Hacker News Reader

Comment by tptacek | Hacker News Reader

I'm a broken record on why I believe this is a distinctively bad use case for public blockchains, for whatever it's worth. But it doesn't matter that I think this use case is doomed; what matters is: it's not a serious mainstream thing right now. If it becomes one, you won't have to argue; it'll be self-evidently an important use case.

View on HN

Sorry to make you repeat yourself, but we’re looking at DIDs and I’d like to know your arguments against them. Do you mind linking to a comment or summarizing?

Two words. Identity theft.

Could you elaborate a bit on that please? How would that pan out?

DIDs have nothing to do with blockchains, it's just that some people are trying to hype DIDs by artificially linking them with blockchains. They use "crypto" but the "old kind", the good one: cryptography.

While DIDs do not have to link to blockchains the standard itself notes:

Many—but not all—DID methods make use of distributed ledger technology (DLT) or some other form of decentralized network.[1]

[1] https://www.w3.org/TR/did-core/

I think it was this comment: https://news.ycombinator.com/item?id=29750165

I went looking (!) and I found this comment[1] from you. To summarize:

> SSI has no human mediated account recovery and so is a toy.

I don't completely disagree, but it's not the take I was expecting.

[1] https://news.ycombinator.com/item?id=29750165

Please explain what this is a bad use case. I'd love to learn why that is.

Not the person you're responding to, but my view on this is: Having one online identity that is publicly viewable by everbody is a privacy nightmare. Yes, I know that it's pseudonymous, but it's possible for a sufficiently motivated entity to find out who is behind a pseudonym.

Zero-knowledge cryptography allows you to post opaque data to the blockchain and selectively reveal properties of it to chosen people and applications.

For example, a zero-knowledge DID can generate a proof to an age-restricted application that you are over 18, without necessarily revealing your birthdate or any other information about yourself.

And this is already covered by the DID/VC W3C specs that don’t need a blockchain. An age restriction in particular will _always require the blessing of a central party_.

Because you can't remove stuff from a blockchain.

Change your name? Old name still there. Change your phone number? Still there. All old email addresses? Still there.

I reckon that's now how SSI solutions use a blockchain (or at least the ones I know of). AFAIK, a blockchain is only used to resolve a DID (an identifier mapping a random ID to a bunch of public keys you control, along with a URL to communicate with your device). Your personal information, as in your name, phone number, email stay in your device and are only exchanged with the party you want to authenticate against.

I'm curious to know if there's something I misunderstood in this space

Reply on news.ycombinator.com