On Simplicity (POSH)
This year will be the year of the semantic web for the common man, in the first place because of the success of microformats. And because of the original semantic web people, who in general have a very positive attitude towards the grassroots adoption, helping a hand with tools and specifications like GRDDL.
This whole movement has led to another change as well. That is, as I see it, people getting to finally understand what the meaning of the original HTML specification was all about: it is the semantics, not the presentation in the first place.
In short, that specification meant to say that a <h1> element should convey the meaning of a top/document title, followed hierarchically by <h2> for chapter titles, <h3> for subtitles and <h4> for section titles and so on. Thus a <p> marked a paragraph, and more interesting, things like <cite> would mark a literal citation. This is called semantic markup.
So far so good, the resulting presentation in the browsers of those days was very bland and we needed a more sexy look for marketing purposes. Thus the markup was pushed towards the presentational limits – presentational markup indeed, where elements were used for their look, rather than their semantic meaning. A <blockquote> element no longer meant some sort of citation, but rather an indented paragraph of text. HTML itsel suffered from presentational elements as well: the <b>old tag and the <i>talics define presentation. For meaning, we have <strong> (rendered as bold type) and <em>phasis (rendered in italics). Quite a mess.
OK, so right now, the POSH movement has come to life. Where POSH stands for Plain Old Semantic Html. It all originated from within the Microformats movement, but discussions on the mailing list indicate that most microformat advocates think of POSH as being not related to microformats in general.
And I agree. Sure, good to see semantic correct use of html on the web, but this approach is way too simple in general to extract real meaning. If the web only existed of sceintific papers and novels, then maybe we could rely on the stricture of html as originally specified. And write a GRDDL transformation, which then says as much as: this <blockquote> is a citation of something else. Of what exactly, would still be left up for guessing.
All in all, using POSH wherever you can is good practice. But this has almost nothing to do with real semantic web. And could someone please help me to explain what the semantic meaning of <strong> versus <em>phasis is?
To finish with a famous qoute:
[Solutions should be] as simple as possible, but no simpler.
— Albert Einstein.