- From: Kingsley Idehen <kidehen@openlinksw.com>
- Date: Tue, 03 Jan 2012 06:41:48 -0500
- To: public-xg-webid@w3.org
On 1/3/12 1:46 AM, Peter Williams wrote: > Excellent, I dumbed it down for Cromwell's army of Puritans : http://tinyurl.com/7kmjo7l. > > > > With this, I can build a distribute root cert store, and get passed Mozilla controlling roots (on behalf of whoever pays to play, in older root stores or more modern websso discovery points).Then, one can start to use the search space to build topologies of cert chains, walking up the AIA URIs (that point back to one or more parents/previousParties in the directed graph/topology). > > > > If it's not already happened, we need to get Mo and Kingley in the same room, so we find just the right balance between technology and message.This is critical to a sound presentation of the technology and the opportunity. We dont sound like technologists, but the technology clearly has class. > > > > I suspect the reification and semantic signature is too advanced. Yes it is at this stage. Ultimately, it provides another pathway. > Right now, we JUST need to (a) do henry's keys (the world of no certs client authn, aping PGP-grade logon), and (b) make the semantic web just deliver a distributed certificate store, for good old X.509 blobby roots that secure the profile endpoints supporting (a). That's fine at this stage. > > > > The only reason I used the data uri (for the cert) was becuase it was the only way I could find to represent a mime-described blob in the semantic web. We can use (or add) terms from an ontology (maybe WOT) for holding the blob. Basically, the blob then holds the carbon copy of the cert. This is what I mean when I say verification can be based on at least one or all of the following: 1. Public Key components 2. Fingerprint 3. Entire Cert. Blob. Verification is about a lookup against a mirror claim (in full or vital splices). > Being a triple-world, semantic web ontologies tens (rightly) to veer away from tagging what would be endless series of blob. Correct. > But, here thats just what I need. Yes, but ultimately, the system has to be a little purer, hence the inevitability of reification en route to graphs with signed triple based claims that mirror what's in an X.509 cert. > I happen to have a blob that is part of an older linked data tradition, that is carrying forward into the world noth of securing and assuring the more modern linked data tradition's protocols (i.e. the https endpoints of simple profiles and the more sparql servers providing sparql verbs, alike). Okay :-) > > > > > > > -- 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 Tuesday, 3 January 2012 11:42:11 UTC