Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

A Primer for Decentralized Identifiers

w3c-ccg.github.io

Post reply on HN
Print view View original

1–10 of 16 posts

Re: A Primer for Decentralized Identifiers

#3

Post by clairity »

networkimprov wrote: post #2

See also: https://en.wikipedia.org/wiki/Self-sovereign_identity https://identity.foundation/ https://human-id.org/ (a non-profit startup)

it's a little disheartening that all of the sites you link to include conglomerated tracking mechanisms, e.g., proprietary tracking for wikipedia (even if this is relatively benign), google for the latter two, plus wordpress for human-id.org.

Re: A Primer for Decentralized Identifiers

#5

Post by capableweb »

max_ wrote: post #4

For privacy preserving IDs I suggest you look at Coconut Credentials — https://medium.com/chainspace/coconut-threshold-issuance-sel...

How is DIDs not privacy preserving? You chose what information you put in it. Also, Coconut Credentials seems to be for "distributed ledgers" while DIDs are agnostic in their usage.

Re: A Primer for Decentralized Identifiers

#6

Post by max_ »

capableweb wrote: post #5
max_ wrote: post #4

For privacy preserving IDs I suggest you look at Coconut Credentials — https://medium.com/chainspace/coconut-threshold-issuance-sel...

How is DIDs not privacy preserving? You chose what information you put in it. Also, Coconut Credentials seems to be for "distributed ledgers" while DIDs are agnostic in their usage.

I don't know much about DIDs this was my first encounter.

I was just sharing.

Re: A Primer for Decentralized Identifiers

#8

Post by capableweb »

lucb1e wrote: post #7

Aside from a spec that tells the "how", I don't understand what or why it is. What problem does this solve? How can or should I use this?

It's mentioned in the "How DIDs Differ from Other Globally Unique Identifiers" section.

> The need for globally unique identifiers that do not require a centralized registration authority is not new. UUIDs (Universally Unique Identifiers, also called GUIDs, Globally Unique Identifiers) were developed for this purpose in the 1980s and standardized first by the Open Software Foundation and then by IETF RFC 4122.

> As a rule, however, UUIDs are not globally resolvable and URNs – if resolvable – require a centralized registration authority. In addition, neither UUIDs or URNs inherently address a third characteristic – the ability to cryptographically verify ownership of the identifier.

TLDR: We've used UUIDs for a while now, they're cool and all but don't work in a distributed network as they assume location-addressing to one central, verifying server. Instead, DIDs work via content-addressing, making them more ideal in distributed networks as they're easier to verify.

Re: A Primer for Decentralized Identifiers

#9

Post by capableweb »

max_ wrote: post #6
capableweb wrote: post #5

Earlier quoted context omitted.

How is DIDs not privacy preserving? You chose what information you put in it. Also, Coconut Credentials seems to be for "distributed ledgers" while DIDs are agnostic in their usage.

I don't know much about DIDs this was my first encounter. I was just sharing.

Sorry, I wasn't meant to be abrasive or anything, it's a genuine question. When you say "For privacy preserving IDs..." it sounds like DIDs are not privacy preserving, while the people who worked on DIDs did in fact think about privacy when designing it. But maybe I'm understanding your previous comment wrong, sorry if that's the case.

Re: A Primer for Decentralized Identifiers

#10

Post by tgbugs »

I find the choice to use ownership language strange. The url/uri based systems are fundamentally inadequate for many if not most of their current use cases (they are not and never were intended to be persistent), and as the authors point out, there the urn system requires a central authority. Thus there is a need for some other system.

The ability to claim and prove authorship and determine the provenance (or tampering) of an identified document is crucial. For the most part people bindly trust that the nytimes.com that they are seeing is the real one, but there are so many ways that the can be compromised it is hard to comprehend.

More importantly, the current ways of protecting oneself against such issues are no longer accessible to the common man except via an interaction with some giant corporate edifice that they also have to trust. If you want to publish a document and distribute it in a way that its integrity and authorship can be determined there aren't good options.

Even more more to the point, there is currently no way general way to give a link to such a document that can be dereferenced to an untrusted host in such a way that the integrity of the document can be verified.

I supposed we could all share magnet links and check that they were signed with the nytimes pgp key, but then we are back to the issue of how to determine and distribute the nytimes pgp key. The WOT doesn't work, and pgp fails for anonymous publishing. Maybe the infrastructure behind DIDs will be able to solve the problems with WOT.

It will be interesting to see what comes of this, but I'm wary of its adjacency to the blockchain buzzwords.

Post reply on HN