[Home] [By Thread] [By Date] [Recent Entries]
Good. You'll need them. Build an open source engine for XML-encoded real-time 3D graphics. It must support real-time routing of events in a cascade, nested transformations, full rich color rendering and if implemented by more than one source, must render precisely the same on ALL platforms. It needs an API that supports real time communications and authoring across the Internet and can support multi-user modes of operation without detectable latency (note: VOIP + 3D has a way of outing cheats here). Oh, you need a 24 fps minimum performance at full screen. I won't bother you with the resolution issues, and btw, you'll have an average of 256 to 512k RAM to allot to your application and the HTML framework plus all of the other objects of the 'operating system within the operating system'. Slow motherboards and cheap graphics cards are part of the challenge too. We won't even discuss real-time 3D sound yet. Still look easy? The difficulty has been underestimated more than once. So what we get are closed box java streaming solutions, patents pending, and closed server side systems. No HTML browser required. Or we get HTML plugins with a considerable interoperability gap among the platforms. Or we get a small handful of robust implementations that run ok in the browser, are outstanding when running standalone, but have teams with decades of experience writing them. Oddly enough, the VRML Cybertown installation was running on two Dells and had as many as 150 avatars in a room simultaneously. SL maxes out at about 64 in a room but it supports in-world building. Sorry, no private client-side building allowed. That's soooo 90s. The power required is required and the server farms aren't eco-friendly. It helps if you give the client something to do but there is no barrier to competitors if you do that and you can't indemnify the IP of the authors, but you can't anyway. See CopyBot. IP is not very valuable if you can't reuse the product anywhere else so it's a content flytrap but hey, the XMLers love it because they think it is good for the web. There are applications for the Internet which aren't made easier by HTML frameworks and they are intensely competitive. XML made some data reuse easier but in more cases than not, the parsers are hardwired and hand-rolled, don't rely on the browser framework, and don't use XML Schemas except for the authoring tools. They can interoperate with the web browser but the utility of that is what is in question, not the ability. That aside, they are a good use case for Relax. That isn't attitude. That's virtual reality. You can't smug your way past it and you can't fake a solution. The wine won't help either. The sun will. You'll learn about shadows and why they are deuce expensive in an environment where things move in 4 dimensions arbitrarily. The W3C? Didn't even try. They left it to the experts. len From: david.lyon@p... [mailto:david.lyon@p...] 3D graphics are not as hard as they look. And there seem to be good experts scattered around the world to help out if need be. > Understand our problem? To bend to your rules, we have to commit commercial > suicide while you jabber and Raph the Kos, Clay the Shirk, Rosey Phil laugh. oh Len.... sounds like you're having an attitude today.
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] |

Cart



