[microformats-discuss] comments please: draft-sayre-atompub-xhtml-micro-00

[Tantek Ç elik]

Hi Robert,

On 9/3/05 4:57 PM, "Robert Sayre" <sayrer at gmail.com> wrote:

> [this message has also been posted to atom-protocol at imc.org, but that
> list is not CCed here]
> 
> Here are a couple thoughts on Atom category editing and Atom
> Publishing Protocol error documents. Also, these thoughts come with
> sample client and sample server implementations written in Python
> (appendices A and B).

Wow!  That's a lot of work.


> -------------------
> (I submitted it, but it's not in the I-D directory yet)
> http://www.franklinmint.fm/2005/09/03/draft-sayre-atompub-xhtml-micro-00.html
> http://www.franklinmint.fm/2005/09/03/draft-sayre-atompub-xhtml-micro-00.txt
> 
> Atom Publishing Protocol [APP] client implementations require a fair
> amount of ancillary server-provided data in order to provide a smooth
> user experience. Rather than invent a plethora of new XML formats,
> this specification chooses to present a number of XHTML profiles
> [XHTML], colloquially known as "microformats". Visit
> <http://microformats.org> for more information.
> -------------------

Makes sense.


> Please consider this a discussion
> (<http://microformats.org/wiki/process>). I had a random thought and
> decided to see if it would work, but the ideas turned into a spec and
> a couple of implementations before I knew it. I'm not attached to the
> actual XHTML markup,

To be fair, as has been pointed out to other folks on this list, it's better
to publish actual *examples*, known previous *formats*, and even raw
*brainstorming* ideas before jumping to a specification, and specific XHTML
markup.

> but I think the basic capabilities are pretty
> solid. Basically, the format builds on Atom's category element and the
> metaWeblog API's getCategories facilities.
> 
> Comments, changes, process suggestions, and anything else welcome. I'd
> be happy to put this on the wiki if that's appropriate.

Robert, I took a quick look at the document, and perhaps I'm missing
something.

The whole "category" concept as described appears overkill, and not
representative of current content/behavior on the web.

Do you expect people this complex category construct instead of simpler
things that already work and are broadly interoperably implemented, e.g.
rel-tag?

 http://microformats.org/wiki/rel-tag

Thanks,

Tantek