I think the point of HTML5 is that it’s not meant to handle those things. Nor even is browser-hosted JavaScript really. I suspect the big thing for HTML5 on the horizon is televisions and tablet computing.
Cas 
OpenGL in HTML5 is supported as Canvas WebGL.
Not sure about the left/right mouse button events, haven’t tested it well enough, but there are plenty webgl demos out there.
You have WebGL with HTML 5. Though its horrible slow compared to Java JNI bindings it is still possible to use and its growing.
//edit, damn work ^^" two posts in the meantime
And still shafted, because it requires Microsoft to implement it and roll it out.
Cas
I can’t wait for all the exploits to come out against WebGL. Does anyone think their OpenGL driver is hardened for security?
Exactly why Microsoft aren’t enabling it. WebGL is really just a small hole though really. Does anyone really think that with Javascript and browsers having basically all the power of an ordinary desktop client that the latest crop of HTML5 browsers aren’t basically going to be riddled with security holes like swiss cheeses to the extent that Java, PDF and Flash holes are going to look like Fort Knox security? And think about it: it’s easy to disable plugins. Not so easy to disable your browser.
Cas
Browser makers can do little with regards of security when it comes to plug-ins for your browser, e.g. Java applets, PDF docs or Flash. The makers of those plugins have to ensure security, so you have browser security as a distributed responsibility instead of as a single responsibility.
If, let’s say, Mozilla implements HTML5, it’s up to them to make it secure. They cannot guarantee security if you use 3rd party plugins, nor should they. And it’s a lot easier nowadays to update your browser (e.g. FireFox) than it is to update your browser + all your plugins. At least for me, it’s a whole lot bothersome to accept to update that annoying Adobe Acrobat reader than it is to update FireFox.
So, in practice, HTML5 might reduce security holes because it reduces the need for plugins, and people will overall be more up-to-date.
I somehow doubt it. Only time will tell if my predictions are right but there appear to be plenty of security questions over the HTML5+JS combo. Won’t be long before we see the big ones being used for sploits.
Cas
I believe some browsers, either Firefox or Chrome (or both?), only enable WebGL on certain cards in order to avoid security holes. Plus most issues found will probably be OS and driver specific.
Many CSS 3 selectors were implemented in IE 8 and even IE 7. IE 9 supports pretty much all of the selectors that other browsers supports, and the ones which are missing are typically also missing or very buggy in other browsers.
For visuals, all the main stuff is already supported by IE 9. This includes border radius, box shadows and CSS transitions (through the -ms- prefix). CSS gradients are missing, but you can use SVG as a background (and of course the old DirectX filters, which even IE 6 supports, but they don’t play nicely with properties such as border-radius). The only noticeable property which is missing is text shadow, for which there is no pure CSS alternative (you’d have to include a piece of text twice in your HTML and use one of them as the shadow). Gradients and text shadow are supported in IE 10.
The reason why Webkit and Firefox support so many more advanced features then IE is because the features are rushed in, at sub-par quality, with bugs and issues left outstanding. Take a look around the Mozilla bug tracker and you can find plenty of known HTML5 bugs which are several years old, and I personally encounter these bugs all the time. For example in Firefox you can break the matrix inside the canvas, leaving it in a state where it’s no longer usable, at all, even if you explicitly set a new matrix using ‘setTransform’. It’s also possible to setup and draw to a canvas, and have the contents appear on a totally different separate canvas, due to rendering bugs.
Pretty much all of my JS ‘just works’ in IE 9, whilst it is rare that I don’t encounter a browser specific bug in Firefox or Chrome (especially Firefox).
That’s not the problem (Well, unnecessary copying is, but that’s not what we are talking about).
The problem is the global environment and other objects caught in the lexical environment when declaring closures. The global variable could be changed. Definitely not like Erlang.
Haha.
Two Web Workers cannot reference the same variable in JS, so yes, it’s like Erlang.
I use GWT. JavaScript is like tacky duct tape. As long as I don’t have to touch it, it doesn’t bother me.
Yes, the support is growing, but to slow! And even thus features implemented were buggy. Attribute Selectors, pseudo-classes and a lot more. I agree that the IE Dev Team has done a great job improving the render engine since IE6 but there is still a lot of work to be done. For example the border-radius, why haven’t they implemented in IE8? Instead they kicked out the interface and done another horrible to use interface. At least they fixed the Box-Modell Issues.
Well and at the other site of their products they stripped of Outlooks HTML caps back to the '90s!
Maybe all this will end some day.
//edit: Another thing i missed out. IE9 hardware acceleration is pretty neat when you hear about it. But when you have to disable it to see any website its pretty lame. Not to mention that the initial 64bit release did not worked either.
I’m looking forward for IE10, but they could have done so much more…
A little unfair, since the box-model was fixed when IE 6 came out. That was quite a long time ago. I also personally use the box-model quite heavily on my pages (through box-sizing), as certain layouts are relatively simple to build using it.
I hate the box-model, been CSS’ing a bit lately, and it still whacks my mind.
Depends on your market. IE is below 30% in some countries. Also, IE’s share is generally a lot lower among gamers.
See box-sizing.
To be honest, I think that box-sizing:border-box would have been the better default. It’s a lot more powerful and also more flexible.
I think that’s one of the worst URL fails I’ve seen in my life. ;D
Well, that’s UBB for you. Sometimes you need quotes, sometimes they are optional, and sometimes they don’t work.