[uf-discuss] class names and uniqueness - hAtom

Tantek Ç elik tantek at cs.stanford.edu
Tue Oct 25 20:54:20 PDT 2005

On 10/25/05 7:00 PM, "Danny Ayers" <danny.ayers at gmail.com> wrote:

> On 10/26/05, Dr. Ernie Prabhakar <drernie at opendarwin.org> wrote:
>>> I have mentioned this several times over the last several weeks,
>>> but read [2] _carefully_ to understand why a generic concept of
>>> "content" will _not_ work for marking up blog content.
> I couldn't find your relevant arguments at that link, but I do accept
> that a generic concept of content won't work.

Worse than won't work -- it's redundant.

>From the microformat perspective it is all *content*.

> But what is to stop the
> naming being like:
> <head profile="http://pwns.google.com/hatom">
> ...
> <body>
>   <div class="entry">
>         <h2 class="title">cat stuff</h2>
>          <p class="content">
>               <div>this is the business</div>
>           </p>
>         </h2>
>    </div>
> ...
> </body>
> The profile URI disambiguates, the hAtom profile can refer to the Atom
> spec normatively, why do the attributes need to carry something more?

They don't, BUT, for "web designer" friendliness, it is good practice to
pick a root class name for the microformat which is relatively unique.

For more on this practice, see:


> However I don't think it's worth losing sleep over (which I literally
> am doing at this point in time ;-) as long as the attribute names used
> within the doc are likely to be unique,

Only the root class name(s) really need to be unique.

> it doesn't matter what they
> are. Use Tantek's unicode wibbly (cedilla?) thing if it looks nice. It
> might even be beneficial to do a little namedropping,,,

Funny yes, but that wouldn't follow the existing patterns.

atomfeed and atomentry seem analogous to vcalendar and vevent (or even

Thus for those, I can see it making sense to keep the "atom" prefix in the
class name.  However, for any classes contained therein, there is no need to
prefix them.

And as for making things more generic, I'd say, at first, don't.

Stick with solving a specific problem.

Which in this case is making the atom/feed semantics explicit in the
semantic (X)HTML of a blog.



More information about the microformats-discuss mailing list