Re: ISSUE 248 - Microformats

Thanks Manrique!

Overall, I completely agree with the usefulness of having more semantics 
in general, and in the mobile context in particular.

I don't think that there is any existing best practice we could 
reference, but given that we've extended the definition of best practice 
to include something that we think will become a best practice in the 
very near future, there could still be some room for that to happen. I'm 
just skeptical though that any of the applications you list matches the 
"very" in "very near future". I'd be extremely happy if I turn out to be 
too pessimistic about that!

See further comments below.
Francois.


Manrique Lopez wrote:
> When I read about microformats, I hope we are also talking about adding 
> semantic content to the markup (semantic annotation?), using 
> technologies like RDFa[1], GRDDL[2] or even SKOS[3] i.e.

That is also my understanding of the discussion. Microformats and more 
generally speaking semantics included in the markup.

> 
> At CTIC we have done research on this topic, and we have found several 
> scenarios where semantic annotated content could be helpful for mobile 
> web context. For example:
> * Semantic annotation for people, geo-position, etc. data could be 
> useful for contextual searching, so browsers could show only the 
> information that would be relevant for the user, or even mobile apps 
> could extract information from the web (dates, telephone numbers, etc.).

I don't think there is any support for microformats in current mobile 
browsers:
- The mobile version of Firefox might have some support but it does not 
really exist yet
- Safari may have some support in the future but not sure when
... but I haven't conducted any extensive research on the subject. Am I 
wrong?

Semantic data could also be used by search engines to index the pages 
more efficiently and present only the results that match the user's 
context, which could be particularly useful in the mobile context (hey, 
isn't that one of the use cases of the mobileOK labelling in POWDER?). 
But again, I don't think search engines in general use this kind of 
information for the time being.

Extraction of information by mobile apps is a good idea. It's not 
completely clear what specificity the mobile context brings here. Would 
you have an example? I guess that's the main lead I would follow to 
include some text on microformats in MWABP.


> * Automatic form filling based in semantic annotation of form inputs.

You mean filling in the first name, last name, email, and other 
recurring info automatically, right?
That would be fantastic, and clearly has a specific resonance on a 
device with limited input. But same thing here, it does not exist yet...


> * Content transformation tools could be helped by semantic annotated 
> content, making web browsing more contextual.

Right. POWDER is typically considered to do such things. But we'll 
mention this in the Content Transformation Guidelines is the "Scope for 
future work" appendix.


Francois.

Received on Friday, 4 July 2008 14:40:45 UTC