Investigate strict mode performance

[mikebolt]

We may be able to get a performance boost by using JavaScript's strict mode in our library. Someone should test performance with and without strict mode in different browsers. The bars example (examples/01_bars.html) could be a good way to test the performance, especially when combined with a profiling tool.

[dalisoft]

I get 3-5fps lower than "non-strict", i interested to this, because, my js based on tween.js

[mikebolt]

I got the same drop in framerate when switching both Tween.js and the script in 01_bars.html to strict mode.

However, it looks like the issue is the following code in 01_bars.html:

var updateCallback = function() {
    this.domElement.style.left = this.x + 'px';
}

This code is taking most of the CPU time and it runs slower in strict mode. I did another test by turning strict mode on only in Tween.js, and the performance was better in strict mode.

We should run some performance tests with the other examples.

[dalisoft]

@mikebolt Yes, we need some examples for performance. I interested. Contact me via e-mail, if i need in tests

[dalisoft]

@mikebolt If we can get performance that you said as improved. We can catch GSAP Performance. It's very amazing, yes?

[sole]

This is probably a good moment to remind ourselves that we actually do not have a way of measuring performance properly! I created this issue to start a discussion about it: https://github.com/tweenjs/discuss/issues/3

[sole]

@mikebolt I removed 'good first bug' because I don't think it's reasonable to expect someone to solve the 'benchmark performance' issue and 'strict mode performance' in just one go =)

[dalisoft]

You can see my note about it here

with "use strict" performance maybe improved by almost 15%, but some Android <5 has bugs with "use strict" and may not load script and tweening doesn't runs. Sorry :(. If you want use it, please just paste

(function(){
    "use strict"; // here paste

and done. You have better performance :)