- From: Paul Cheyrou-Lagrèze <paul@sketchfab.com>
- Date: Thu, 24 Nov 2016 09:01:30 +0000
- To: public-webvr@w3.org
- Message-ID: <CAKAy3J9RwvVphCWRKuO-kfco4hz3TvriqYgLJm5y1yru8g0cFQ@mail.gmail.com>
WebVR "requestPresent" currently requires user gesture to be activated. In our case, we do load new page or new Iframe on user clicks, and it’s in those new document that webVR requestPresentVR is done, at Document Load event. (link to google doc explaining our current website flow: https://docs.google.com/document/d/1R6Wt14K5jvdXLQc7niYSed_YFSAB3tHPMqkS4b7hEeU/edit?usp=sharing ) So current "user gesture" requirement doesn’t work for us. What would work for us, and seems flexible, would be to have the "requestPresent" be permitted once and for all on our domain (sketcfhab.com), after asking the user, and not requiring new "user gesture" even on different webpage of the same domain. (staying in VR across webpages). Seems using “Permission API”, ( https://github.com/w3c/permissions ) as done with "Geolocation, GetUserMedia, , etc" would allow that ? We would do a demand using permission API once and for all for our dommain " https://sketchfab.com", prior to any requestPresent, and then we can requestPresent without user gesture at any place in the code we want (in our case, once everything is preloaded and ready to be viewed). Webapp not having "permission” using API would still need "user gesture", so that would be a non-breaking change. Paul, Sketchfab.
Received on Thursday, 24 November 2016 13:34:51 UTC