Re:Ask HN: Does HTML5 excite you? Why or why not?
by Semiapies on Sunday November 29, 2009 @05:06PM
(#966891)
None of these things require HTML 5, and all of the worthwhile things had workable implementations before HTML 5.
Formal specifications for things that had de facto cross-browser support for multiple browser generations are nice, but not amazing.
"This is the first time our core platform has seen any serious evolution since HTML4.01 was published in 1999!"
And that's just simply untrue, but oh, well.
"I'm amazed to see so many web developers unexcited by HTML 5. This is the first time our core platform has seen any serious evolution since HTML4.01 was published in 1999!"
And I'm boggled to see such giddy fanboyism modded up so high.
Re:Ask HN: Does HTML5 excite you? Why or why not?
by rimantas on Tuesday December 01, 2009 @12:39AM
(#969305)
Can you list _worthwhile things had workable implementations before HTML 5_ and the stuff they implemented before HTML5?
<canvas>? <video>? local storage?
Re:Ask HN: Does HTML5 excite you? Why or why not?
by Semiapies on Tuesday December 01, 2009 @04:18PM
(#970769)
With Canvas, you might have a point, though it makes a poor justification for the rest of HTML 5. I've yet to need Canvas for anything, though. Video is silly to even ask about. Google Gears rather predated HTML 5 local storage.
Now, give me something I couldn't do before with nearly-universal plugins.
HTML 5 is clearly aiming to bring the open web platform up to par with desktop APIs. I expect many desktop applications to be developed using HTML5+CSS+JavaScript within just a few years.
And even if that doesn't spike your interest, there's the critically important role the HTML 5 spec pays in standardising what we already have. It's ridiculous that XMLHttpRequest, one of the most useful modern browser APIs, has no spec - every browser implements it by reverse engineering the others, who in turn reverse engineered IE (which is closed source, so who knows if they got it right?). HTML5 fixes that. It also specifies things like innerHTML, ContentEditable, the execution order for script files, all that DOM level 0 stuff that never made it in to an official spec.
Finally, HTML5 specifies a formal error model for HTML for the first time. This means it will actually be possible to implement a new browser from scratch without having to spend years reverse engineering the behaviour of IE and friends when faced with invalid markup.
I'm amazed to see so many web developers unexcited by HTML 5. This is the first time our core platform has seen any serious evolution since HTML4.01 was published in 1999!
As for browser support, it's happening right now. If you develop for the iPhone / mobile WebKit stuff most of the above features are sitting there waiting for you to use them. Try this page in your iPhone for example:
http://simonwillison.net/static/2009/navigator.geolocation.h...
Even IE is catching up - IE8 supported a bunch of HTML5 concepts (key/value based storage, cross document messaging, hashchange event and more) and IE9 is set to support even more.
HTML 5 is the most exciting thing to happen to web development in a very long time.