FYI and dirty if it doesn't matter: ES6 adds an invoke trap to proxies. Cc'ing Allen.
/be On Sep 11, 2013, at 2:37 AM, Kannan Vijayan <[email protected]> wrote: > On 13-09-10 6:08 PM, Bobby Holley wrote: >> On Tue, Sep 10, 2013 at 2:25 PM, Kannan Vijayan <[email protected]> wrote: >>> I think in addition to those steps, we could also change the semantics of >>> __noSuchMethod__ such that it only gets called for lookups of non-existant >>> properties, as opposed to all lookups which return a primitive value. The >>> interpreter would be changed to implement that behaviour, and the jits >>> basically stay the same as they are now. >>> >>> How does that sound? >> That's probably fine IMO, as long as we communicate it. How much more >> painless would that make it to maintain __noSuchMethod__ for the >> medium term? >> >> We should certainly remove uses of it from the tree, and flag it with >> deprecation warnings. But our capital for guns-blazing >> break-your-addon platform changes is finite, and it's not clear to me >> that this is the most worthy cause. If we can make the feature >> reasonably painless to maintain (albeit ugly), I think we should just >> remove it from the tree and the addon-sdk, and see where we are in two >> years. There are a number of other addon-related unknowns on the >> horizon (like e10s and servo) as well. >> >> bholley > That behavioural change would make it pretty painless for the JITs. I'll > start working on a patch for that. > > How would we communicate it and what sort of lead time would we want between > communicating our intent and landing the changes? > _______________________________________________ > dev-tech-js-engine-internals mailing list > [email protected] > https://lists.mozilla.org/listinfo/dev-tech-js-engine-internals _______________________________________________ dev-tech-js-engine-internals mailing list [email protected] https://lists.mozilla.org/listinfo/dev-tech-js-engine-internals
