Re: Thinking about Webfinger

There's no need for any changes for any URIs with a host component (any containing an @ or //, broadly)

Though you're correct that it's not defined in the case of other structures, those are comparatively rare (though some of them are certainly interesting); and some of them come with discovery methods implicit in their design

On Sun, 7 May 2023, at 00:58, Melvin Carvalho wrote:
> 
> 
> ne 7. 5. 2023 v 0:12 odesílatel Bob Wyman <bob@wyman.us> napsal:
>> Johannes,
>> Okay. I understand why people don't like content-negotiation, even if it is explicitly mentioned in the WebFinger RFC. But, can you explain what, if anything, is wrong with:
>>  • Using WebFinger to resolve did:* URIs, and
>>  • Returning a did document, or a link to one, as either a property or link in a JRD response?
>> Why isn't the existing WebFinger specification sufficient to allow useful lookup of did documents?
>> 
>> I'm thinking that:
>>> https://example.com/.well-known/webfinger?resource="did:web:w3c-ccg.github.io:user:alice"
>> might return either:
>> 
>> 1. A link to a did document:
>>> {
>>>   "subject": "did:web:example.com:user:alice",
>>>   "links": [
>>>     {
>>>       "rel": "http://example.com/rel/did-doc",
>>>       "href": "https://example.com/user/alice/did.json"
>>>     }
>>>   ]
>>> }
>> (Note: If this was commonly used, someone might want to register the appropriate rel type with IANA. But, that's not necessary.)
>> 
>> or,
>> 2. A did document as a JRD property:
>> 
>>> {
>>>   "subject": "did:web:example.com:user:alice",
>>>   "properties": {
>>>     "http://example.com/did-doc": {
>>>       "@context": [
>>>         "https://www.w3.org/ns/did/v1",
>>>         "https://w3id.org/security/suites/jws-2020/v1"
>>>       ],
>>>       "id": "did:web:example.com:user:alice",
>>>       "verificationMethod": []
>>>     }
>>>   }
>>> }
>> (Note: If this was commonly used, someone might want to register an appropriate property type with IANA. But, that's not necessary.)
>> 
>> Is there some reason why either of these would not be useful and appropriate?
> 
> Webfinger is primarily designed to look up user@host identifiers, which are conceptually similar to mailto:user@host. The "acct:" prefix might be considered unnecessary, but that is the current convention.
> 
> Webfinger was not intended to handle other URI schemes, and accommodating them could involve significant changes. With hundreds of schemes available, including over 200 sub-protocols for DIDs, each scheme typically has its own lookup mechanisms. User@host identifiers are prevalent on the social web, and integrating them into the world of linked data, including lookup, will be essential.
> 
> It would be intriguing to examine Webfinger and JRD usage statistics to gain insights into their adoption and deployment levels.
> 
>  
>> 
>> bob wyman