Re: Microdata design philosophies

On Thu, Oct 15, 2009 at 4:33 PM, Martin McEvoy <> wrote:
> Itemscope, itemprop, itemref...etc..etc don't sound human friendly at all,
> If microdata were ever to become popular ( god forbid ) think of all
> gigabaytes of data that's going to take up just because you thought that
> pre-pending everything with "item" was somehow intuitive, and TBL worries
> about forward slashes, Im suprised you can sleep at night ;)

You kidding?  Prepending everything is *awesome*, because it
immediately calls out that this is a group of related attributes.
Since they're all global attributes, it's useful to have an immediate
visual clue about what you're dealing with.  It also lets you avoid
worrying about name collisions with existing attributes.

> Why use microdata at all when microformats already exist,  I cant think of a
> single microdata use case that isn't already solved by microformats and
> RDFa,

Huh?  There are tons of cases not addressed by microformats (at least,
not addressed easily).  That's the entire reason stuff like RDFa

> by the way microformats and RDFa work very niceley together on the
> same page, I would hate to try and mix micro data with microformats or RDFa

Again, huh?  RDFa and Microdata both mix exactly equally well with
microformats.  Both of them are nothing more than a set of attributes
(+ 2 elements for Microdata, but neither of them are visible so they
don't interact with Microformats at all).  Microformats just uses
@class, @title, and the contents of elements, which shouldn't
interfere with either RDFa or Microdata.

I'm pretty sure you can even mix RDFa and Microdata together, since
Microdata mints entirely new attributes for everything it does.  I
don't believe there would be any weird collisions.

> I have said this before I still don't think you will be landing any planes
> with microdata sorry.

I would hope that no one attempts to land a plane with data embedded
in an html page!  ^_^


Received on Thursday, 15 October 2009 21:46:35 UTC