To clarify, microformats specs don't seek to constrain HTML at all. They only describe expected behavior among those who have arrived at a rough consensus about specific bits of HTML. But you're free to, and furthermore encouraged to, use other bits of HTML as well. Not every bit of useful HTML semantics belongs in a microformat. Maybe this bit does, or maybe it doesn't, but the lack of a microformat should never prevent you from making your own HTML markup more descriptive.
I agree with the idea that we should first use the semantics offered by HTML before creating a microformat. But microformats are an agreement among community members on certain standard usages of semantic class names and other attributes when the HTML spec either doesn't supply what's necessary, or more specific uses are required (as in standardized class names). What I meant by the word constraint was the fact that the spec for Vote Links originally stated that you should exclusively use the rel="" attribute to mark up a Vote Link, and then was changed to more correctly reflect the nature of the relationship by using the rev="" attribute--once again exclusively. This means that according to the microformats spec for Vote Links, a vote link goes one way. Sure, you can use a rel attribute to link back, but that doesn't mean that the usage will be standardized (aggregator friendly, etc...). Hence my proposal to specify the proper usage of both rev and rel. I don't agree that you can simply use rel="vote-for" because it incorrectly gives the impression that you're "voting for" when really you were "voted for", hence my suggestion to use the following: rev="vote-for" rev="vote-against" rev="vote-abstain" rel="voted-for" rel="voted-against" rel="vote-abstained" Because really, the relationship isn't the same in both directions. Your thoughts? A. _______________________________________________ microformats-discuss mailing list [email protected] http://microformats.org/mailman/listinfo/microformats-discuss
