[webvr] Medium post: How can we ensure WebVR is not an off-ramp for the web?

rob.manson at awe.media rob.manson at awe.media
Sun Sep 18 11:12:27 UTC 2016


Hey Florian,

yep I get what polyfill/shim means...thanks 8) But in WebVR's case it's 
not quite the same as there's completely parallel implementations of IMU 
APIs - VRPose and DeviceOrientation/Gyro. Yet the only real use that's 
come up for DeviceOrientation/Gyro is really VR/AR - however the focus 
on Cardboard and WebVR nearly broke that at one point too.

Like I said, I understand why we're here - but it doesn't mean we can't 
plan a better way forwards that benefits both VR and AR.

To develop concrete suggestions of how an integrated view looks I've 
been working on a presentation that maps out how AR and VR relate to 
each other in terms of web API's and the general scene/camera geometry.

For one example of how this model suggests WebVR should be extended - 
WebVR has defined fov, etc. for the HMD displays for each eye 
(VREyeParameters). In AR this model needs to extend right through the 
display and out into the viewable world using the Pinhole Camera Model. 
Mathematically it's easy to integrate these two geometries and it would 
be good to model this as an integrated whole. It's also possible to map 
planar touch devices like tablets/mobiles into this world too so 
developers don't each have to re-invent that.

Lots more info on that level of detail coming soon.

roBman


On 18/09/2016 8:34 pm, Florian Bösch wrote:
> On Sat, Sep 17, 2016 at 11:46 PM, <rob.manson at awe.media 
> <mailto:rob.manson at awe.media>> wrote:
>
>     Even if you look at the samples page:
>
>     https://webvr.info/samples/
>
>     There's the primary link to the WebVR experience - then some sad
>     little links in brackets that say "Use Polyfill" 8)
>
> Use Polyfill means it uses existing APIs (such as device orientation) 
> and best-guess distortion mechanisms to produce a VR-like view. 
> Polyfilling is a mechanism whereby missing functionality (such as 
> WebVR) is emulated for the purpose of compatibility with non WebVR 
> supporting browsers, and for developers who don't have VR hardware. 
> It's got nothing todo with relegating the 2D web to a second class 
> citizen.
>
> https://en.wikipedia.org/wiki/Polyfill
>
>>     /We should extend WebVR to make it a seamless part of the web of
>>     the future. /
>     This is the web's opportunity to differentiate itself from all the
>     other VR and AR platforms.
>
>     This is a hard problem...but one that can set the web head and
>     shoulders above all the other platforms.
>
> Still no concrete suggestion.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.mozilla.org/pipermail/web-vr-discuss/attachments/20160918/7af5ca33/attachment.html>


More information about the web-vr-discuss mailing list