- From: Nicolas Torzec <torzecn@yahoo-inc.com>
- Date: Thu, 29 Jun 2017 22:38:41 +0000 (UTC)
- To: Christopher Allan Webber <cwebber@dustycloud.org>, <public-schemaorg@w3.org>
- Message-ID: <1580049434.1800404.1498775921054@mail.yahoo.com>
Hi, You can test Google's parser output via its Structured Data Testing Tool. It looks like Google recognizes the type schema:Person but not the type customfoo:Citizen, as expected. Bing has a similar tool here: https://www.bing.com/toolbox/markup-validator Cheers.-Nicolas. On Thursday, June 29, 2017, 12:20:13 PM PDT, Christopher Allan Webber <cwebber@dustycloud.org> wrote: Hi there, I'm doing some data modeling for a client that would like to ensure that their content is discoverable via Google / Bing / etc. We're using schema.org everywhere we can, but there are some custom terms. Say we have a Person object posted in a context like: {"@context": {"schema": "https://schema.org/", "customfoo": "https://foo.example/ns#"} "@type": ["schema:Person", "customfoo:Citizen"], "schema:name": "Gertrude McBell", ...} This should technically be valid data modeling as far as I can tell, and it should also convey that this Person is also a "Citizen" under their custom vocabulary. (This is not the actual vocabulary we're using, I'm just making up an example.) Composite types are allowed in the linked data world in general... but whether they are recognized by common systems is another thing. Google / Bing / etc probably have no idea what a Citizen is here, even though it is important to our data, and that's fine. But will they still be able to recognize the presence of Person with the type as an array? If not, maybe we should stick to just a singular type, as schema:Person. Thanks in advance! - Christopher Allan Webber
Received on Thursday, 29 June 2017 22:39:16 UTC