[Sigia-l] Web patterns & web semantics

[Listera]

Donna Maurer:

> John Allsop (a very smart Australian web standards dude) has written
> an article...

...which would greatly benefit from separation into at least two parts: the
notion patterns (for those new to it) and semantics (in the context of HTML
markup). And I wish it would declare unambiguously just what the
problem/issue/objective is.

In the absence of that, it intermingles a lot of *different* and not
necessarily related matters, such as markup naming conventions, web
navigation, tagging, algorithmic parsing, etc. Nothing would "solve" all
those problems simultaneously. :-)

People have been advocating common usage of nomenclature for, well,
everything, since 345 B.C. In 20 years of coding I have never met two
programmers who agree on how to name their variables, even when using the
same language. One way or another, structuring and "standardizing"
information has been a chimera since the dawn of the web. So color me
skeptical.

But different "solutions" have also emerged in this regard. Flickr,
del.icio.us, etc popularized the notion of loosely related tagging. Google
and now Microsoft just introduced the notion of structured data entry. And
if you notice, these are not entirely page-based. They are more granular and
targeted at the object/field/entity level. Even if your semantic definition
of a Google Base entity has not been blessed by a Pattern Standardization
Authority, I can still query/parse/relate it because it's well-structured
and mostly self-describing.

In short, while thinking in terms of structure is extremely important, I
expect solutions to come from different directions.

----
Ziya

"Innovate as a last resort."