How convenient for a web company to present a graph with no units or measure of "user experience" and, surprise, the web is almost as good as the desktop now. How... made up and fake.
How one earth did they measure "user experience"? What is responsible for the up and down fluctuations (note, on a logarithmic scale) in the relatively flat line for "native" user experience?
Finally, of course Google's own Chrome is the best.
I would recommend anyone interested in Canvas to check out this exhibition of experiments. There are a lot of "I can't believe it's not Flash" projects collected here:
At this point, if MS does not implement these features then it will fall further behind and lose relevance. There are enough people using non IE browsers, that one could create a new site leveraging these features and ignore the IE users, or provide a significantly downgraded experience. The thing is, it would not take much to do so. A couple of fun games here and there with some useful apps at a significantly reduced cost and there goes more of IE's market share.
But we can hope that it happens the other way around: what we need is a definitive "killer app" that can only be done given near native speed JS + HTML 5 capabilities.
If it truly IS a killer app, it will kill IE in short order.
If Tesla, Lamborghini, and Ferrari start making all electric cars, but Ford doesn't, that's consensus right?
Market share matters. Maybe for specialized hacker sites we can cater to only the good browsers, and graceful degradation is a must anyway, but it's a hard sell to build a successful mainstream site by eliminating 40% (or more) of users right off the bat.
That's... such a wonderful comparison. Because we've never seen an apparently dominant American car company lose serious ground to competitors because it refused to pay attention to what consumers actually wanted, have we?
And yet, how much have any of those companies changed the car industry? (Although I'll grant that Tesla has a chance).
My point is that the standards aren't set by the number of companies that support them, they're set by the market share. Talking about browser makers like IE doesn't exist is wishful thinking at this point, just like talking about car makers as if GM and Ford have already stopped producing cars is wishful thinking.
It doesn't mean that you shouldn't innovate around them, just that you can't call your work a 'standard' until at least a majority of the market is on board. (I mean standard in the market sense here, not standard in the W3C sense).
Just because the product still runs on gasoline, has a steering wheel, and puts four tires on the ground, that doesn't mean the industry around the product hasn't changed fairly dramatically. It's not a revolutionary change, sure, but honestly, neither is implementing web standards. You're still rendering html, applying css, and executing javascript.
As for the point about what constitutes a 'standard', you're just saying there's a difference between an official standard and a de facto standard. I won't dispute that. I don't think that's the story here, though. The story is that IE used to be the de facto standard through merit before growing lazy as hell and losing its technical lead, and that people are increasingly willing to think they can move forward without it. It's been steadily losing market share since 2004, while the web has absolutely exploded in that same time frame. Are you really saying that its a bad bet at this point to not particularly care about IE's attempts to bully the market by not playing nice with the standards the other browsers are agreeing to support?
Consumers care about car internals at anything more than a superficial level?
If HTML5 is actually important to improving the end-user experience, if browsers that implement it have a real, demonstrable advantage over those that don't, then it doesn't matter if the consumer cares about HTML5 or not. They'll still want the results that technology enables.
Right, the typical consumer can crank out exactly which relevant industry standards their car's internals adhere to.
When consumers say they care about the internals of their cars, what they actually care about are things like power, fuel efficiency, torque, whatever. They have nothing more than a superficial understanding of how the actual mechanics of the vehicle translate into these qualities. Talking about V8 engines, etc, is, for the typical consumer, nothing more than marketing.
HTML5 could certainly be a marketing device (5? That's one more than 4, it must be better!), especially if you showed end users cool things that web sites can do with HTML5 that they can't do with 4.
I think that's where we disagree. I haven't come across any normal folk who even remotely cares about HTML version or javascript library, etc. I have however come across many people from various backgrounds talking about how different cars perform.
edit: ultimately we'll probably need some numbers to back our claims. I've got way too many bugs to fix to do any significant digging. But I'm open to counterexamples if you can find some numbers.
You've never heard anyone complain that a particular page looks weird, or that it loads or runs really slow? You seem to mixing up the levels of abstraction as we cross the boundary moving into the car analogy. Cars have industry standards, too. Nobody knows shit about them. But they still care, indirectly, because adhering or not adhering to those standards affects in some way the experiences that the end users do care about. That's the only point I'm making. If you actually disagree with that, I'm really surprised.
I'm disagreeing with the degree of care. I don't think most people give a hoot about web standards. I do think most people care about their car standards/internals.
I'm wondering if it'd be practical to port Gecko or WebKit to Silverlight or Flex, so a site could simply launch itself in an embedded browser, transparently from inside IE?
While a very interesting idea, I don't think it would be practical. Both platforms have some very strict sandbox regulations which would make it difficult.
How one earth did they measure "user experience"? What is responsible for the up and down fluctuations (note, on a logarithmic scale) in the relatively flat line for "native" user experience?
Finally, of course Google's own Chrome is the best.