Re: [whatwg] [2D Canvas] Proposal: Losing and restoring rendering contexts

2013-10-22 Thread Mark Callow
On 2013/10/19 3:19, Justin Novosad wrote: Please share your thoughts. I think it these events are needed, unfortunately, as it doesn't look like GPU contexts will be virtualized any time soon. For the open issues, I say yes to the first. Note that this matches what is said about the

Re: [whatwg] [2D Canvas] Proposal: Losing and restoring rendering contexts

2013-10-22 Thread Justin Novosad
On Tue, Oct 22, 2013 at 3:07 AM, Mark Callow callow.m...@artspark.co.jpwrote: I say a weak no to the second. Testing can be done with shims and there are probably better ways for apps to do resource management. After doing some digging, I totally agree. I found that the webgl debug utilities

Re: [whatwg] [2D Canvas] Proposal: Losing and restoring rendering contexts

2013-10-20 Thread Rik Cabanier
Hi Justin, I spoke with some game developers and they are really pleased with this proposal. Adobe introduced a similar concept in Stage3D. Some of the pain points/mistakes of that implementation (such as the throwing of exceptions during a context loss) are addressed in this approach. On Fri,

[whatwg] [2D Canvas] Proposal: Losing and restoring rendering contexts

2013-10-18 Thread Justin Novosad
Hi All, A few weeks ago there was a discussion on the graphics-dev@chromium.orgmailing list about the idea of adding support for discarding and restoring 2D canvas contexts. https://groups.google.com/a/chromium.org/forum/#!topic/graphics-dev/CQJXpXxO6dk This idea has been discussed in the past