- From: Gregg Kellogg <gregg@kellogg-assoc.com>
- Date: Sun, 10 Jul 2011 13:55:24 -0400
- To: Manu Sporny <msporny@digitalbazaar.com>
- CC: Linked JSON <public-linked-json@w3.org>
- Message-ID: <6D2DDA50-7BDB-4D60-90AC-90FC4CDE3C31@kellogg-assoc.com>
Manu, this is great! A couple of points: 1) The Turtle output could be a bit better, I think the only reason to show Turtle is that it makes the graph easier to understand. This is the output from my Turtle writer: @base <> . @prefix : <http://rdf.data-vocabulary.org/#> . @prefix xsd: <http://www.w3.org/2001/XMLSchema#> . [ :ingredients "1 cup ice cubes", "1 tablespoons white sugar", "1/2 cup club soda", "1/2 lime, juiced with pulp", "12 fresh mint leaves", "2 fluid ounces white rum"; :instructions [ :description "Pour white rum over ice."; :step "3"^^xsd:integer], [ :description "Garnish with a lime wedge."; :step "5"^^xsd:integer], [ :description "Fill the rest of glass with club soda, stir."; :step "4"^^xsd:integer], [ :description "Fill glass to top with ice cubes."; :step "2"^^xsd:integer], [ :description "Crush lime juice, mint and sugar together in glass."; :step "1"^^xsd:integer]; :name "Mojito"; :yield "1 cocktail"] . (It could be even better if, e.g., "1"<xsd:integer> were just replaced with a bare '1'). 2) Is there a framing spec for JSON-LD? I like the concept of using the structure from a template document to control chaining/framing output. Gregg On Jul 9, 2011, at 2:24 PM, Manu Sporny wrote: Hey folks, There is now a live, Web-based JSON-LD processor available here: http://json-ld.org/playground/ It uses the latest experimental JavaScript JSON-LD processor (Forge's JSON-LD processor) written by Dave Longley. The processor includes some of the latest changes we've been discussing on the mailing list. "@" is now "@subject". "a" is now "@type". It also implements the latest normalization algorithm that Dave's been working on, which he will explain in a separate e-mail. The new normalization algorithm works with all of the known, really nasty graph isomorphism inputs that the old algorithm didn't handle. At present, we do not know of a single case that the algorithm cannot handle. The website also got a bit of a facelift. Not much of one since I'm no graphic designer and my CSS skills are fairly lame. The biggest addition is the Playground code, which allows you to: 1. Type out arbitrary JSON-LD and view it in compact form, expanded form, normalized form, framed form and as TURTLE. 2. Specify a JSON-LD frame to re-frame an arbitrary set of triples into a JSON data structure. 3. Select a set of pre-made JSON-LD examples and view the output. 4. Immediate feedback on the output of the JSON-LD processor. 5. Syntax highlighting for JSON-LD and TURTLE, making it easier to spot and debug errors. 6. Very minimal error handling - need to integrate a JSON linter at some point. Give it a shot and let us know what you think of it. All code for the site and the playground is available here: https://github.com/json-ld/json-ld.org -- manu -- Manu Sporny (skype: msporny, twitter: manusporny) President/CEO - Digital Bazaar, Inc. blog: PaySwarm Developer Tools and Demo Released http://digitalbazaar.com/2011/05/05/payswarm-sandbox/
Received on Sunday, 10 July 2011 17:56:09 UTC