"Why I Hate Microformats"
Robert Cooper: They are a hack to get around a problem that is going away VERY shortly.
I’ve been using microformats more and more in my recent development. Even though I have yet to really find them useful personally, I think they are great concept. Anything that makes content more accessible is a step in the right direction.
But I have to say, Robert Cooper makes a great point: “The problem is, this is a serious abuse of HTML.” XML already has a solution to deal with this problem, namespaces. The problem is just that IE doesn’t support namespaces in CSS, so you can’t style anything but plain HTML. Well, with 7 on the horizon, this problem goes away.
So why are we still using microformats? You could say it’s so that they are accessible to IE users, but I imagine microformats are really only used by a segment of the early adopters. How many IE users are seriously using them? Besides, it wouldn’t be to difficult to write some Javascript to transform the XML into style-able HTML. That would at least be a temporary solution until IE 7 is released main stream. And By that time, maybe the use of the new “microformats” would be more mainstream.
5 Comments
Ah, I disagree almost completely.
First, while Microformats are still not obviously useful to the average user, they easily can be without them knowing. Mark up your event calendar with hCalendar and you can make an instant .ics file download that can be automatically polled for updates! I do this to a lot of sites we create, and pull them into Apple’s iCal or Google Calendar. I would love to see Apple’s Address Book allow me to subscribe to an hCard… maybe someday.
Second, the main motivation for microformats is to make them easy to create. All the XML stuff, while great, fails in simplicity. HTML is popular not because its perfect (far from it) but because it is simple enough for average people to pick up.
Microformats don’t strive to be perfect—they don’t have to be. They need to be simple and good enough. Microformats will succeed because of, not in spite of, their simplicity.
Sure namespaces would be ideal, but the current method is not an abuse of HTML at all. It is valid and semantic, while also being a common convention for developers. All very good things.
Finally, your average surfer won’t (and shouldn’t) know what Microformats are. This is fine the same way them not knowing what HTML is is fine. They will use the next generation of tools like Tails, like today, have easy links to .ics files that the web developer didn’t have to take the time to create.
I think you make some great points, but I still like Robert Cooper’s better.
I think you’re right in when you say “while Microformats are still not obviously useful to the average user, they easily can be without them knowing.” Any good technology should work exactly like that; it should be completely unobtrusive.
I disagree. How is the cryptic HTML syntax easier to create? Every time I go to use an hCard or hCalendar, I have to open up the documentation to figure out how to mark it up properly. If it was XML, not only is it referencing a schema definition, which would allow me to verify that it is valid—whereas, with the HTML I just have to pass it through one of the parsers and check if it comes out right—I could also get autocomplete in the editors that support it.
Microformats are an abuse of HTML because they’re taking semantic information and moving them into classes. Now, the only way to extract semantic information is to create special parsers that ignore the tags and use the attributes instead.
This is exactly the kind of problem that XML is suited for, and in my opinion, the reason for the move to XHTML: to overcome the limitation of only using HTML tags, and to be able to embed other data types inside of HTML.
…and this is why XHTML 2 has some really nice support for embedding RDF data in a document. There are alternatives to microformats.org where the namespace problem has been taken care of (google for embedded RDF). For some reason, everyone seems to think that microformats.org is the only microformats initiative.
Peter, do you have any links to these alternatives?
Brandon, I believe Peter was referring to eRDF
RDFa is another that isn’t much use until XHTML2 but is clearly where things are headed. You may also want to read about GRDDL for Gleaning Resource Descriptions from Dialects of Languages; that is, for getting RDF data out of XML and XHTML
Post a Comment