WebGL and HTML5 Challenges For The Future | Hacker News Reader

WebGL and HTML5 Challenges For The Future

codeflow.org

32 pointstbassetto12 comments

My main concern here is that Adobe's new flash plugin will allow some of these things for 3d on the web and set back adoption of WebGL to the point that it's also DOA.

I know mouse capture seems like a security concern to many browser vendors, but just have a dialogue to agree to mouse capture and a clear use of ESC to exit it. If they don't do it, some plugin will and we'll be set back into another X years of flash dominance in online gaming again.

It's a risk but at the same time Adobe has a lot of the same issues to fix as what webgl has. It's not going to be easy for either solution to make a dent in the adoption of the 3D web. Some of the biggest problems with WebGL and Flash will be drivers, security and prevention of abuse.

At least with WebGL you have multiple groups working on solving the major issues. I don't see Flash fixing these things alone.

There also Silverlight that will come out with their own solution but like with Flash I don't see it as an easy win over WebGL.

"At least with WebGL you have multiple groups working on solving the major issues."

I'm not sure this falls under the "advantage" column in this scenario. Number of "groups" doesn't count for anywhere near as much as "number of people", and there's no particular reason Adobe can't field as many people on their one side as the other interested groups can in total, unless one of those groups really commits. Having a lot of big names on the list doesn't guarantee than any given name has more than "a guy we told to subscribe to the mailing list and scream if somebody proposes something that doesn't serve our interests". (That's by way of example. I know there is non-zero real commitment, my point is that it isn't at all the case that this group must outweight Adobe's ability to solve the problem.)

I wouldn't worry too much about that. IMHO Flash is already a moribund technology. 3D acceleration won't change that.

The real competitor here are native mobile applications.

Also, I have to wonder about the security implications of WebCL (not the C not a G), like malicious sites using idle GPUs for password cracking. That sort of thing.

I really don't understand this obsession with shoving everything into the browser. In native applications 3d graphics are great for games and certain scientific applications but camera control issues etc make them too awkward and confusing for almost anything else.

The UI demands of games seem to me to be strongly at odds with the UI needs of just about every other kind of app. I wish all this effort was directed instead at making HTML5 less painful for laying out information in an elegant and straightforward way. CSS still leaves a lot to be desired in this respect.

Any front end coders here using HTML5 over using HTML?

What are the advantages of using <article> , <section> and <footer> vs. <div id="footer>, per example?

The markup changes don't give any performance advantages, they just help to standardize things. They aren't essential to HTML5 being useful. LocalStorage is one of the big advantages. Canvas has some interesting uses, although still needs work to be competitive. Video and Audio tags could be useful in the future, if a consensus is ever reached about codecs.

Basically, you're asking the wrong question. Switching to HTML5 for markup changes is silly. The power of HTML5 isn't in semantic changes, it's in everything else.

Not sure why I was downvoted here?

I am mid-level front end coder (HTML/CSS & plug and play Jquery scripts) who hasn't used HTML5 yet (i do use CSS3 often to design sites & code).

I've only glanced over the mark up and wondered why I would use that over what I've been using past 2 years.

I know there are <video> and <audio> mark up tags that in some years will kill the need to use flash. What are the other advantages - the everything else?

I think the down vote is mostly because the article is centred on WebGL and nothing to do with the new mark-up elements in HTML5.

I didn't down vote, so I can't answer to that. The parts I found useful from "everything else" I listed above. LocalStorage is the big one (IMO). You get access to a built in, cross session data store natively. The canvas element has some interesting potential applications. I don't think it's quite "there" yet, as far as performance goes on most systems.

The reasons for the markup changes are to help standardize things. If you don't feel the need to do that, you don't need to.

You won't find a whole lot in html5 that will affect what you do from a design standpoint (unless you start throwing in some canvas elements). Your plug and play jQuery scripts might take advantage of some html5 elements without you knowing, but they've sufficiently abstracted those details away from you, so that you don't need to worry about them. Those elements won't break if you're using HTML4 markup instead.

When people refer to html5 they are often including css3 within those technologies, html5 has by most people agreed to be a marketing term that just means the recent wave of web tech, html itself is now versionless

http://slides.html5rocks.com/#landing-slide

is probably the best overview of the new "html5" features, my personal favourites would probably be localStorage, websockets and webgl, along with a lot of the css3 features

It's more a matter of everyone is doing exactly what you pointed out which makes it a convention. The HTML5 folks are concentrating more on standardizing what people already do instead of trying to tell everyone what to do.

Reply on news.ycombinator.com