[uf-discuss] RDFa Basics video (8 minutes)

[Manu Sporny]

Guillaume Lebleu wrote:
> Manu Sporny wrote:
>> Constructive feedback would be great, as I'll probably be doing the
>> "advanced" RDFa tutorial in a month or so, and will need to know what
>> worked and what didn't in the RDFa Basics video.
>>   
> I'm relatively new to RDFa and this is a great introduction. I'm
> probably going to say things you've already heard, so please bear with me.

Any feedback is really appreciated, so I thank you for taking the time
to write out your thoughts on the video[1] :)

> So, when comparing uf and RDFa, adopting RDFa seems a
> huge leap of faith and a lot of work considering it requires XHTML 2.0

Ooops, looks like my post lead to some confusion. RDFa works with any
version of XHTML. You can write valid RDFa today. t's not an "it's going
to happen in the future" thing - you can author RDFa today (for example,
Operator already has RDFa support).

> and your introduction does not mention any application supporting RDFa...

That's true, there aren't many applications that use RDFa today...
mostly because the standard is just being finalized. The semantic web
stuff has been brewing in academia for almost a decade. We have the
Microformats community to thank for getting the message about why
semantics are important in web pages... so, I think that we're really at
the cusp of this stuff taking off.

> So, if the goal of this video is to evangelize RDFa to a large audience,

Hmm... the goal of the video wasn't necessarily to evangelize RDFa - it
was to educate web developers about some basic semantic web concepts. It
was supposed to teach people very briefly about the power of RDF, what a
CURIE is, what N3 notation is, and how you can express semantic
information in XHTML web pages using RDFa. Do you think it achieved that
goal, or did it fall short for some reason? Was there too much going on
in each part of the video?

> I think it would be great to explain why (ex. why id, class and
> a/href/rel haven't been leveraged more to represent RDF triples)

[FULL DISCLOSURE: I am on the W3C RDFa task force... take what I have ]
[                 to say with a grain of salt.                        ]

I'll explain here, just so you know. I don't know how many people know
this, but XHTML is extensible in that you can add modules to extend
XHTML. These modules define new attributes and tags that can be used in
a valid XHTML document. RDFa is an extension to XHTML, and works with
all current browsers that support XHTML.

@id isn't used because it must be unique to a page. It also already had
a different set of semantics associated with it in XHTML and it doesn't
allow CURIEs, AFAIK. @about is used instead.

@class wasn't used because they didn't want to stomp on the Microformats
community's implementation, among other reasons. In certain RDFa
implementations, bad things happened when you mixed RDFa and
Microformats on the same page. @about is used instead.

@href and @rel were re-used. @href is used to denote a object, while
@rel is used to denote predicates.

The current implementation makes sure to take the semantics of the
pre-existing XHTML standard into account, the following attributes were
used from XHTML:

@rel, @rev, @href, @src

These attributes were added in XHTML+RDFa:

@about, @instanceof, @resource, @property, @datatype, @content

I should mention that while RDFa may seem more complex than Microformats
(and in some cases it is)... there are no issues with ABBR or @title
mis-use/abuse. These endless debates we seem to have over "where do we
place the ISO machine readable data in the Microformat?", are a
non-issue in RDFa.

> also explain how an implementation can best leverage the
> backward-compatibility/evolutionary benefits of microformats with the
> RDFa more formalist and consistency with other w3 standards.

Thanks for the suggestion - I'll try to put that in the next video, or a
video comparing/contrasting Microformats and RDFa. We need both... like
any syntax, each has its benefits and drawbacks.

> Last, if you believe like me that applications drive standards, not the
> reverse, then I think what will ultimately get people excited is to have
> a demonstration of how this content can be leveraged by an application
> once published and gathered in a RDF store. 

You're right, the proof of the greatness of any technology is in it's
adoption rate and the number of applications that are written using that
technology.

> if I had to evangelize RDFa, I would
> start with the application, for instance showing content in a browser,
> then showing how it's possible to type queries against this content in a
> browser plugin. Then only I would show the implementation, and at the
> end possibly, adoption numbers such as how many people have downloaded
> the plugin so far.

We hope to do just that this year... using the hAudio Microformat as
well as hAUdio RDFa. I guess we have to start somewhere, and this video
was an attempt at just that...

Thanks a bunch for your feedback, Guillaume - it is definitely
appreciated :)

-- manu

[1] http://www.youtube.com/watch?v=ldl0m-5zLz4