Re: Should we complete the WebID spec?

pá 3. 11. 2023 v 18:00 odesílatel Kingsley Idehen <kidehen@openlinksw.com>
napsal:

>
> On 11/3/23 9:47 AM, Melvin Carvalho wrote:
>
>
>
> pá 3. 11. 2023 v 14:12 odesílatel Kingsley Idehen <kidehen@openlinksw.com>
> napsal:
>
>>
>> On 11/2/23 9:48 PM, Melvin Carvalho wrote:
>>
>>
>>
>> pá 3. 11. 2023 v 1:09 odesílatel Kingsley Idehen <kidehen@openlinksw.com>
>> napsal:
>>
>>>
>>> On 11/2/23 5:53 PM, Melvin Carvalho wrote:
>>> > 2. urn scheme for webids:  urn:webid
>>>
>>> What's that, and why?
>>>
>>> A WebID has always been an HTTP based URI used to name an Agent.
>>>
>>
>> Yes indeed.  But not every authentication process gives a URI (sadly).
>> Some will give a string of characters that denote an Agent, which need to
>> be turned into a URI.
>>
>>
>>
>> I didn't see our focus right now being on solving for authentication
>> protocols and URI combinations. I believe the primary objective is to
>> propel WebID forward by delineating its essence and its potential usage,
>> especially within the context of Solid.
>>
>> The shift from understanding WebID as an HTTP URI that identifies an
>> Agent to something else doesn't align with the aforementioned goal.
>>
>> Presently, we utilize the term NetID for a URI that identifies an Agent,
>> grounded on the reality that we have authentication services accommodating
>> a range of schemes right from the outset.
>>
>> The discourse around WebID and Solid Lite seems to diverge slightly from
>> the issues pertinent to a NetID. Furthermore, with billions of pages on the
>> Web currently utilizing HTTP URIs to identify Agents, it seems prudent to
>> harness this established practice with the existing WebID definition.
>>
>> Isn't the ultimate aim to employ HTML profile documents, encompassing RDF
>> structured data islands (i.e., metadata articulated through various RDF
>> notations), via a more streamlined derivative of Solid?
>>
>
> Kingsley, you are completely correct.  I agree with everything.
>
> I'm just outlining the things I would like to complete to make a working
> system, bearing in mind that this comunity group may close.
>
> The problem of turning a string into a URI post authentication, is
> something I'll need to do.
>
> Turning:  "userid" into urn: <something> : userid is a quick hack.
> Whether it's webid or netid, the software doesnt care, so long as it's
> consistent.
>
> It would be in this case something like an "indirect identifier" as
> described in awww.  Example:
>
> "Today 10 Downing Street said that ..."
>
> Of course the building 10 Downing Street didnt say anything.  It's an
> incorrect sentence.  But the consumers of the sentence understand it well
> enough.
>
>
> You handle that using blank nodes, which is basically the default in JSON
> or JSON-LD without explicitly indicating an object id. Bascially, your
> example is ground zero for JSON and JSON-LD where both denote subjects
> using indefinite pronouns (a/k/a blank nodes).
>
>
> Example:  What is a WebID URN?  A webid URN is a web identifier that is a
> URI but where the software was unable to locat an HTTP URI but wishes to
> store the authenticated username as a URI.
>
> I need to get something working in any case, in order to have a pluggable
> auth system.
>
>
> My point is that you are swimming against the current if you want to
> venture down this path. More importantly, it cannot be named WebID -- since
> that's a massive change from the original definition.
>

Not sure I agree on this, but I'll back burner it for now, given the push
back.

Let me give some context:

There are two things, one is a string as a name "kidehen" the other is a
universal name aka a urn which is also a URI

So one issue with the urn spec is:

I'd like to just say:

urn:kidehen

Problem solved.

Unfortunately this brakes the URN spec, because for some strange reason
it's not allowed, you have to subclass it:

so it has to be:

urn:<something>:kidehen

And whatever goes in there needs to be consistent.  This is simply a
constraint imposed by RFC8141.  I wish it wasnt there, then this would not
be an issue, but I need to put in something logical to pass the Test of
Independent Invention (TOII).

I am unsure I agree with your assessment on a webid namespace, but I'll
give it more thought.  In any case it will likely be an internal software
matter, at this point.


>
> --
> Regards,
>
> Kingsley Idehen 
> Founder & CEO
> OpenLink Software
> Home Page: http://www.openlinksw.com
> Community Support: https://community.openlinksw.com
> Weblogs (Blogs):
> Company Blog: https://medium.com/openlink-software-blog
> Virtuoso Blog: https://medium.com/virtuoso-blog
> Data Access Drivers Blog: https://medium.com/openlink-odbc-jdbc-ado-net-data-access-drivers
>
> Personal Weblogs (Blogs):
> Medium Blog: https://medium.com/@kidehen
> Legacy Blogs: http://www.openlinksw.com/blog/~kidehen/
>               http://kidehen.blogspot.com
>
> Profile Pages:
> Pinterest: https://www.pinterest.com/kidehen/
> Quora: https://www.quora.com/profile/Kingsley-Uyi-Idehen
> Twitter: https://twitter.com/kidehen
> Google+: https://plus.google.com/+KingsleyIdehen/about
> LinkedIn: http://www.linkedin.com/in/kidehen
>
> Web Identities (WebID):
> Personal: http://kingsley.idehen.net/public_home/kidehen/profile.ttl#i
>         : http://id.myopenlink.net/DAV/home/KingsleyUyiIdehen/Public/kingsley.ttl#this
>
>