[uf-discuss] Process to handle decentralized creation of new microformats?

[Tantek Ç elik]

On 10/1/06 7:44 AM, "Costello, Roger L." <costello at mitre.org> wrote:

> Hi Folks,
> 
> This community is awesome!  Microformats are awesome!

Thanks much Roger!


> At the present time the creation of new microformats is centralized, to
> this list.

Or certainly to microformats.org.  Many folks prefer different communication
channels so sometimes more is discussed on IRC than the mailing list for
example.  It is possible (though I haven't seen it happen yet), that a
microformat could be created simply through use of IRC, wiki and process,
but unlikely, as it is more likely that one or more participants will email
the list at some point, at least to announce the interest in the microformat
and request help with research.


> And this community has an excellent process established for
> creating microformats.

Thanks Roger.  It has certainly been a bit of a learning experience, and
improved along the way thanks to the participation of the community.


> What happens when creating microformats becomes decentralized;

IMHO, this akin to asking what happens when Wikipedia becomes decentralized?


> that is,
> when communities go off and create their own new microformats?

What Scott said:

> On 10/1/06 10:56 AM, "Scott Reynen" <scott at randomchaos.com> wrote:
> 
>> I'd guess what happens is it doesn't work.  By analogy, what happens
>> when communities go off and create their own dictionaries?  I'm not
>> sure, but I suspect they end up with a dictionary that's not very
>> descriptive of how people actually use language.

I would expect that if other communities did go off an specify their own
semantic class names, even with profiles, that it might get some uptake, but
not much.

I expect that *this* community would bring any such effort to the attention
of the list and then either there is enough interest / universal / practical
applicability (80) or there isn't (20).  If there is, anyone here could walk
an effort through the microformats process.


  
> For example, one community might create a microformat for aquarium
> tanks.  The "tank" microformat might be used in something like this:
> 
>    <div class="tank">
>        ...
>    </div>

Unlikely.  What is more likely is that such a community could simply use a
generic "product" microformat.  There have already been quite a few
discussions here about the research towards creating a product microformat.


> A separate, independent community might create a microformat for
> artillery tanks.  Their "tank" microformat might be used in something
> like this: 
> 
>    <div class="tank">
>        ...
>    </div>

Again, unlikely.

Unfortunately this is is a theoretical example, and as such, would not even
be worth the effort to document research toward creation as a microformat.

One of the reason the practical and research requirements/steps are in the
process are precisely to avoid this kind of theoretical proposal or even
discussion.


> In general, as microformat creation becomes decentralized

We can worry about it when Wikipedia becomes decentralized.

Scott answered this as well.

In addition, profiles address this for folks that want to create their own
semantic XHTML class names and provide specific well defined semantics.


> This "microformat name collision" problem might lead to the suggestion
> to use namespaces for microformats.

Already suggested and rejected.  Profiles solve this problem without
namespaces.

 http://gmpg.org/xmdp/


> But using namespaces in microformats will trigger problems in other
> technologies, such as CSS.

Actually - you just explained why namespaces have failed on the Web.


> There are lots of problems lurking with decentralized creation of
> microformats.

Which is why it is unlikely to happen, thus we should not be worried about
it.


> On the other hand, centralized control of the creation of microformats
> doesn't seem practical on a massively decentralized Web.

A centralized Wikipedia seems to be working reasonably well.



> As
> microformats become increasingly popular people will go off and create
> their own microformats (it's already happening).

They are creating their own sets of semantic class names and incorrectly
labeling them as microformats.

They are not creating microformats.


> Centralized control of the creation of microformats is akin to the XML
> community mandating that all XML tags be created by a central
> clearinghouse.

XML was essentially designed for babel - for decentralized, re-invented,
creation of tags which would result in an inevitable proliferation of
non-interoperable formats, and that is exactly what has happened with *very*
few exceptions that you can count on one cartoon hand (XHTML, RSS, Atom).


> There are lots of really smart people on this list.  There must be a
> solution.

Or, perhaps there is no problem.

IMHO Roger, this is unlikely enough to be a problem that we can ignore this
possibility and focus on immediate practical problems instead.

It is a matter of prioritization.


> Can you think of a process for allowing a decentralized
> creation of microformats, without the ensuing chaos alluded to above?

Given that decentralized microformats would cause problems, why would we
bother spending any time enabling them by creating a process for them?

Tantek