W3C home > Mailing lists > Public > public-webpayments@w3.org > March 2014

Re: "Web Identity" -> "Web Credentials"

From: Manu Sporny <msporny@digitalbazaar.com>
Date: Thu, 13 Mar 2014 22:03:34 -0400
Message-ID: <53226376.2060708@digitalbazaar.com>
To: public-webpayments@w3.org
On 03/11/2014 01:14 PM, Melvin Carvalho wrote:
> Loosely speaking, In Foaf you have a Person, and you have a the
> super class which is an Agent which can be a robot, human, group or
> corporation.

Ah, so here's the disconnect. I'm not talking about FOAF at all. I'm
talking about the English language definition of "Agent":

a·gent ˈājənt/ noun
1. a person who acts on behalf of another.
2. a person or thing that takes an active role or produces a specified

You are talking about foaf:Agent, which deviates from the English
language definition quite wildly (and is thus very confusing to people
that don't know about FOAF, which is the vast majority of the Web
Developer population):


Agent - An agent (eg. person, group, software or physical artifact).

We should go with English language definitions because that's what
people are familiar with (not some esoteric Semantic Web thing). Dan
Brickley, who created FOAF, now works for Google on schema.org. The spec
is dead. Schema.org killed it. We should stop using it.

> The super class of Agent I think is a "Thing".
> "Agent" itself is not tied to foaf in the Web Identity spec, it seems
> to be more or less the same thing you are saying.

You could say that what we are talking about is the identity of a Thing.
Again, going back to the English language:

i·den·ti·ty īˈdentitē/ noun
1. the fact of being who or what a person or thing is.

Let's stop using esoteric Semantic Web vocabularies that are now over 14
years old. Primarily, let's stop using them because they don't line up
with the standard English definition, which is going to confuse Web

> "identity" A set of information that can be used to identify a
> particular entity such as a person, agent, or organization. An entity
> may have multiple identities associated with it.
> "identity URL" An identity URL consists of an HTTP or HTTPS scheme
> and denotes an identity.
> "identity document" (don't know if this is necessary) A document that
> exists at an identity URL and contains statements about an identity.
> May help with an example
> I am a type of Person.  Which is a type of agent.  In your
> terminology am I an Identity, I think not.  So what am I?

You're down one level too many. You're at the Semantic Web Vocabulary
level. "identity" is at the conceptual level. You are a thing that is a
person. Information about you is your identity, and you can have
multiple identities depending on which subset of information about you
is used. The standard English definition of "identity" is what we're
using in the spec.

> The digital string that "denotes" me on the web, is 
> http://melvincarvalho.com/#me

That's a perfectly fine way to denote yourself on the Web.

> The (profile/identity) document which contains more information about
> me can be dereferenced at http://melvincarvalho.com/ and specific 
> statements about me found at the #me anchor.

Fine as well.

> So I wonder is identity a super class of agent?

No, that's a strange FOAF-ism. I'm arguing that we should use the
English language definition of "agent", which is:

a·gent ˈājənt/ noun
1. a person who acts on behalf of another.
2. a person or thing that takes an active role or produces a specified

And an identity is this:

i·den·ti·ty īˈdentitē/ noun
1. the fact of being who or what a person or thing is.

Agents can have an identity.
Things can have an identity.
People can have an identity.

> In the terminology above which is the Identity, the Identity URL and
> the Identity Document?

Identity is abstract.

Identity URL is http://melvincarvalho.com/#me

Identity Document is the thing that you have when you HTTP GET

> From my POV, I already implement turtle, but would not be a huge
> effort for me to add json ld I think.

For you it's not a huge effort because:

1) You studied mathematics under Stephen Hawking, which probably means
   you're incredibly intelligent. :)
2) Have a deep knowledge of the Semantic Web and its intricacies as well
   as programming
3) Already know TURTLE and would find it easy to implement just about
   any data expression language.

Most Web Developers do not easily fall into any of the 3 categories
above. This is something that is typically lost on most Semantic Web
folks. So, the question becomes: What are the set of technologies that
we could use that are most familiar to Web developers that also provide
the benefits of the Linked Data stack?

That's why we're using JSON-LD.

> We're currently working on a proposal such that domains that are not 
> email providers may still vouch for an email address. So, for
> example, you could still login to a website using "melvin@gmail.com 
> <mailto:melvin@gmail.com>", but your personal website
> "http://melvincarvalho.com/" could still vouch for the email
> address.
> Great.
> So what about content addressable identifiers like bitcoin:
> addresses?

Yep, should also work just fine.

> At this point I'm unsure DHT's give you anything the web doesnt!
> I'm convinced that decentralized systems can be built using the web,
> tho I've yet to prove it!

One of the Web's fundamental flaws is that the protocols it uses, like
DNS, HTTP and TLS, lead to centralization as a byproduct of its design.
It's the reason we have things like Gmail, Facebook, Twitter, WhatsApp,
Google Search, etc.

Things like Telehash and Kademlia tend to foster more decentralization
than the Web because no player on the network has "special privileges"
(like the root CAs, core network routers, etc.).

You could argue that any decentralized system can be built using the Web
just like any program can be written as long as you have a Turing
complete language. Just because you could re-implement the Web using
Fortran doesn't mean that you'd want to. Specialization is a good thing. :)

-- manu

Manu Sporny (skype: msporny, twitter: manusporny, G+: +Manu Sporny)
Founder/CEO - Digital Bazaar, Inc.
blog: The Worlds First Web Payments Workshop
Received on Friday, 14 March 2014 02:04:05 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 19:07:28 UTC