Clarify relation between target system and DID for https method about specs

[oed]

In the current spec, there is only on single DID for each target system.

@friedger Yes I would say this is by design. If you want to specify an ethereum address as your DID then you could simply use did:ethr:0xb9c5714089478a327f09197987f16f9e5d936e8a or similar. Same goes for any other type of identifier. The https-did is intended to be used for a single DID per domain.

[friedger]

@oed Then the DID should look something like did:https-w3c-ccg.github.io:1 for the first registered DID on this domain, or did:https-w3c-ccg.github.io:2 for the second after the domain name was transferred ?

[oed]

No, the domain should be the identifier, not a part of the method. For example you could have did:https:1.example.com and did:https:2.example.com to achieve the same thing you are suggesting.

[friedger]

@oed But then the target system is wrongly specified. The DID method name should describe the target system. And you run into problems how a DID client can create a domain name, that proof of control is equal to access to the web host and the DNS registrars can delete your ID.

Currently, it says:

The target system of the HTTPS DID method is the web host that the domain name described by the DID resolves to when queried through the Domain Name System (DNS).

By that specification the target system is the web host of example.com.

Following your argument, the target system is the DNS system.

[oed]

Following your argument, the target system is the DNS system.

Yeah I think that's exactly what is intended with this DID method.

[awoie]

@friedger @oed I agree, the target system would be the DNS system.