- From: Robin Berjon <robin@berjon.com>
- Date: Fri, 18 Mar 2016 12:45:37 -0400
- To: Ivan Herman <ivan@w3.org>
- Cc: Gareth Oakes <goakes@gpsl.co>, Peter Murray-Rust <pm286@cam.ac.uk>, W3C Scholarly HTML CG <public-scholarlyhtml@w3.org>
On 18/03/2016 12:36 , Ivan Herman wrote: >> On 18 Mar 2016, at 17:10, Robin Berjon <robin@berjon.com> wrote: A >> first step for this would be to use the HTML validator. It's not >> constraining enough, but it would certainly already catch issues. >> The only problem with that is that it only supports RDFa Lite, >> whereas SH uses a little bit of RDFa Full. Ideally, I'd like to get >> rid of that and rely only on Lite — input on achieving this >> welcome. > > @darobin, can you point out which are the features that are not RDFa > Lite? I am happy to look at it (though I cannot promise that > something can be found) There is at least one case in which we use @about (for roleAffiliation). I think we could probably use @resource instead. When more information is available than ought to be *shown* (but it still should be shared), we recommend meta@content. We could use spans with CSS display: none but that's a bit dodgy, or we could have JSON-LD islands (but that gets verbose and increasingly weird). We also use @datatype to mark how precise dates are (it is common to only have a year). I guess that could be skipped since it's easily inferred from content. But we also use it to mark content as rdf:HTML — removing that capability would be more problematic. Actually, writing all this down I think it would be simpler to fix the HTML validator :) -- • Robin Berjon - http://berjon.com/ - @robinberjon • http://science.ai/ — intelligent science publishing •
Received on Friday, 18 March 2016 16:46:01 UTC