Google isn't yet parsing RDFa. Dan Brickley just confirmed this for me, saying "it's being looked into plenty seriously. i can't promise any schedule (any more than the WG can) and of course there are various things parsing earlier rdfa 1.0 idioms".
Honestly, it upsets me that this has been made less than clear to users. I think that we should let the merits of the technology stand on their own and not be less than honest about what results people can see from them.
I feel bad that I have been inadvertently complicit in misleading users, who will spend hours trying to figure this technology out without realizing it won't give them the intended benefit. I would greatly appreciate it if you would make the real benefits clear.
Comments
Comment #1
scor commentedThe general statement "Google isn't yet parsing RDFa" is inaccurate. You very well Lin that Google has been parsing RDFa since 2009. The testing tool started to display Rich Snippet with the schema.org namespace in RDFa in September 2011 (maybe before, but that's as early as I noticed it). If you make a typo in the namespace it won't work, that shows it was intentional from their part to enable the schema.org namespace support.
You should have put Danbri's response into context. Your question was "is Google parsing RDFa at this point for Schema.org stuff? \cc danbri". I would need to check with Danbri, but I bet he understood this in the context of the new RDFa 1.1 Lite syntax that the schema.org sponsors agreed to support. This syntax recently went into last call and people are just starting to upgrade their parsers (including Google). I believe that's what Danbri was referring too.
I've added this note regarding syntax on the project page:
Comment #2
Anonymous (not verified) commentedI don't believe that Google is supporting RDFa for Schema.org at this point. When they announced Schema.org, they made it explicit that they would continue supporting RDFa for Data Vocabulary but had no plans to support RDFa for Schema.org. Since this module is specifically placing Schema.org terms, it doesn't matter that it works with the Data Vocabulary.
As you point out, you found that the testing tool supported RDFa in September 2011. As you also might note, I retweeted that tweet. In addition, you could note that Guha, the driving force behind Schema.org, said the following the next month:
And YOU know this very well. Because I have pointed it out to you before.
Now Dan Brickley says that they are working on it. In this, the full context, I take it to mean exactly what he said... that they are working on it. That doesn't mean that it is implemented or that it is even going to happen.
I find what you added to the project page to be even worse than the original, implicit implication that the module will produce Rich Snippets in Google.
Comment #3
Anonymous (not verified) commentedThe only way to know conclusively would be to have confirmation from Google's team. However, a report from an implementer using RDFa:
I've heard reports from implementers using microdata saying it takes between 2-3 weeks.
As I said above this is in no way conclusive. However, barring proof otherwise it provides weak confirmation for my understanding.
Comment #4
scor commentedI've just updated the project page with a few live search results powered by 100% schema.org+RDFa 1.0:
- Event: World Women's Curling Championships 2013
- Listing of events
- Recipe
I've also created a new section to explain the situation re. syntaxes (with a link to the microdata module:
I hope that clarifies some of the confusion which have spread in this thread.
Comment #5
Anonymous (not verified) commentedYes, I believe that is an accurate statement. Thank you for making sure the current status is clear to users.
It is also interesting that there are now examples of pages that do have live Rich Snippets based only on Schema.org placed with RDFa, this makes me feel much more comfortable about the module's promises.