new-reality-after-the-uns-global-digital-compact/) >>>>>>>> of power centralizing "in ways that threaten the open and bottom-up >>>>>>>> governance traditions of the internet." When critics from across the >>>>>>>> geopolitical spectrum — from sovereignty advocates to digital rights >>>>>>>> groups — all sense something amiss, it suggests a fundamental >>>>>>>> architectural problem that transcends ideology. >>>>>>>> >>>>>>>> Perhaps it's time for a new architecture: one that acknowledges >>>>>>>> these inversions and builds resistance into its very foundations. >>>>>>>> >>>>>>>> But that's a longer conversation for another day. >>>>>>>> >>>>>>>> --- >>>>>>>> >>>>>>>> *Christopher Allen has been architecting trust systems for over 30 >>>>>>>> years, from co-authoring TLS to establishing self-sovereign identity >>>>>>>> principles. He currently works on alternative approaches to digital >>>>>>>> identity through [Blockchain Commons]( >>>>>>>> https://www.blockchaincommons.com/).* >>>>>>>> >>>>>>>
Received on Friday, 18 July 2025 08:26:12 UTC