Re: [webfinger] [appsawg] #12: Semantic gap for the client side

['Stephane Bortzmeyer' <[e-mail]>]

On Fri, Jun 14, 2013 at 03:09:44AM -0400,
 Paul E. Jones <[e-mail]> wrote 
 a message of 31 lines which said:

> I don't mind adding a statement that makes it explicitly clear that
> the example is not an IETF-approved means of provisioning an email
> client.

Then saying clearly that the example is not an example of present and
actual use: it's brainstorming about what WebFinger could do in the
future.
 
> With the examples in the spec, there have been comments ranging from
> "the example is too abstract" to "the example looks like it could be
> real" (the case with this email provisioning example).  We could
> remove all examples,

Indeed, some of the examples are wrong. In -14:

> Locating a User's Blog

OK for me (except the erroneous "en-us" that I already mentioned here)

> Identity Provider Discovery for OpenID Connect

OK for me (I assume not OK for Pete Resnick because there are no
details about the possible "rel" in the query and the possible links
in the response)

> Auto-Configuration of Email Clients

Useless and dangerous. 

> Retrieving Device Information

Dangerous, for security reasons, as mentioned in the IESG ballot