The committee behind ECMAScript wants a quicker release schedule to keep up with the pace of Web innovation.
InfoWorld: When are the planned arrival dates of ECMAScript 6 and 7?
Husain: The ES specification for 6 is planned to be finalized, I think, June 2015. ES 7 currently has no date of arrival, but the committee members have talked about a more regular cadence, and certainly I think you’ll see that release for ES 7 is quicker than the one for ES 6, which is a very, very large release because it was so long overdue.
InfoWorld: Why the parallel development?
InfoWorld: Would you say async programming is the main feature planned in ECMAScript 7?
I don’t think it’s fair to say that async programming is the main focus. I would like it personally to be a large focus, and some of the proposals that I’ve put forward with representatives from Mozilla do a lot to make async programming easier, but it’s very early in the stages, and there’s no reason to believe that these are going to get rubber-stamped and go through. They might be too big and have to wait for ES 8, or they might never get accepted. Currently async functions are for ES 7, and async generators I’m proposing for ES 7.
InfoWorld: Explain the difference between async functions and async generators.
Husain: A function returns a value. The thing about a function is the consumer is in control. The consumer calls a function, and the world blocks until that function returns the value. Now an iterator is, basically, think of it as a function that can return multiple values. The consumer asks for an iterator, then pulls by calling next, pulls values out. Now in a reasonable time, the consumer requests a value, again everything blocks and the world stops until that consumer gets their value. I call it synchronous programming.
InfoWorld: So it makes it easier for developers to do async programming?
Husain: You said it shorter than I did, but yes.
InfoWorld: Which important features of ES 6 and 7 have already been implemented in browsers?
Husain: Certainly, my perception is that there’s broad support for SIMD, and I think Intel has very impressive demos that show that SIMD performance can dramatically improve rendering. I think the committee is generally supportive of anything that’s going to improve the speed of the Web, and it certainly looks like SIMD is a good choice.