On Wed, Dec 19, 2012 at 1:29 AM, Robert O'Callahan <rob...@ocallahan.org>wrote:
> On Wed, Dec 19, 2012 at 5:39 PM, Rik Cabanier <caban...@gmail.com> wrote: > >> On Tue, Dec 18, 2012 at 1:14 PM, Robert O'Callahan >> <rob...@ocallahan.org>wrote: >> >>> It seems to me a synchronous constructor that creates an ImageBitmap for >>> a subimage of another ImageBitmap (but not other kind of source!) will be >>> needed. >>> >>> >> Did you mean to say a synchronous constructor? >> > > Yes. That's what I said. > > >> Reading the new spec ( >> http://www.whatwg.org/specs/web-apps/current-work/#imagebitmap), I still >> fail to see how this is better than a drawimage with dx=dy=0 and >> dw/dy=canvas width/height >> > > You mean, how is using createImageBitmap better than drawing the input > ImageBitmap into a temporary canvas and then using that canvas instead of > the result of createImageBitmap? > > Basically, createImageBitmap is a more direct approach that's easier to > optimize to avoid copying the image data. Would it also handled HD backing stores automatically, like drawImage does?
