Not all semantic HTML is a microformat (Was: [uf-discuss] [Zen of Microformats] Two Fundamental Principles of Information Design)

[Scott Reynen]

On Mar 8, 2007, at 2:26 PM, Ian Davis wrote:

> Hi Ara,
>
> On 08/03/2007 19:43, Ara Pehlivanian wrote:
>> What I think your contrast of microformats' bottleneck with the web's
>> free growth is missing is the notion that there indeed /was/ a
>> "bottleneck" in the development of the web in the form of Tim
>> Berners-Lee. It's just that it was all up front and not distributed
>> throughout its growth. When he invented HTML, he did the same kind of
>> work we're trying to do with microformats, but he did it by himself
>> and presented it to the community as a fait-accompli.
>
> Maybe I'm misunderstanding but you appear to be saying that  
> creation of each microformat is equivalent to creating HTML, i.e. a  
> set of useful elements people can use for creative expression.  
> Under the microformats process that effort needs to be centralised  
> to prevent name collision, semantic drift and duplication of  
> effort. There's no way to allow parallel development and then work  
> out how to coordinate differences later.
>
> This means there's a limit to the scalability of the mf development  
> process.

Yes, there is.  I think that's made pretty clear on the about page:

> microformats are not:
> ...
> - infinitely extensible and open-ended
> ...
> - a panacea for all taxonomies, ontologies, and other such  
> abstractions
> - defining the whole world, or even just boiling the ocean

http://microformats.org/about/

> I think the agreements can occur in parallel between interested and  
> motivated parties.

They can, and they do.  But they don't often follow the microformats  
process.

> But that's not the impression I get of the microformats process.

That's because such smaller-scale agreements fall outside the  
microformats process.  It should be self-evident that two people can  
talk to each other and agree on what something means.  It baffles me  
that we would need to explicitly state that.  But since it's such a  
common misunderstanding that we somehow control other people's HTML,  
it would probably be good to make that clearer where we can.

> It isn't a microformat until it's blessed by the mf community.

No, it isn't a microformat until it follows the process.  The  
microformats community just happens to be the largest community  
currently dedicated to following the microformats process.  There's  
nothing preventing anyone else from following it in another  
community, but that doesn't seem to happen much, I expect because  
it's easier to just use this community.

> Here's a great microformat that's been developed outside the mf  
> community:
>
> http://selfdescription.org/

I'd say that's semantic HTML, not a microformat, as it doesn't appear  
to have followed the microformat process.  That doesn't mean it won't  
be useful, but calling any semantic HTML a "microformat" is like  
calling all feeds "RSS" or calling all JavaScript "AJAX" (both  
mistakes I see often).  It's not a big deal, but it's confusing to  
people who don't understand the difference, and if enough people do  
it without anyone correcting them, the terms will lose meaning in  
common use.  I often find I need to clarify if someone talking about  
"RSS" is really talking about feeds in general, including Atom.  I'd  
rather not feel the same need to clarify what people mean when they  
say "microformat."

> Has anyone here heard of it, evaluated it or have an opinion on it?  
> Its development followed the mf principles with an extensive survey  
> of existing practice and a distillation of that into a set of  
> common and reasonably humanly accessible terms.

That's the semantic HTML aspect of the microformats process, but  
there's more to it than that.  For example, early in the process is  
this sentence:

> We want to involve all interested parties in the discussion.

http://microformats.org/wiki/process

That doesn't seem to have been followed in selfdescription.org.  It  
looks like the work of only one person.  Also, he seems to be  
repeating some of the semantics already available in HTML (e.g. the  
idea of nodes as containers).  Again, none of this prevents the  
concept from being useful semantic HTML, but I think it's a mistake  
to call it a microformat.

> I provided some mentorship to the author as part of his Masters (of  
> which this formed the thesis)

You might encourage him to bring his work to a larger community (this  
or any other).

Peace,
Scott