Re: Chairing, Superspec, subspecs and WebID 1.0

On Sat, Nov 11, 2023 at 8:51 PM Aaron Coburn <aaronc@inrupt.com> wrote:

> This sounds really good.
>
> In addition to what Jacopo wrote, one thing that might help us reach
> consensus on a WebID specification is to start with a collection of clearly
> articulated use cases. If we agree on the use cases, it can be easier to
> find agreement on the specification that follows.
>
> While this would add an extra step at the beginning of this process, this
> can help clarify the scope and direction for WebID.
>

Considering this was a lot harder than I thought it would be. Given:

1) we can assert for example <x> foaf:knows <y>, and that entails that <y>
a :Agent, and <x>, via range and domain.
2) all other aspects are factored out (for example media types, profiles,
auth, keys, etc)

Then the use case reduced down to, why would I need that single statement
(<x> a :Agent ) to be established via dereferencing the URI <x> rather than
any other means (see 1.)

I am left with: there is value in me saying "you are an Aaron" (<you> :name
"Aaron"), and then in you confirming "I am Aaron" (<#me> :name "Aaron"),
and that value is established via the dereferencing function.

Secondly, it promotes good form, and introduces the benefits of
dereferencing which are greatly enhanced when you factoring in the things
which come after (2.)

Finally, if you encounter wild data, out of a follow your nose context,
it's useful for all manner of functionality to be able to: if(x an Agent)
then.

That's all I can come up with.