Now, as a microformats lover and semantic dork, this was big news for me. Big, exciting news. Of course I should write a blog post … Unfortunately, I'm busier than a one-legged man in an ass-kicking contest these days.
So, since I already wasted a few hours of time chatting and Twittering (twatting?) with folks yesterday about the topic, I'm going to recycle those thoughts and conversations for you here now (you're welcome).
There Is No Versus
First, though, I want to stress that I actually don't believe it is about "microformats vs. Schema.org." Just like I've never believed it was about microformats vs. RDFa or microformats vs. microdata (which is what Schema.org uses).
Yes, I love microformats. They are beautiful in their simplicity and ease–of–use. I dig the Open Web nature of the microformats process and community. And, most of all, they deliver that key machine-readability I love about semantic markup.
HTML5 microdata and RDFa do the same. They are, comparatively, more complex and verbose, but they, too, deliver that rockin' machine-readability. And all can happily exist together in the same universe, each filling their niche. Anyone who tells you otherwise is likely just in the mood for a "my semantic technology is better than yours" argument that won't get anyone anywhere.
Microformats, RDFa and microdata: three stones, one bird.
Keep Calm & Carry On
I first heard about Schema.org from Todd Barnard on Twitter:
In response, I twatted (yep, I've decided that's the correct term):
I understand the immediate sense of alarm that many microformats devotees may feel upon hearing the news about Schema.org. I certainly had a moment's thought of "oh great, now my book is completely irrelevant," and "fuck, now I need to replace my microformats with microdata!?"
Even my friend Ian Pitts pinged me on IM yesterday:
RuhRoh …Beginning of the end for MF?
But I really don't think all is lost.
If Schema.org proves to truly become a standard for consupmption by the search engines and other machines … If Schema.org proves to become a standard for publication that developers willingly adopt, then how awesome is that? I mean, seriously, think about it.
We will finally have a semantic technology that has a very real chance of gaining wide adoption. Search will get better. Content syndication could get better. New applications and other machines will arrive that make functional use of this new structured data (Did you see Watson on Jeopardy?)
That all sounds so good to me.
And if Schema.org proves to disappoint, then we are where we are today, with, hopefully, some valuable lessons learned. From my initial perspective, Schema.org is nothing more than well-organized and (I must say) well-documented HTML5 microdata. And HTML5 microdata will be around regardless of Schema.org.
Plus, in either scenario, I'm still an expert in structured markup and will have work and opportunities. As I twatted to Todd:
This idea of a "semantic mindset" is basically the gist of my In Control presentation earlier this year, How I Learned to Write POSH & Learn to Love the Semantic Web. My interest in standards led to my interest in POSH, which lead to my interest in microformats, which led to my passion for both the semantic web and The Semantic Web.
Thanks to microformats, I already think semantically in my development. My markup is more semantic than ever. I can analyze content quickly and easily to apply semantic vocabularies. The syntax I use is, frankly, moot.
I won't lie though. From a front-end standardista perspective, microdata is ugly and complex. RDFa is even more complex and more appropriate for uncommon content.
Microformats are so simple and borrow from already-understood conventions, so I do prefer them. But, as microformats has struggled to gain adoption — particularly of the consumption variety — these benefits don't really give developers much incentive to actually learn and use.
Ahead of Its Time
Perhaps Schema.org does herald the end of microformats as we've come to understand them. If that happens, though, won't that be the natural evolution that microformats, ultimately, wanted?
Microformats were, undeniably, ahead of their time. The founders wanted Sir Tim Berners Lee's Semantic Web, and they wanted it back in 2003. And so they embraced Open Web methodologies to come up with a semantic vocabularly for describing common web content using existing standards. It was (and still is) brilliant, forward-thinking and chock full of awesomesauce.
During my IM chat yesterday with Ian, I commented:
[microformats] did good work getting people to understand the value of semantic markup and design patterns for machine readability … [but] adoption was always a challenge.
As everyone knows, microformats have remained a very niche subject. Awareness is comparatively limited. Machines, particularly native machines, that parse and consume microformats are definitely limited.
Publishing microformats has steadily increased over they years, but without consumption tools, most of that has been with the hope that "semantics are so awesome, of course the machines will come … someday." At least that's how I feel about microformats … and I happen to be someone who really digs them.
Is Schema.org the herald of that "someday" I've been waiting for? Frankly, I hope so. Even if that means changes for microformats.
Some Practical (and Pedantic) Thoughts
Schema.org does not validate and it uses an XHTML Transitional DOCTYPE.
I don't get this. I know that validation for validation's sake isn't the goal, but seriously? Fix your fucking markup (especially, if it is just a DOCTYPE change).
Speaking of DOCTYPES, if you are promoting microdata, why wouldn't you use the HTML5 DOCTYPE. Again, I don't get this. Fix your fucking DOCTYPE. This just scream laziness to me and isn't at all in the spirit of (lowercase) semantic web.
Schema.org could use some POSH
Just like the microformats wiki has been plagued with non-POSH and (frankly, pretty crappy) markup examples, Schema.org isn't that much better (well, a bit.) C'mon people. If we want to encourage broad adoption, let's provide good, practical and POSH examples of how to publish.
Schema.org syntax is valid markup and can be combined with microformats but …
If you are using the HTML5 DOCTYPE, the microdata syntax that Schema.org is built on is completely valid. If you aren't using that DOCTYPE, your markup won't be valid.
Seems simple and straightfoward, but it is worth reiterating that just because a page/site doesn't validate, doesn't mean you don't have support for microdata. The search engines will parse, regardless of validity. Just like a modern browser will still render things like
<audio> with an invalid DOCTYPE.
Microdata and microformats can existing in the same markup together and be completely valid and appropriate. However, if you are utilizing the microdata or microformats syntax for Google's Rich Snippets, beware!
If you’ve already done markup on your pages using microformats or RDFa, we’ll continue to support it. One caveat to watch out for: while it’s OK to use the new Schema.org markup or continue to use existing microformats or RDFa markup, you should avoid mixing the formats together on the same web page, as this can confuse our parsers.
Like I said, I'm busy. So this entire post reflects my initial impressions of Schema.org, having not actually implemented, tested or experimented.
Also, while I've talked about and written about HTML5 microdata, I've never actually used it (instead, favoring microformats). So, I also have only superficial impressions about it, too.
But even with just superficial knowledge, I'm super psyched about Schema.org: