On Wed, Oct 14, 2015 at 11:30 AM, Bobby Holley <[email protected]> wrote:
> On Wed, Oct 14, 2015 at 11:06 AM, Boris Zbarsky <[email protected]> wrote: > >> But Bill is right. If we're trying to expose isolated world semantics to >> addons, we need to think some about how to actually do it. >> > > I don't think we can/should get there, tbh. Trying to do general emulation > of the corner-case behavior of Blink's reflection model doesn't seem > possible without switching to that model, which is basically off the table. > > What I think we should do instead is targeted compatibility fixes of the > stuff that actually bites us in practice. > I'd like to understand in more detail how isolated worlds are implemented. We might be able to get some ideas about how to improve our model without totally changing it. Or we'll find out that they're bad or have serious security problems--that's fine too. I worry that saying "we'll spot fix as needed" isn't going to work long-term. Isolated worlds seem a lot more appealing to add-on authors since they generally "just work." Developers are going to complain a lot if we tell them we're not compatible and we don't give them a good reason why our model is better. > This is all to say that I think we may just be able to fix the TypedArray > thing by, say, just allowing slow indexed access over Xrays with an opt-in > flag on the sandbox. I'll see if I can hack something up for bug 1198934. > OK, thanks. That seems to be one of the biggest issues so far. -Bill
_______________________________________________ dev-fxos mailing list [email protected] https://lists.mozilla.org/listinfo/dev-fxos
