Change microdata chapter so it points to schema.org instead of data-vocabulary.org? about diveintohtml5 HOT 3 CLOSED

It is a consideration to put it into the text, but http://manu.sporny.org/2011/false-choice/ brings me pause.

from diveintohtml5.

@jonathantneal that post of Manu's was written a year ago. Since then there have been a number of more positive developments. The RDFa WG took on the challenge of making RDFa as easy to use as Microdata, and I think delivered there: the RDFa Lite view of RDFa 1.1 is, from a publishers perspective, pretty much equivalent to Microdata in complexity. It's a bit more work for parser writers because you need to parse all of the RDFa 1.1 syntax, but as there are many many more publishers than parser writers, I think that's a reasonable design. Secondly at semtechsf last month in the schema.org panel we announced - see http://blog.schema.org/2012/06/semtech-rdfa-microdata-and-more.html - enthusiasm for RDFa Lite. And we're starting to see search engines offer support for it. So a lot of the issues from last year are now historical...

from diveintohtml5.

To resolve this issue, I asked Tantek Çelik for some advice. This is what I understand.

The chapter is simple and straightforward, whereas schema.org is a confusing mess. Compare the sample PERSON VOCABULARY, which is short, based mostly on hCard with a little XFN thrown in there, with http://schema.org/Person.

At least with data-vocabulary.org, Google stuck with a few small vocabularies that it actually supports. Google seems to keep reinventing how they want to do their custom version of semantics every 2-3 years. Perhaps this issue contains its own demise: "seems to be the shiny new common thing".

The best "future-proofing" for the microdata chapter might be to add parallel microformats markup examples to each microdata example.

I'll look into adding the microformats markup, but I'm going to close this particular issue. If you feel that this judgement is off, feel very free to comment or reopen the issue.

from diveintohtml5.