[microformats-discuss] Re: Proposing RelSource

[Andy Skelton]

Getting back to the rel="source" or rel="cite" idea: Ryan and I
discussed it off-list and I sorted out a lot of my own half-baked
thoughts. Posted here for interest.


On 7/13/05, Ryan King <ryan at theryanking.com> wrote:
> And to take things  a step back, I don't think your proposal is a bad
> one- in fact its very good. I just think there may be some other good
> ideas around this problem and we need to explore all of them. Also,
> if you think <cite><a></a></cite> is not the best idea, I want to
> hear your reason. I'm not defensive about it by any means, but I want
> dialogue about microformats to be full and reasoned-out. I hope you
> understand.

The rel="cite" idea was easy to generate but it's difficult to defend.
I needed to step back and think about what problem I was trying to
solve. When I did that, I saw rel="cite" becoming less and less
attractive. Several times during the process, I wanted to give up and
remove myself from the discussion but I think I'll eventually come up
with something worthwhile. I'll try to work out my thoughts in the
next several paragraphs. I hope some good comes from all this...

My rel="source" idea was born out of some observations I have made
while watching the blogsoftheday.com lists evolve and react to
breaking news and surging memes. The most significant cases to come to
my attention in the past week have been the London bombing and the
Longhorn screenshots. What I saw were primary sources at first,
followed by many bare gateway articles (just the link and maybe a few
words) pointing toward primary sources and a lesser number of articles
offering the same link but with some additional content such as
editorial/commentary, supporting/refuting evidence, or additional,
related links.

When I'm looking for topical information the masses of bare gateway
articles are like so many iron filings when what I want is the magnet
at the center. The filings have little or no intrinsic force or value
but they do serve an excellent purpose: their numbers help to indicate
the strength of the magnets. By counting the filings, machines like
Technorati can measure the strength of the magnets. To a human,
however, they are generally not worth the time it takes a UA to load
the page.

The way I see it, the division between bare gateways and other kinds
of resources is the division between mere dissemination and actual
participation. That is an important distinction to make because it is
also the difference between "interesting to machines" and "interesting
to humans."

There is a big monkey wrench in here: the proliferation of blogs and
their posts as forums for discussion. What might otherwise be a mere
gateway (a one-line blog entry linking to a more substantial article
elsewhere) now has the /potential/ to house a rich discussion.

While we're throwing monkey wrenches around, consider the author's
motives for writing and hosting these one-line gateway posts. Aside
from being just plain helpful, these authors are usually trying to
snag a few passers-by with ads or other content. Take for example one
of the biggest gateway blogs ever: Slashdot. All they do is publish
links (provided by readers even!), add some snappy comments and give
readers a place to discuss.

What's the lesson here? Don't count on authors to spend their time
telling people "don't visit my site, it's just a gateway." They won't
do it.

Now that I've settled my own selfish desire to make the chaff kindly
separate itself from the wheat, let me explore another aspect of the
problem: if each and every one of these interlinked resources is
potentially interesting, can we at least classify the links such that
we can map cross-domain discussions in a useful manner?

Now we're back into the exact territory you were exploring a couple of
months ago with hVia. I'll end and send this email and organize my
thoughts for the next one after lunch.

Andy