[uf-discuss] Microformats collection and VoteFor

[Sven Fuchs]

I am currently playing around with an application for collecting VoteFor 
links for some kind of "distributed votings" I envision. This is my 
first experience with a microformat-targeted app and it's definitely fun.

I've chosen the FIFA worldcup for a first test voting and (not really in 
time) managed to put it online: [1]. Also I've blogged about some 
details [2]. It's really in an experimental state currently so I don't 
know if everythings working as expected right now - but please feel free 
to try it out if you like. :)

[1] http://folksr.de
[2] 
http://www.artweb-design.de/articles/2006/06/05/distributed-votings-using-microformats


Two questions.

I do know of the following approaches/practices regarding the collection 
of microformats:

1) spiders (requires actively searching the web, expensive)
2) track/pingback (requires usage of specially crafted publishing tools)
3) a form to manually enter the URL (requires additional user action)

Ok, of course there's additionally:

3.a) ping-service that allows to distribute pings for pages that aren't 
  capable to do it themselves.

But this requires extra specialized knowledge that users probably don't 
have and that could make up an additional hurdle IMO.

I'm currently wondering if there's another reliable approach like:

4.) a transparent pixel (called a "webbug" in wikipedia [3]) in 
combination with a MicroID [4]

[3] http://en.wikipedia.org/wiki/Webbug
[4] http://microid.org/

In a context where I can't rely on 1.) (because I don't want to) and 2.) 
(because I don't want to exclude to potential users) the latter approach 
could probably spare the user one of three steps that are required to 
get a microformat known by an application:

a.) put together/obtain the microformat (e.g. hcard creator)
b.) publish it on my website, copy the URL
c.) go somewhere else and paste the URL into some kind of "ping-me!" or 
"digg-this!" or whatever webform

I wonder if I can get rid of c.)

I could requiring the user to add the URL (and an email address) in step 
a.) so that I could put a MicroID into the VoteFor link. This would 
allow the application to respond to a webbug by digging the URL and 
collecting the microformat. (Putting a MicroID into the VoteFor link 
could also solve another issue: when the same vote is displayed on 
different URLs it would likely be counted multiple times.)


What am I missing here? Could this be made reliable in your opinion?

Of course the difference of 3.) (enter the URL in a webform after having 
published the microformat) to 4.) (enter the URL in advance) might be 
subtle. In 4.) the user has to understand "two things" while in 3.) 
there are "three things".

But from my experiences with no-so-super-geeky users I imagine that this 
could actually be a crucial extra hurdle. That's why I'm currently 
wrapping my head around this.


The other question:

As far as I understand from the RelFaq [5] page in the wiki (section 
"Is/Was VoteLinks a proper use of rel?") it would be right to markup the 
backlinks published in my application (e.g. on [], links on the right 
side, pointing back to the pages where a vote was found) as 
rel="vote-for". Is this correct?

[5] http://microformats.org/wiki/RelFAQ

Also, the wiki states that using 'rel' for VoteLinks (as specified in 
some previous draft) has been depracated. So, is there any known 
real-world usage of this deprecated behaviour?

Even if this might not be useful in any other way (is it?) I'd probably 
want to markup my backlinks as rel="vote-for" so that I could use it as 
a richer example for explanations.

Thanks in advance!

--
Sven