[microformats-discuss] [admin] before posting new pages to the wiki...

Tantek Ç elik tantek at cs.stanford.edu
Fri Aug 26 11:56:52 PDT 2005

On 8/26/05 8:54 AM, "mark rickerby" <coretxt at gmail.com> wrote:

> Regarding the content-id page, I admit that I was trying to be
> provocative, the idea wasn't so much about solving a problem, as it
> was documenting widespread current behaviour and questioning what
> actually constitutes a design pattern in HTML.

Documenting current patterns of semantic XHTML usage is certainly a good
thing, whether or not we want to make them full fledged microformats depends
on a number of other factors, but the fact that pattern exists is a strong
starting point.

The *key* to documenting such patterns is by listing examples *explicitly*
including hyperlinking to the actual examples on the web.  Without such
actual examples that anyone can verify, the examples/patterns are purely

> I think it's unclear
> where elemental formats end and standard HTML starts,

Or perhaps the opposite, where usage of purely standard semantic XHTML ends,
and elemental formats begin.

To put it simply, as soon as you begin to capture semantics in your markup
using terms (and semantics!) which are not in the XHTML specification, then
you have taken the first steps towards using a microformat.

E.g. by using semantic class names, ids, or a rel/rev value (other than what
is already defined in the "link types" list in HTML4.01), you begin
experimenting with the key pieces of a microformat.

Such "experimentation" or even "design" is not new.

Modern web designers have for some time been using semantic class names, ids

Thus microformats itself is not really an invention.

It's more of a recognition of that broader pattern of web content design and
authoring itself.

Now some folks who don't know semantic XHTML very well write things like:

 <div class="heading">

instead of simply using <h1>, <h2>, <h3> etc.

I've even seen:

 <div class="para">

instead of <p>,

or even:

<div class="list">
  <div class="item">


Those are not microformats, because they are representing semantics that are
already built into XHTML.  Those are just reinventing the wheel.

Also note that folks that invent a <list> tag or an <item> tag when <ul> and
<li> would do are doing no better than those that would use <div
class="list"> or <div class="item">.

Such unnecessary reinvention is one of the things we seek to avoid wasting
time with in microformats.

One could even say <div id="content"> is semantically redundant with <body>
for example.

> and the idea of
> documenting this "design pattern" raises this in an interesting way,

Existing similar markup is not necessarily a design pattern.  "design
pattern" is a loaded phrase that implies good design.  The sad thing is,
that due to much poor markup (most of it output by poorly constructed CMS
systems), much existing markup on the web is not even close to being a
design pattern.

Thus, if you see existing similar markup, bring it up on the list and see
what people think of it.  Whether it is a pattern or perhaps an

> showing how the intuitive reactions of designers over the past 3-4
> years of CSS development has led to certain commonalities in
> structure.

That certainly has taken place.  Some of those commonalities are patterns,
and some are workarounds for browser bugs and limitations.  Thus it is
useful to examine/deconstruct them.

> Maybe it belongs in a more design focused context, I'm happy to remove
> it, my apologies for misinterpreting the purpose of the wiki pages.

Or perhaps such examples belong in a more *general* context of
web-markup-patterns where we can then discuss and maybe separate them into
design patterns and anti-patterns, and provide the reasoning for each so
other folks can discern between examples of the two.



More information about the microformats-discuss mailing list