4 ms·
Ironic that this is sponsored by IE, and IE is the one browser holding back cutting-edge browser-based games.
by james33 14y ago
Ironic that this is sponsored by IE, and IE is the one browser holding back cutting-edge browser-based games.
- harrylove 14y agoExcept IE 10, which, based on my development experience with it, is actually caught up in most respects and on the cutting edge in a few others. Grab a copy of the Windows 8 preview, load it up in a VM, and give it a try. I thought the IE 10 demos (http://ie.microsoft.com/testdrive/Views/SiteMap/ http://ie.microsoft.com/testdrive/Views/SiteMap/) were particularly impressive.
- adamman 14y agoDoes it support WebGL?
- TomGullen 14y agoDoes it support Ogg audio files? And as the other commentator mentioned does it support WebGL? Microsoft and IE drag their heels at every opportunity with the web. I'd be happy to see IE completely fade into obscurity going into the future. It's slowing progress down regardless of what their new fangled marketing campaigns say and we don't need it anymore.
- harrylove 14y agoHand waving without substantive evidence to support your position is generally frowned upon here. I do not see how the non-support of WebGL and Ogg supports your position that the IE team is dragging its heels. Chrome and Firefox do not support CSS Grid Layout or CSS Regions by default, but that doesn't mean their teams are dragging their feet.
- harrylove 14y agoFor questions about support for your favorite spec, see IE 10 Guide for Developers[1]. http://msdn.microsoft.com/en-us/library/ie/hh673549(v=vs.85).aspx http://msdn.microsoft.com/en-us/library/ie/hh673549(v=vs.85)...
- Kiro 14y agoTimes have changed. "IE9's smoothness on the other hand is remarkable. Of all browsers and systems I tested, IE9 subjectively produced the best results." http://www.phoboslab.org/log/2011/08/are-we-fast-yet http://www.phoboslab.org/log/2011/08/are-we-fast-yet
- tomjen3 14y agoIt may be smooth, but my job is to make cross-browser HTML5 games (a great gig if you can get it) and almost all of my worries are 'solved in Chrome and Firefox' but that fix somehow is not available in IE9. IE10 might, be better but then you have to deal with all the legacy people who don't upgrade. Also when it ships, it will still be behind Firefox and Chrome.
- james33 14y agoNot really. Our game studio builds browser-based exclusively, and IE is the biggest paint point we have.
- angersock 14y agoHow 'bout that WebGL? Without that, I don't give a fuck. EDIT: I'll elaborate. Go here ( http://www.caniuse.com/#cats=JS_API http://www.caniuse.com/#cats=JS_API ) to follow along if you're playing the home game. The problem is that for games--for smooth games with low invocation overhead--we really really need to be able to use the graphics card. There are cute tricks you can do with a 2D context (including, in Three.js's case, software fallbacks for some of WebGL), but nothing that really compares to being able to write modern programmable shading pipeline code. Without that (and the Audio API, and Orientation events), game developers don't really care about your platform. Now, there is (in my opinion) a storm brewing for WebGL. Go here ( http://www.khronos.org/registry/webgl/extensions/ http://www.khronos.org/registry/webgl/extensions/ ) to see the coming madness. There is nothing worse in the world than being under the stewardship of Khronos/ARB. The fact that W3C et al. haven't taken over or provided an alternative to this is bad. Microsoft could do something like make DirectWebX or Web3D or what have you. That would be annoying, but sure--we've been writing cross-API rendering backends for a decade and a half now, and so this wouldn't be terrible. Even better, Microsoft could go ahead with a WebGL implementation that calls down into DirectX. Best, Microsoft could leverage their experience with good API design--don't scoff, seriously; compare OpenGL and the latest DirectX versions--and work with the W3C and browser vendors to fix some of the yawning holes in WebGL. ( For example, killing off the extensions mechanism entirely in favor of "Is this Web3D 10 compliant? If so, you get this and don't worry.", or a useful capabilities structure. getParameter() is kind of useful, but having a guaranteed caps baseline is nice. )
- Qworg 14y agoHilariously, at least one of the people writing a tutorial uses Chrome. =)