W3C home > Mailing lists > Public > public-xg-webid@w3.org > January 2012

Re: hCard and WebID

From: Kingsley Idehen <kidehen@openlinksw.com>
Date: Wed, 11 Jan 2012 15:05:55 -0500
Message-ID: <4F0DEBA3.3010704@openlinksw.com>
To: public-xg-webid@w3.org
On 1/11/12 1:53 PM, Peter Williams wrote:
> You need to layout the hcard thing, a little. its very hard keeping religions straight, and in their right box - when along comes the semantic web and smushes it all together again. And lets not focus on formats, but on how the semweb wants to work with it - as the ultimate religion killer.

Yes!

>
>
>
>
> So there I was on my webid proxy profile navigating around in the uriburner "data space", looking at stuff its proxying from my windows page, suitably crawled/sponged/Pinged. Said data space then offers a view on the declared entities in that data set, replete with links to n recastings, n describings, and n formats. I now udnersatnd that another "data space" could do the same - and even offer a different perspective by default (and different mashup of sources, perhaps). This "data space" is quite different to my little personal facebook graph that such a "data space" might "connect to".
>
>
>
> Ill distinguish it as "a data cloud", in the public-data-cloud sense. It can agument my crappy windows profile's endpoints (with all the linked data tech limits) and give it's (very same) profile "full presence", at a different endpoint (with full compliance). Full compliance enables it to work with an engineered security policy, assuming actual secure nameservers for the public virtual name space (and several "well known" competititors, no doubt)
>
>
>
> Ok, nirvana is close by. We have moved beyond pgp-isms in the security space, dribbling about me and my 3 friends and their keys, as we plot to overthrow Assad and his lot. And, lots of legacy can participate in the full compliant web (so crucial for the the "trusted name resolution' to work, and work suitably for engineering grade security policies).
>
>
>
> now nirvana is not quite there, since I eliminated my https (and http) endpoint. Yet http://uriburner.com/about/html/http://uriburner.com/about/id/entity/https/idweb.cloudapp.net/Home/About%01me is still around, giving a false impression.Should a validation agent with SAN URI citing that work? remember, this is acccess control (not just fun consumer-grade following stuff). Does the nuke explode or not?
>
>
>
> For this kind of architecture, what specifically must the validation query do, beyond the simple ASK?

Authentication is covered by ASK.
Authorization requires much more than that. But, without Authentication 
one can't really nail Authorization in the loosely coupled manner we are 
all seeking.

>   It seems to need some liveness test for the original endpoint (and not just a caching expiry test). This doesnt just mean DNSSEC (or the US's desire to have the final offswitch)
>
>
>
>
>
> Back to hcard, though. There isn't one. Or is there?

There is a Microformat for hCard.

Example, but *ignore* the URI since its for example purposes only. It's 
also straight out of QA (which is in progress).


<div id="hcard" class="vcard">
<a class="url fn" 
href="http://id.myopenlink.net/mv/data/afacf6b868e6cb36f71d2cecc6d531a26ce1f2ef#this">@kidehen 
(BrowserID)</a>
<a class="email" 
href="mailto:kidehen@openlinksw.com">kidehen@openlinksw.com</a>
<a class="key" 
href="data:application/x-x509-user-cert;base64,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">Public 
Key</a>
</div>


You can cut and past the output above into an HTML doc. Publish the doc 
to a space on the Web and then use the URI <{ResourceURL}#this> in the 
SAN of the x.509 you generate and persist to your local keystore.

>
>
>
> We learn from uriburner://....https/.../idweb that its the same as uriburner://...http/...yorkporc in blogspot land.
>
>
>
> http://uriburner.com/about/html/http://uriburner.com/about/id/entity/http/yorkporc.blogspot.com/%01me does not seem to show (or link obviously) to the vcard, once crawle ...even, if I wander up to the documument, type.

We have a translator for vCard (binary) and hCard (microformats). Have 
you tried passing the URL of a Microformat bearing resource via URIBurner?
>
>
>
> According to http://morph.talis.com/?data-uri%5B%5D=http%3A%2F%2Fyorkporc.blogspot.com&input=&output=html&callback= we see vcard information that the blogspot template (around my bookmarked post) is publishing - above and beyond the facts that I publish in my bookmarked posting.
>
>
>
> And this is where I get lost in the model.

It could be issues with the cartridge/driver for vCard or hCard. Thus, 
if I have a resource URL  I can decipher what's going wrong.

>   There are so many views, and some have some and not other data instances. Sometimes I see the vcard pop up in the alternate view (of my RSS feed) in semweb landm, and other times not. While this VERY good for my immediate need (membership entity federations, all n of them), I cannot actually fathom the ones I'm looking at, and have.

I will be testing all of this very soon anyway, we are code complete, 
just doing QA re. hCard and WebID :-)


>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>    		 	   		
>


-- 

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








Received on Wednesday, 11 January 2012 20:06:18 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 11 January 2012 20:06:19 GMT