Re: Shapes question

On Wed Dec 24 2014 at 1:16:09 PM Eric Prud'hommeaux <eric@w3.org> wrote:

>
> On Dec 20, 2014 1:03 AM, "Judson Lester" <nyarly@gmail.com> wrote:
>
> I don't recall seeing mail to public-rdf-shapes@w3.org (as was Richard's
> suggestion). I wonder if your approach looks like what Google submitted to
> the RDF Validation Workshop. Can you send an example?
>
There's a RDFS with comments at
https://github.com/nyarly/road-forest/blob/master/path.ttl . The same
project implements part of the idea. The server can send Paths as part of
resource affordances to describe the kinds of graphs that it accepts.
Clients can use the affordances to produce subgraphs of their own modified
state in order to submit it back to the server.

The captures sound like the semantic actions in Shape Expressions (see
> later slides in <http://www.w3.org/2014/Talks/1209-shex-egp/>).
>
> Well, captures are a best a wishlist feature at the moment. In particular,
though, the difference is that captures more resemble Regex capture groups
- the idea being that a client could sift through server responses both for
data and for graph nodes as they interpret the responses. It's possible
SPARQL might be more appropriate for this task, since those queries aren't
meant to be transmitted between client and server.

Judson