- From: Christopher Allan Webber <cwebber@dustycloud.org>
- Date: Mon, 16 Oct 2017 15:32:14 -0500
- To: Manu Sporny <msporny@digitalbazaar.com>
- Cc: Melvin Carvalho <melvincarvalho@gmail.com>, W3C Credentials CG <public-credentials@w3.org>
Manu Sporny writes: > On 10/16/2017 10:58 AM, Melvin Carvalho wrote: >> Why would that want to be removed? > > There were a few folks from the Bitcoin BTCR DID camp that asserted that > you can assume who the key owner is if the key is listed in the DID > Document, which I believe is true (without putting much thought into it). Right... it's not strange to us, Melvin, because we're thinking about the general linked data case. Some people haven't been thinking about signatures from a general linked data perspective until this group, and so it "appears" that having a "single" object with an embedded fragment identifier sub-object that it's strange that the key is pointing back at itself. Eg, highly abbreviated: {"id": "did:example:foo", "somethingSomethingKey": { "id": "did:example:foo#key1", // ... rest of key stuff here ... "owner": "did:example:foo"}} If you're *just* looking at this from a single "nested json object" and the assumption that you'd always have a "nested" object with a fragment identifier, it looks like the object is pointing back at itself. But of course if you look at it from a triples/graph perspective, *especially* in the general case when that key may even be a different document living somewhere else and not "nested" like that, that this ownership would not be implied and it would be necessary to spell out. Anyway, that's why it was so confusing. It was confusing to realize where the confusion was, too! :) But the confusion is understandable if you aren't coming from the general linked data world...
Received on Monday, 16 October 2017 20:32:40 UTC