[microformats-discuss] Re: Educationg Others

[Danny Ayers]

On 10/5/05, Chris Hibbert <chris at commerce.net> wrote:
> > Ok, the long tail idea is not just about the case of 3 people who post
> > a picture of their albino cat jumping over a fence while wearing a
> > sombrero, versus the 10,000 people doing the same. It is about those
> > first 3 people, plus the 4 fence-jumping fedora-wearing albino cats,
> > plus the 2 fence-jumping fedora-wearing black-and-white cats...the
> > aggregated total of all of which may be well into the millions.
>
> Amazon, Netflix, and eBay didn't make it big because they had separate
> categories and formats for each of the tiny niches in books, movies, and
> p2p sales.  They each supported a few categories, and lots of ability to
> describe differences, and allowed their customers to search the
> descriptions.  Flickr may want to allow arbitrary captions on the
> pictures, but it's unlikely they need a structured markup language for
> cats that includes what they're wearing, or what their color or breed
> is.  Standard simple markup with room for differentiation in the
> descriptions is sufficient, and I think that's what microformats push for.
>
> Amazon wants to display titles, authors, cover photos and reviews.
> Netflix wants to let people browse by genre, title, actors, and
> directors.  Neither needs to come up with genre-specific markup to serve
> all the tiny niches the companies never thought of before they became
> part of the phenomenon.

Yep, sure. But I was thinking of microformats more from the point of
view as an approach to creating markup specifications, more of a
meta-level. If the process systematically excludes edge cases, then
you're looking at something that might expand one particular 1% to
1000% (say Amazon's use case), but arbitrarily leave another 1% at 1%
(say Netflix's) because it doesn't mesh with the Intelligent Design of
Mark's 80%.

I honestly don't know whether the long tail kind of analysis works at
the format level. "XML has failed on the Web" because the leading
formats are HTML or (often broken from an XML pov) RSS. There are
numerous successful XML formats in use on the Web, but I doubt their
aggregated document count outnumbers that of HTML.

Cheers,
Danny.

--

http://dannyayers.com