Re: RDFa 1.1 Primer

Brilliant!

I have added a first set of comments, see below. They are not in priority order, just as it came to my mind while reading.

I understand the reasons you kept @xmlns and it is fine with me; using Facebook as an example is a good idea, so...

Two more general questions, though (not necessarily to you but to the group in general): 

- Shouldn't we have a Primer for the RDFa API, too?
- This is a primer to HTML authors. I wonder whether we should have a separate primer aimed at SW people? The emphasis would be different on a number of issues (I am in the process of finalizing a tutorial for ISWC...:-) 

Cheers

Ivan

Comments
--------

- Abstract: I hope what you say about Yahoo is still true...

- Abstract: the normative spec is indeed rdfa core, but that does not cover the html case of all your examples. Maybe it is worth making that clear and refer to the html document, too

- 1.1 HTML vs XHTML: maybe it is worth referring to the fact that rdfa can also be used with non-html, for example svg, even if all the examples here are in rdfa+html

- A general issue about facebook as a core example. Are we sure that the facebook engine really extracts all the rdfa information (as long as it is valid) and does not restrict itself to the <meta> elements? I would hope that is the case but we should make it sure, otherwise the primer may become very misleading...

- 2.3 Links with flavor: I wonder whether the usage of the rel="license" is good here. Up until now, we refer to full URIs via curies, and we also use that for the og:thumbnail property. But, suddenly, we refer to, essentially, a term with license, ie, hinting at the fact that we have yet another mechanism to generate those uris. This may be a bit disturbing to the user; it may be better to avoid that here and come back on the term issue later...

- same section, og:thumbnail example: I think we should emphasize that @src plays the same role as @resource in this case. It may be a bit confusing otherwise...

- the subject for the triples in 2.2 is ../alice/posts/trouble_with_bob, whereas, in 2.3, it is /alice/posts/42. I guess there is no realy reason to have different subjects here, it would be better to use one.

- for 3.1: I think there are two rdfa1.1 features that is worth emphasizing here: the fact that the user can use a full URI for a @rel or a @property if she wishes (eg, an example might be the usage of some esoteric property only once, and using the full uri might be easier than going through the @prefix declaration). The other is the fact that @about can use a curie, too...

I know you do not like the prefix declarations via @profiles:-) but I would still think that it is worth mentioning. Google expands the vocabularies it understands regularly and it would be a good idea if they simply added new prefixes to some sort of document they publish and people could then go and use foaf:, dc:, you-name-it:

- Section 5: I would prefer to refer to Turtle rather than N3. This is what we do in our documents and, well, N3 is (much) more than just a serialization format...

- A Acknowledgement: this should be changed to the RDFa WG... (see the other documents)

- I miss something on @resource (which I believe is very important) and chaining, ie, when object become subject. I believe these should be covered in the primer. Another question is whether we want to touch datatypes. I would think we should but, well...

 
On Sep 1, 2010, at 06:22 , Ben Adida wrote:

> 
> Hi folks,
> 
> It's been a while, but I think my distance helped me write a new draft of the Primer somewhat effectively. At least, I hope!
> 
> Here it is:
> 
>  http://www.w3.org/2010/02/rdfa/sources/rdfa-primer/
> 
> Some tweaks to diagrams and references are needed, but I think now's the right time for feedback from the group.
> 
> I realize that some in the group want to do away completely with @xmlns in the Primer. I tried that in an early draft, but I could never reconcile it with Facebook's current markup recommendation, which is probably the way that most users will be introduced to RDFa.... so bear with me on that approach long enough to read through the whole thing and tell me what you think!
> 
> -Ben
> 


----
Ivan Herman, W3C Semantic Web Activity Lead
Home: http://www.w3.org/People/Ivan/
mobile: +31-641044153
PGP Key: http://www.ivan-herman.net/pgpkey.html
FOAF: http://www.ivan-herman.net/foaf.rdf