Do they want a car from 2025 as well? Surely it is part of the professionals job to explain that HTML 5 is still in the early stages of development and should NOT be used for anything other than web pages set up specofocally to test whether its features do what they should.
Yes, it is my job to explain to them why I think HTML5 is not the ideal option to use at this time, but so far, I've managed to convince only a few, by far not everyone. And how could I? If you look at the massive support HTML5 is getting from the "big" industry players, then many see that as enough justification to go the HTML5 route. In fact, I've been told numerous times that I was hindering progress by suggesting to use HTML4.01.
Like I said, I find it's too early to use HTML5 and don't use it whenever I can, but sometimes I have to respect a client's wish unless I didn't care about paying my bills.
Since html5 will be in development for at least a few more years I think it is better to code now in a way that will definitely not break after html5 becomes official rather than use html5 now which will most likely need to be fixed in your code after html5 is official. Using html5 also requires a html5 doctype which currently causes all sorts of markup validation issues, so I don't see any point in using html5 atm.
AussieJohn: by the time I found the problem, the code was huge and in production. My only tip-off was that browsers who supported HTML5 input types were failing and the rest did fine. Debugging found the number-type inputs weren't being looped through in various functions, but never did a reduced case to figure out why. Since I wasn't on the project anymore, all I could do was fix it as soon as possible by changing types back to text. I don't know what the ultimate reason was.
I'm not blaming HTML5 for the problem, but that it seemed to cause an unexpected result when I wasn't looking. Next significant JS I do with forms will test for the new form types while in the basis writing phase.
In fact, I've been told numerous times that I was hindering progress by suggesting to use HTML4.01.
Obviously they probably don't understand or appreciate craftsmanship if they are coming up with such weak paper-thin arguments for using Fred as "hindering progress". Since it is non normative and is not supported in current browsers obviously they've been mainly blinded-stupid by buzzword hype. Hindering progress would be creating websites that require JS to even have basic functionality and are a web accessibility nightmare.