Some old e-mails regarding exposing APIs in Workers... On Thu, 14 Mar 2013, Tobias R. wrote: > > I can see efforts in exposing canvases within a worker [1]. Does that > mean the following APIs are exposed and available within a worker as > well?
Since you sent that e-mail, I updated the spec to use the new "Exposed" WebIDL annotation to say what is exposed in workers. > - CanvasPathObject [2] Assuming you mean Path2D, yes. > - SVGMatrix See: Path#addPath [3] That's up to the SVG folk, but presumably yes. On Tue, 9 Apr 2013, Rik Cabanier wrote: > > It's unclear at this point how fonts would be resolved in the canvas > path object. If this did not yet get resolved, please let me know. On Thu, 14 Mar 2013, Simon Pieters wrote: > > The new canvas proxy stuff is supposed to make canvas work in workers, > but I don't see any normative text that makes the necesary interfaces > and constructors actually available in workers. The spec needs to > explicitly say that CanvasRenderingContext2D, e.g., is to be available > in workers. This is now done, using [Exposed=...]. On Tue, 9 Apr 2013, Rik Cabanier wrote: > > Is there a list somewhere that lists all the interfaces that are > available to a worker? I don't think it's listed explicitly in one place currently, no. (This information is spread across many specs, so generating an up-to-date list would be non-trivial.) On Wed, 10 Apr 2013, Boris Zbarsky wrote: > > We should solve this problem. At this point we have some interface and > prototype objects that should only appear on Window, some that should > only appear in Workers and some that should appear both places.... I > wonder whether it makes sense to specify in the IDL which kind the > interface is. This is how we ended up doing it. -- Ian Hickson U+1047E )\._.,--....,'``. fL http://ln.hixie.ch/ U+263A /, _.. \ _\ ;`._ ,. Things that are impossible just take longer. `._.-(,_..'--(,_..'`-.;.'