Re: WebID 1.0 -- Section 3 -- Removal of Note

On 2/18/13 11:13 AM, Michael Hackett wrote:
> On 17 February 2013 17:18, Stéphane Corlosquet <scorlosquet@gmail.com 
> <mailto:scorlosquet@gmail.com>> wrote:
>
>     We use hash URIs in all our examples, and people who are new to
>     WebID and looking at implementing this will use hash URIs.
>
>
> Just to echo what Mo said, people won't use them if they don't know 
> what they are, or rather why they are used. To someone just looking at 
> WebID as a distributed identity system, who might have lots of 
> experience building web sites (or maybe not so much), but no 
> experienced with Linked Data, the hashes don't stand out as 
> significant. I think it would be very helpful if the spec included a 
> brief explanation of their use and a link to more in-depth reading. 
> (Don't just point to a long external document, as developers will not 
> be compelled to read a long doc if they don't immediately see the value.)
>

Links from a prior posts in this thread:

1. http://lists.w3.org/Archives/Public/public-webid/2013Feb/0029.html -- 
TimBL presentation link (covers indirection via hash and hashless URIs; 
note that the hashless slide has a few typos that leads to confusion)

2. http://lists.w3.org/Archives/Public/public-webid/2013Feb/0060.html -- 
provided references to materials that address the tradeoffs associated 
with either style of HTTP URI in the context of Linked Data.

All:

For those of you that don't understand my fundamental frustration with 
Henry (as Chair) and Andrei (as Editor) it boils down to this:
They are selective and obstructive when dealing with responses that 
differ from theirs. For instance, they reacted to the vote on the 
definition of a WebID by inserting an unnecessary notice.

How should this have been addressed, if this was genuinely about clarity 
as opposed to a backdoor mechanism for negating the vote? They could 
have done exactly what's outlined above, add links to relevant documents 
that cover the different types of HTTP URIs and their usage 
implications. Even better, they could have looked to the following 
document collection to really make things clear:

1. WebID -- definition document.
2. WebID oriented Profile Document -- defines the fundamental 
characteristics of a WebID oriented profile document.
3. WebID oriented Profile Document publishing -- how to publish said 
document (a natural place to shed light on HTTP URI choices) .
4. WebID authentication protocol -- defines the WebID+TLS protocol 
that's used to verify a WebID based on its association with a WebID 
profile document.

1-4 would be cross referenced using their respective URLs from the 
relevant documents.

Here are some important points about Linked Data:

1. Linked Data is about Data Representation and Access that leverages 
RDF, HTTP, and core architecture of the Web.
2. HTTP URIs in this context have tradeoffs that affect consumers, 
publishers, and agents that oscillate between either role.
3. Implementing a Linked Data solution is fundamental to understanding 
its many nuances.
4. Demonstrating Linked Data comprehension is best done via Links.

Ironically, I've been through this loop with RDF/XML, and today RDF is 
finally (12+ years later) decoupled from that classic example of 
conflation gone wrong. The only thing you get from failure to separate 
powers is confusion that ultimately compromises the fundamental goal 
i.e., in this case: leveraging Linked Data en route to producing a spec 
for Web-scale verifiable identity.


To conclude, a W3C community effort isn't about the personal preferences 
a chair person or editors. It's about a communal effort to fashion a spec.

-- 

Regards,

Kingsley Idehen	
Founder & CEO
OpenLink Software
Company Web: http://www.openlinksw.com
Personal Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter/Identi.ca handle: @kidehen
Google+ Profile: https://plus.google.com/112399767740508618350/about
LinkedIn Profile: http://www.linkedin.com/in/kidehen