It seems that my HTML5 switch has sparked some thoughts.
On the del.icio.us page for the article, Jonathan Rentzsch writes: “Circa 1999 I predicted XHTML would Win™. I failed to consider W3C’s complete bungling and now agree with Jesper, the future is HTML5. So long XHTML.” Also on that page, Alessandro Fulciniti wonders: “”From this moment on, Waffle is served in HTML5″… Is that a good idea at the moment?”
The story as I’m concerned is very simple. When it comes to things like Windows vs [Your Linux distribution] vs Mac OS X and so on, it’s often really not all that clear-cut. When you have powerful tools that you’re comfortable with, it’s often hard to change just for the sake of changing to something that’s technically better and more capable - ie some folks stay with Windows 2000 and a bunch of programs that they are happy with and don’t wish to upgrade to Windows XP or Vista, or even jump ship to something that’s got plenty of programs and take up less resources than Windows 2000.
That’s not how it is with HTML5. HTML5 is technically clearly superior at being a specification, and as has been discussed at length, mostly by me, it’s much more closely mirroring the direction in which the people who actually use HTML would like to take it. From what I’ve seen, there’s very little debate that HTML5 is the better spec. What HTML5 has going against it is just that it’s not gained wide-spread acceptance yet. Opera and Mozilla are behind it; Google is funding Ian Hickson (the editor) and, I assume, Mark Pilgrim; at least the better, non-legal-dept., half of Apple are throwing it out there and poking it with a stick. Other than that, the rest of the world, at least the rest of the world that lies in Redmond and heads W3C chairs, doesn’t seem to want anything to do with it.
Luckily, this is where common sense comes in. Adopting HTML5 full-out would have taken a few tag changes (and not actually work at the moment); adopting HTML5 investigatingly takes literally less than two minutes. Very few changes need to be made and your site will work just fine. What you’re doing is symbolic, but it’s important. You’re running a stake into the ground, and you’re saying: “This is HTML5 country. I am for a future of the web that means your uncle will not have to stop at the fact that you forgot to close an entity reference on line 20, column 8, when you or your HTML generator mess up. I know a great spec when I see it, I look forward to it, and I want to help make it happen.”
I encourage you to do so. The HTML5 spec runs 191 printed A4 pages. What you need to know is that they are mostly really great and that you won’t need to read 189 of them.
Update: Again from the del.icio.us page; the signature ‘numberless’ notes “ah crap. Now do I need to re-learn HTML? I wish I understand this stuff better.”
You don’t have to re-learn HTML at all. Nearly everything is drop-in ready. There will be a bunch of new tags that you’ll want to use to improve the readability of your markup - <div class="x"> and <div id="y"> only take you so far, and the closing tags are hell - but you can’t use them just yet.
Update: Apparently, depending on how you define it, Safari may be thunderously incompatible with feed autodiscovery.
According to the Atom autodiscovery specification: Well, the ‘RSS’ autodiscovery button stops appearing when you put in another relationship in addition to alternate in the <link> element pointing out the feed. In every single current (X)HTML specification specifying the <link> element, the rel attribute is defined to be a space-separated list of relationship strings. The correct behavior is to watch for “alternate” as one of these values in the list - Safari bails out after encountering another.
The RSS autodiscovery specification, on the other hand - to be fair the one n8han should use since he serves RSS, but nevertheless just one of the two standards Safari should adhere to - specifically acknowledges that the HTML specification allows more, but dictates that “alternate” must be the only value and must be in lower-case.
See also: Safari RSS autodiscovery sucks.