incaseofstairs

ES6 Feature Performance

2015-06-26T00:25:06+00:00

With ES6 features landing quickly in native browsers and readily available for use through Babel and Traceur, it seemed like it was time to look not just at support, but also the performance impact of using these features under the current implementations.

While there is great promise for the future, the picture of ES6 feature performance today is very muddled and depends on the specific feature being used. Some such as bindings and simple arrow/destructuring are ready for use today, others such as generators and tagged template strings might require analysis before using.

The standard warnings of premature optimization and recommendations to profile your own code apply to the comments here. These tests are very micro in their scope and might not be representative of your particular use case. It’s also possible that the ES6 version is fast enough for your use case and is not worth additional time spent refactoring to a more complicated but faster implementation.

Results

The results below are a snapshot from when this post was written. kpdecker.github.io/six-speed/ has the most recent results.

Arrow Function

Arrow functions invocation has little performance impact under transpilers. Their native implementation under Firefox is 40-70x slower for calls than the equivalent ES5 operation. Internet Explorer’s performance is approximately that of ES5.

Arrow function declaration on the other hand is slightly slower than the most optimized ES5 implementation, under most environments. Here too Firefox’s implementation shows a large performance hit and IE shows a slight performance hit.

		node		chrome			firefox			internet explorer		safari
		0.10.39	2.3.0	43	44	45	38	39	40	11	12	8
arrow tests	babel	1.3x slower	Identical	Identical	Identical	Identical	Identical	Identical	Identical	Identical	1.2x faster	Identical
	traceur	1.3x slower	Identical	Identical	Identical	Identical	Identical	Identical	Identical	Identical	1.3x faster	Identical
	es5-bind	10x slower	15x slower	19x slower	14x slower	20x slower	6x slower	6x slower	7x slower	4x slower	2.6x slower	3x slower
	es6						38x slower	41x slower	52x slower		Identical
arrow-args tests	babel	Identical	Identical	Identical	Identical	Identical	Identical	Identical	Identical	Identical	1.2x faster	Identical
	traceur	Identical	Identical	Identical	Identical	Identical	Identical	Identical	Identical	Identical	1.2x faster	Identical
	es6		2.9x slower				64x slower	89x slower	68x slower		1.3x faster
arrow-declare tests	babel	1.3x slower	1.3x slower	Identical	Identical	Identical	8x slower	10x slower	13x slower	2.3x slower	1.6x slower	1.4x slower
	traceur	1.3x slower	1.3x slower	Identical	Identical	Identical	5x slower	7x slower	8x slower	2.3x slower	1.4x slower	1.4x slower
	es6		Identical				53x slower	78x slower	78x slower		1.3x slower

Issues:

[Firefox Native Implementation](https://bugzilla.mozilla.org/show_bug.cgi?id=1177518)

Classes

With classes we start to see some differences in behaviors. Traceur and the V8 native implementation operate at partity with the ES5 tests when looking at instantiation. Babel’s implementation does suffer a 1.5-60x performance hit for the operations tested. When compiling using Babel’s loose mode, the hit is lessened to 8x.

The super keyword has some fairly large performance issues under all implementations, with the best case being 3x slower and the worst case being 60x slower than the respective baselines. V8’s native implementation also sees a 15-20x performance hit.

Babel’s loose implementation of super is akin to that of the ES5 implementation, utilizing C.prototype.bar.call(this) rather than the slower getPrototypeOf lookup operation that while more accurate technically, incurs an additional cost. This is controlled by the es6.classes parameter but Babel’s authors cite a number of warnings with this flag that may impact compatibility when migrating code to native implementations.

		node		chrome			firefox			internet explorer		safari
		0.10.39	2.3.0	43	44	45	38	39	40	11	12	8
classes tests	babel	2.3x slower	1.5x slower	1.3x slower	1.4x slower	1.3x slower	27x slower	26x slower	30x slower	1.4x slower	1.5x slower	Identical
	babel-loose	2.3x slower	1.5x slower	1.3x slower	1.4x slower	1.3x slower	6x slower	6x slower	8x slower	1.3x slower	2.4x slower	Identical
	traceur	Identical	Identical	Identical	Identical	Identical	1.5x slower	1.7x slower	1.8x slower	Identical	Identical	Identical
	es6		Identical	Identical	Identical	Identical
super tests	babel	24x slower	60x slower	52x slower	48x slower	45x slower	60x slower	62x slower	61x slower	24x slower	24x slower	16x slower
	babel-loose	2.0x slower	1.7x slower	1.7x slower	1.7x slower	1.7x slower	5x slower	5x slower	5x slower	3x slower	3.0x slower	1.2x slower
	traceur	12x slower	26x slower	18x slower	18x slower	18x slower	26x slower	30x slower	27x slower	10x slower	12x slower	11x slower
	es6		14x slower	19x slower	21x slower	18x slower

Issues:

[V8](https://code.google.com/p/v8/issues/detail?id=3759)
[Firefox Class Performance](https://bugzilla.mozilla.org/show_bug.cgi?id=1167472)
[Firefox Super Property Performance](https://bugzilla.mozilla.org/show_bug.cgi?id=1169745)
[Firefox Super Call Performance](https://bugzilla.mozilla.org/show_bug.cgi?id=1169746)

Enhanced Object Literals

Object literal extensions generally provide an overhead of up to 147x the baseline. Under the transpiler implementations this is due to the use of defineProperty rather than the much more optimized field assignment. This is done to bullet proof code from potential edge cases discussed here. Loose mode is effectively the same as the ES5 implementation as of 5.6.7

		node		chrome			firefox			internet explorer		safari
		0.10.39	2.3.0	43	44	45	38	39	40	11	12	8
object-literal-ext tests	babel	4x slower	60x slower	72x slower	71x slower	80x slower	1.8x slower	1.6x slower	1.6x slower	2.6x slower	2.0x slower	2.5x slower
	babel-loose	Identical	12x slower	Identical	Identical	Identical	Identical	Identical	Identical	Identical	Identical	Identical
	traceur	8x slower	120x slower	143x slower	122x slower	144x slower	3x slower	3x slower	2.9x slower	5x slower	4x slower	4x slower
	es6		25x slower		23x slower	26x slower	1.4x slower	1.5x slower	1.3x slower		Identical

Issues:

[Babel Optimization](https://github.com/babel/babel/pull/1830)
[V8](https://code.google.com/p/v8/issues/detail?id=4246)

Template Strings

Template strings are a mixed bag. In the basic form, transpilers are able to hit parity with the baseline implementation under most environments. The native implementations are hit or miss. Under Chrome they execute at half the speed and under Firefox up to 650x slower.

Tagged template strings unfortunately do not have such a nice outlook. Their performance ranged from 2x slower for IE’s native implementation to 2000x slower for Babel’s implementation under Firefox. Babel’s loose implementation (es6.templateLiterals) lessens much of the overhead of this operation, at the cost of not having a fully compliant String.raw implementation.

Issues:

[Chrome](https://code.google.com/p/chromium/issues/detail?id=504212)
[Firefox](https://bugzilla.mozilla.org/show_bug.cgi?id=1177318)
[Babel](https://github.com/babel/babel/issues/971#issuecomment-115344157)

		node		chrome			firefox			internet explorer		safari
		0.10.39	2.3.0	43	44	45	38	39	40	11	12	8
template_string tests	babel	Identical	Identical	Identical	Identical	Identical	Identical	Identical	Identical	1.5x faster	1.4x faster	Identical
	traceur	Identical	Identical	Identical	Identical	Identical	Identical	Identical	Identical	Identical	Identical	Identical
	es6		Identical	2.0x slower	2.0x slower	1.9x slower	627x slower	622x slower	591x slower		1.3x faster
template_string_tag tests	babel	567x slower	578x slower	722x slower	755x slower	820x slower	2395x slower	2349x slower	2028x slower	90x slower	82x slower	61x slower
	babel-loose	2.0x slower	1.6x slower	1.7x slower	1.6x slower	1.8x slower	94x slower	87x slower	84x slower	1.4x slower	1.4x slower	2.0x slower
	traceur	7x slower	13x slower	17x slower	13x slower	16x slower	346x slower	307x slower	261x slower	13x slower	11x slower	8x slower
	es6		8x slower	9x slower	13x slower	9x slower	68x slower	64x slower	59x slower		Identical

Destructuring

For destructuring, the average use case effectively matches that of the ES5 counterpart. Unfortunately complex use cases, particularly those around array destructuring, often have large performance overhead. Under Babel an unoptimized helper is used to access the data and under Traceur an entire iterator structure is created, both of which provide fairly substantial memory and CPU overhead over the simple array accessor logic that hand coded ES5 can utilize. Loose mode is effectively the same as the ES5 implementation.

		node		chrome			firefox			internet explorer		safari
		0.10.39	2.3.0	43	44	45	38	39	40	11	12	8
destructuring tests	babel	1.8x slower	1.4x slower	1.4x slower	1.2x slower	1.3x slower	24x slower	22x slower	26x slower	3x slower	5x slower	2.2x slower
	babel-loose	1.2x faster	Identical	Identical	1.2x faster	Identical	Identical	Identical	Identical	Identical	Identical	Identical
	traceur	26x slower	13x slower	12x slower	10x slower	11x slower	163x slower	152x slower	176x slower	57x slower	25x slower	8x slower
	es6						170x slower	148x slower	185x slower			Identical
destructuring-simple tests	babel	Identical	Identical	Identical	Identical	Identical	Identical	Identical	Identical	10x slower	1.2x slower	Identical
	traceur	Identical	Identical	Identical	Identical	Identical	Identical	Identical	Identical	Identical	1.3x slower	Identical
	es6						Identical	Identical	Identical			Identical

Issues:

[Firefox](https://bugzilla.mozilla.org/show_bug.cgi?id=1177319)
[Babel Strict Optimization](https://github.com/babel/babel/pull/1821)

Default Parameters

Default parameters were universally slower for all transpiler implementations. They effectively compile to the same thing, utilizing the arguments object to set a local variable vs. using a named paramemter. This appears to be unoptimzed under all engines and consequently performance was 4-2000x slower. Sadly, this is required in order to properly implement the fn.length behavior defined by the spec:

NOTE The ExpectedArgumentCount of a FormalParameterList is the number of FormalParameters to the left of either the rest parameter or the first FormalParameter with an Initializer. A FormalParameter without an initializer is allowed after the first parameter with an initializer but such parameters are considered to be optional with undefined as their default value.

These scaled numbers should be taken in context. The ES5 equivalents are highly optimized, Firefox pushing over 833 million operations a second in one test, so the net performance of the transpiled versions may very well be sufficient for most use cases, particularly those not on the hot path.

The only native implementation, Firefox, performed identically to the ES5 implementation.

		node		chrome			firefox			internet explorer		safari
		0.10.39	2.3.0	43	44	45	38	39	40	11	12	8
classes tests	babel	2.3x slower	1.5x slower	1.3x slower	1.4x slower	1.3x slower	27x slower	26x slower	30x slower	1.4x slower	1.5x slower	Identical
	babel-loose	2.3x slower	1.5x slower	1.3x slower	1.4x slower	1.3x slower	6x slower	6x slower	8x slower	1.3x slower	2.4x slower	Identical
	traceur	Identical	Identical	Identical	Identical	Identical	1.5x slower	1.7x slower	1.8x slower	Identical	Identical	Identical
	es6		Identical	Identical	Identical	Identical
defaults tests	babel	17x slower	11x slower	11x slower	9x slower	8x slower	1842x slower	2051x slower	2043x slower	229x slower	72x slower	4x slower
	traceur	16x slower	12x slower	12x slower	10x slower	9x slower	1759x slower	2305x slower	1974x slower	210x slower	73x slower	4x slower
	es6						Identical	Identical	Identical

Rest Parameters

Rest parameters are as fast or faster than the ES5 equivalent under almost all implementations. Native implementations provided a performance boost up to 40x (with the exception of V8 where it causes a known deoptimization). Use them with a transpiler. They’re great. Death to arguments.

		node		chrome			firefox			internet explorer		safari
		0.10.39	2.3.0	43	44	45	38	39	40	11	12	8
rest tests	babel	Identical	Identical	Identical	Identical	Identical	Identical	Identical	Identical	2.0x slower	1.5x slower	Identical
	traceur	Identical	1.5x faster	1.2x faster	1.2x faster	1.2x faster	31x faster	34x faster	27x faster	2.4x slower	1.6x slower	1.3x faster
	es6		3x slower				40x faster	42x faster	32x faster		8x faster

Spread Parameters

Under Babel, spread parameters for arrays perform identically to the ES5 counterpart as they are effectively both an apply call. Under Traceur and all of the native implementations the implementations are 1.3x to 17x slower.

		node		chrome			firefox			internet explorer		safari
		0.10.39	2.3.0	43	44	45	38	39	40	11	12	8
spread tests	babel	Identical	Identical	Identical	Identical	Identical	Identical	Identical	Identical	Identical	Identical	Identical
	traceur	13x slower	6x slower	5x slower	6x slower	6x slower	2.7x slower	3x slower	3x slower	9x slower	712x slower	17x slower
	es6						3x slower	4x slower	4x slower		1.3x slower	2.5x slower
spread-generator tests	babel	150x slower	55x slower	58x slower	66x slower	71x slower	153x slower	143x slower	157x slower	656x slower	423x slower	64x slower
	babel-loose	114x slower	46x slower	41x slower	45x slower	53x slower	60x slower	60x slower	61x slower	28x slower	51x slower	22x slower
	traceur	27x slower	10x slower	9x slower	10x slower	13x slower	11x slower	10x slower	11x slower	15x slower	1551x slower	9x slower
	es6						7x slower	8x slower	7x slower
spread-literal tests	babel	Identical	Identical	Identical	Identical	Identical	1.2x slower	Identical	1.2x slower	3x slower	1.7x slower	2.0x slower
	traceur	6x slower	3.0x slower	2.1x slower	2.5x slower	2.3x slower	6x slower	5x slower	5x slower	26x slower	451x slower	9x slower
	es6						8x slower	7x slower	8x slower		1.8x slower	3x slower

Let + Const

let and const bindings were pretty much identical across the board. While these don’t offer performance improvements (yet), they shouldn’t negatively impact performance.

		node		chrome			firefox			internet explorer		safari
		0.10.39	2.3.0	43	44	45	38	39	40	11	12	8
bindings tests	babel	Identical	Identical	Identical	Identical	Identical	Identical	Identical	Identical	Identical	Identical	Identical
	traceur	Identical	Identical	Identical	Identical	Identical	Identical	Identical	Identical	Identical	1.3x faster	Identical
	es6	Identical	Identical	Identical	Identical	Identical	2.3x slower	Identical	Identical	1.3x slower	Identical

For..of

for..of is universally slower, ranging from 3 to 20x slower for array iteration over classical array iteration. When iterating over an object with a custom iterator, the performance is also much slower than for..in iteration with hasOwnProperty checks.

		node		chrome			firefox			internet explorer		safari
		0.10.39	2.3.0	43	44	45	38	39	40	11	12	8
for-of-array tests	babel	21x slower	8x slower	8x slower	9x slower	9x slower	14x slower	20x slower	19x slower	15x slower	489x slower	7x slower
	babel-loose	Identical	Identical	Identical	Identical	Identical	1.4x slower	1.7x slower	1.7x slower	1.3x slower	Identical	1.8x slower
	traceur	12x slower	7x slower	7x slower	8x slower	8x slower	26x slower	31x slower	32x slower	9x slower	6x slower	6x slower
	es6		6x slower	6x slower	8x slower	8x slower	5x slower	7x slower	7x slower		2.9x slower	4x slower
for-of-object tests	babel	10x slower	7x slower	8x slower	9x slower	6x slower	6x slower	6x slower	5x slower	60x slower	11x slower	6x slower
	babel-loose	10x slower	6x slower	7x slower	9x slower	7x slower	4x slower	4x slower	4x slower	6x slower	449x slower	6x slower
	traceur	8x slower	7x slower	8x slower	9x slower	7x slower	9x slower	10x slower	9x slower	5x slower	4x slower	4x slower
	es6		6x slower	7x slower	8x slower	6x slower	3x slower	3x slower	3x slower		3x slower

Generators

Much like for..of, generators are also quite a bit slower than a raw ES5 implementation of the iterable protocol, with performance ranging from 10x to 750x slower. There is hope here as the V8 implementation achieves parity with the ES5 implementation.

		node		chrome			firefox			internet explorer		safari
		0.10.39	2.3.0	43	44	45	38	39	40	11	12	8
generator tests	babel	601x slower	65x slower	67x slower	63x slower	69x slower	754x slower	723x slower	716x slower	79x slower	499x slower	78x slower
	traceur	77x slower	10x slower	10x slower	10x slower	12x slower	48x slower	50x slower	46x slower	18x slower	112x slower	10x slower
	es6		Identical	1.2x slower	Identical	Identical	22x slower	20x slower	19x slower

Issues:

[Regenerator Init Optimization](https://github.com/facebook/regenerator/pull/207)
[Regenerator Exec Optimization](https://github.com/facebook/regenerator/pull/208)

Maps and Sets

Map and Set all have insert performance that is about 10x slower for a moderately sized data set. All of the implementations show massive improvement on the lookup operations, with Firefox’s native implementation showing a 200x speed increase for a dataset of size 500.

Traceur appears to delegate to the native implementation via their polyfill where possible so performance is closely linked to improvements in the native layer. In runtime mode, Babel does not appear to delegate and performance suffers as a result. Babel’s polyfill mode should behave as Traceur does but this was not directly tested.

Take caution with these numbers. The tests use a data set of size 500 and other data sets will have varying performance but it appears that these features are ready for general use if you have many reads and few writes or need to have objects as keys.

		node		chrome			firefox			internet explorer		safari
		0.10.39	2.3.0	43	44	45	38	39	40	11	12	8
map-set tests	babel	13x slower	3.0x slower	5x slower	5x slower	3.0x slower	105x slower	12x slower	15x slower	60x slower	19x slower	79x slower
	traceur	5x slower	Identical	2.2x slower	2.3x slower	1.6x slower	20x slower	2.5x slower	2.4x slower	28x slower	2.2x slower	16x slower
	es6	5x slower	Identical	2.2x slower	2.3x slower	1.6x slower	21x slower	2.2x slower	2.7x slower	28x slower	2.2x slower	15x slower

Promises

Promises are across the board faster with both the polyfill and native implementations. This particular benchmark is dubious as it’s both async and inheriently tied to long running behaviors where execution overhead has little impact.

		node		chrome			firefox			internet explorer		safari
		0.10.39	2.3.0	43	44	45	38	39	40	11	12	8
promises tests	babel	Identical	Identical	6x faster	7x faster	7x faster	21x faster	17x faster	16x faster	Identical	1.2x slower	3x faster
	traceur	Identical	Identical	2.0x faster	2.0x faster	1.9x faster	36x faster	31x faster	Identical	Identical	1.8x slower	3x faster
	es6		Identical	2.0x faster	2.0x faster	1.9x faster	37x faster	29x faster	Identical		1.9x slower	3x faster

Testing methodology

For each of the ES6 features in question, a ES5 implementation of that functionality was written along with a ES6 version. It should be noted that the functionality is frequently the same, but in some cases the “common” vs. “correct” version was written, i.e. using x[key] = value vs. defineProperty which is faster but can be hit but a particular nasty edge case for those who deem it fun to extend Object.prototype.

Babel, in both loose+runtime and runtime mode, and Traceur were then used to compile the ES6 version to a ES5 compliant version, utilizing the runtime over polyfill to maintain test isolation and avoid native implementations where possible.

All of these test instances were then benchmarked in the given JavaScript engine using Benchmark.js and then the operations per second compared to the ES5 implementation. Cross browser and cross execution comparisions are avoided as much as possible to isolate environmental issues when executing on VMs in the cloud.

All of this data, including any updates from more recent test runs is available at http://kpdecker.github.io/six-speed/ and the test suite is available at https://github.com/kpdecker/six-speed for review/feedback.

Takeaways

As noted above, these results might not be representative of your own application since they only test very small subsets of inputs and behaviors of these new features. If you are finding that you have performance issues with your code using these features, you should test them within your own environment to see what the actual behavior is.

While some of these features are a bit slow as of this writing, their performance should only improve as the native implementations mature and are optimized as real world use is applied to them. For the transpilers, there are some performance optimzations that can be made but much of the overhead they are experiencing is due to spec compliance, which they go to great lengths to achieve, but this comes with unfortunate overhead under ES5 implementations as they stand. Babel’s loose mode does offer a bit of a performance boost, but care must be taken when using loose mode as this could cause breakages when code is migrated to the standard native implementations.

Personally I intend to start using most of these features where they make sense but will avoid designing core APIs around these features (perhaps with the exception of Promises) until the native implementations have matured a bit.

May Conference Recap

2013-06-03T00:00:00+00:00

I had the pleasure of attending portions of Google IO, CSS Conf, and JS Conf and wanted to summarize my notes on these conferences. These are my own take aways and may not have been what the presenter intended. If I’m completely off base, please feel free to drop me a line. This is also not inclusive as there are only so many sessions that one can attend :)

Major Trends

Componentization

Biggest theme of all of these conferences was attempting to create ways of compoentizing our code so that we can leave the smell of the poo all in one place, to paraphrase @danwrong.

Polymer

Google has the Polymer framework which polyfils web components. The premise is that you can create truly isolated components that are drop in and do not have worries such as CSS impact from other components.

<element name="my-element">
  <section>
    I'm a my-element!
  </section>
  <footer>nothing to see here</footer>
  <script>
    // When <element> is in document, might run in wrong context.
    // Only do work when this == <element>.
    if (this !== window) {
      var section = this.querySelector('section');
      this.register({
        prototype: {
          readyCallback: function() {
            this.innerHTML = section.innerHTML;
          }
        }
     });
    }
  </script>
</element>

This looks pretty amazing and I think will provide some awesome separation of concerns between content and behaviors in the future. Polyfilling this behavior seems outright genius but at the same time scares me as you are basing a lot of your implementation on something that is being hacked into the browser.

I’m of the opinion that this is amazing future looking work but is a bit more risk than I’d like to take on for my user base.

With standards support for this moving forward I think this is where we will be in 5-10 years depending on the user base that a project is required to support. I also believe that there is a huge opportunity here for developer tooling as this changes some things up that will require changes and improvements to our existing hint/test/minimize/etc tool stacks.

React

Facebook is doing similar work with their React framework. They have created a new compiled to JavaScript language that allows one to embed html into the source file:

var HelloMessage = React.createClass({
  render: function() {
    return <div>{'Hello ' + this.props.name}</div>;
  }
});

The framework will then manage the all of the rendering lifecycle without data binding or blanket re-rendering. This is apparently done by a DOM diffing algorithm that walks the DOM tree and updates only the components that need to be on a property change. The presenting team highlighted that you can provide hints to this implementation to improve performance so I do question how well this scales.

I have not had a chance to play around with this implementation but I am not against the html embedded in a JavaScript file like many people in the blogosphere are/have been. I think both this and polymer tackle separation of concerns in a different manner than the HTML/CSS/JS paradigm that has been beat into web developers for so long. I see them as implementing separation of concerns at a content vs. behavior level which may be better for some workflows.

Flight

Flight from Twitter attempts to handle componetization at a much smaller level. They focus on defining interfaces between discrete components. Each of these components can focus on only their implementation.

This is lower-level than the systems above and in many ways can be used to build other framework systems.

Performance

Performance was a huge concern at all 3 conferences with many sessions devoted to different aspects of performance improvements. My major take aways are:

Don’t follow a coding pattern because something is “faster”.

What exactly is faster will change as runtimes optimize for the code that they are seeing. Instead look at what the tools are saying and react accordingly.
Don’t prematurely optimize.

You are probably wrong. Your code is most likely going to be harder to understand and there is plenty that you can spend time as engineering time is finite.
Performance is holistic.
- JavasScript, layout, rendering, compositing, GC, etc. can all impact performance.
- Newer architectures are moving behaviors out to other threads but this does not work if you try to implement your own behavior. Custom scrolling is the biggest offender here.
- Under current Chrome implementations touch handlers can cause scrolling issues as it disables certain scrolling optimizations.
Chrome is the leader in debugging tools and ability to investigate performance
- Chrome dev tools timeline panel
- about:tracing - Google’s internal performance tool
- V8 tools - In depth V8 profiling tools
- Continuous painting mode lets you test performance impact of specific elements or properties
- Remote debugging API (This powers Chrome for Android debugging and topcoat performance monitor)
Higher level console commands are useful
- console.time
- console.group
- These are not always evil and can help programmers with onboarding
Topcoat has a turbo mode which attempts to remove layout and style calculation from overhead.

All elements have style element and are absolutely positioned. On a personal note this seems a bit overboard for 99% of use cases.
High-speed cameras can be used in the name of performance testing! :)

Many V8 specific performance recommendations as well:

V8 has an optimized/unoptimized state machine that a particular piece of code may jump between.

Can’t tell at compile time if something will be promoted to optimized once hot. When profiling should look for * token next to function names to verify that it’s optimized.
Memory allocation pattern between optimized and unoptimized are dramatically different. Optimized will often remove temp variables wereas unoptimized will have to GC temp variables.
Can monitor state machine with --js-flags="--trace-deopt --trace-opt-verbose". WARN: Very verbose. Pipe & Grep

Better Tooling

Adobe and Google had strong developer tools showings at these conferences.

Adobe demoed some very cool features for their brackets IDE including:

PSDLens - PSD inspector assistant
Response - Responsive design editor
Theseus - Brackets debugger concept
Instabug - Live code inspector with visual unit test executor/generator

Additionally they covered some of the tools involved with their topcoat project:

On the Google front, most of the improvements were in the Chrome Developer Tools including:

Flame charts
Full remote dev tools on Chrome For Android
SASS support, generically via source maps
Live editing with mapped update support

JavasScript everywhere

One of the other fun themes was the use of JavaScript in more and more places. This included a hacking day on various bots and AR Drones for those who choose those sessions. This culminated with a Nodebot shooting down a NodeCopter and calls for Node-based Aircraft Carriers.

Random Coolness

jankfree.org
nodesecurity.io
webplatform.org
@rem demoing game with RTC and getUserMedia
@seb_ly giving an amazing presentation including crowd participation froaway and live coding over VNC
@necolas - Get over yourself. Tech changes every few years and all otoday will likely be moot in a few years.

Outside of the above there we’re a lot of fun and interesting sessions that could be discussed in depth but suffice to say that the JavasScript and Web Development communities are very strong and there is a lot of interesting stuff coming down the pipes.

OSX Frontend Toolchain

2013-01-30T07:08:10+00:00

I’m currently in the middle of inventorying my machine so I can retire the time machine image that has spanned far too many years and far too many machines. Rather than stuffing this in Evernote to be used once I figured I’d post this to the community.

What other frontend tools are missing? Which ones have better alternatives?

Editors

SublimeText 2

Common plugins include:

Package Control - Allows all other plugins to be installed
BracketHighlighter
Git
Handlebars.tmBundle
JsFormat
JsHint
PrettyJSON
Stylus
SublimeSaveOnBuild
TrailingSpaces

Useful config options:

{
  "bold_folder_labels": true,
  "draw_white_space": "all",
  "ensure_newline_at_eof_on_save": true,
  "file_exclude_patterns":
  [
    "*.tmproj",
    "*.sublime-workspace",
    "*.class",
    ".DS_Store"
  ],
  "folder_exclude_patterns":
  [
    ".svn",
    ".git",
    ".hg",
    "CVS",
    "node_modules"
  ],
  "highlight_line": true,
  "highlight_modified_tabs": true,
  "rulers":
  [
    100
  ],
  "tab_size": 2,
  "translate_tabs_to_spaces": true
}

SCM

Git Command Line

Your mileage may vary but by far the strongest git client out there.

Most common git commands

git checkout -b $newBranchName
git stash
git pull --rebase (Warn do not do after a branch merge)

Ensure that your version of git has completions enabled (this might require manual installation)

Customizing your bash prompt for git status is also very helpful:

This script whose origin has been lost will output the current branch and change status in your bash profile.

function parse_git_branch {
  git branch --no-color 2> /dev/null | sed -e '/^[^*]/d' -e 's/* \(.*\)/(\1)/'
}
function last_two_dirs {
  pwd |rev| awk -F / '{print $1,$2}' | rev | sed s_\ _/_
}

c_cyan=`tput setaf 6`
c_red=`tput setaf 1`
c_green=`tput setaf 2`
c_sgr0=`tput sgr0`

function proml {
  PS1='\h:$(last_two_dirs)\[$(branch_color)\]$(parse_git_branch)\[${c_sgr0}\] \u\$ '
}
function branch_color() {
  if git rev-parse --git-dir >/dev/null 2>&1
  then
    color=""
    git diff --quiet 2>/dev/null >&2
    if [[ $? -eq 0 ]]
    then
      color=${c_cyan}
    else
      color=${c_red}
    fi
  else
    return 0
  fi
  echo -ne $color
}

proml

Gitx

GUI frontend on various commit related git operations. The staging view is very valuable for doing interactive commits and self-code review at commit time.
SourceTree

General git repository viewer. Helpful for viewing stashed changes that may have accumulated on your local tree.
git-extras

A collection of useful tools for doing higher level git commands such as delete-branch. When installing make sure that the command line tab completion tools to no error out as these are very helpful.

Debugging and Optimizing

Charles

HTTP proxy and debugger with easy to use interface.
xScope

Inspect/measure mocks and final rendered pages.
Gradient Scanner

(Shameless plug) Extract CSS gradients from flatted mock images.
ImageAlpha/ImageOptim

Image optimization tools. Performs both lossless and lossy compression of image assets.

Documents

Dropbox

Easy way to backup remotely and share content across the team.
Evernote

Free notetaking service.
MS Office

Because somethings you can’t escape.
MS Remote Desktop Client

Again somethings you can not escape.

Communication

Adium

Chat over many protocols. Most common among the team are GTalk and AIM.
Propane

Team chat. Setup growl notifications and stay connected.
Skype

Sometimes you need to use your voice. Horrible at IM.
ScreenFlow

Generate screen casts for sharing content with coworkers or debugging transient behaviors.

System

Alfred

Search-based system launcher.
Growl

Common notification system for Lumbar builds and Propane and others.
Divvy

Window position manager
Stay

Utility which will restore your windows when you connect or disconnect a monitor.
BetterTouchTool

Keyboard/mouse/touchpad gesture mapper.
Cloud App

Dirt simple screenshot sharing app. Useful for sharing screenshots and other local content over IM and email when attachments might not be supported or supported easily.
Homebrew

OSS package manager.
nvm

Node version switcher. Easily allows for switching between different versions of node.

Current recommendations for the latest 0.8 node branch for running the development build stack as the watch issues on the 0.8 branch seem to have been resolved.

Lumbar: Modular Javascript Build

2012-01-17T18:01:22+00:00

Since joining @WalmartLab’s mobile web team close to a year ago we have been very busy rebuilding the mobile web platform using Backbone, Handlebars, Stylus and other awesome open source tools and frameworks

Even though these tools allow for some pretty impressive feats of javascript, we kept hitting inefficiencies both in execution and developer time due to resource loading and management.

When dealing with handlebars for example, we consistently found that we were trying to fight against the system to manage the templates used by a view. We neither wanted to manually inline the template in the javascript as we all know that escaping a language within a language is just painful, ala Java running SQL, nor did we want to add the additional overhead of an additional request for deferred loading of templates or worse inlining all application templates within the html file.

Being both constrained by the mobile environment and needing to operate at the scale of Walmart it became apparent very quickly that we needed to have some sort of build-time utility to manage this so the developer can focus on creating the application logic, throw it into the build tool and out comes a nicely optimized module of code and styles.

Lumbar along with the associated Lumbar projects such as lumbar-loader and lumbar-long-expires are the fruits of these efforts.

Overview

The Lumbar suite focuses on the management of all resources required to generate complex javascript applications across many different platforms and environments. It’s primary goals are to easily package, optimize, link to resources such as javascript, html, templates, styles, images, and any other client-side resources that web application may require.

From it’s initial implementation, Lumbar has focused on making client performance the number one goal, to this end it allows for chunking resources into distinct code modules, easy minification, inlining, and cache management.

Configuration

Lumbar is configured through a simple JSON config file, defining modules which are simple, demand loaded, segments of the application.

A base module that loads the core application:

{
  "base": {
    "scripts": [
      {"src": "js/lib/zepto.js", "global": true},
      {"src": "js/lib/underscore.js", "global": true},
      {"src": "js/lib/backbone.js", "global": true},
      {"src": "js/lib/handlebars.js", "global": true},
      {"src": "js/lib/thorax.js", "global": true},
      {"src": "js/lib/script.js", "global": true},
      {"src": "js/lib/lumbar-loader.js", "platform": "web"},
      {"src": "js/lib/lumbar-loader-standard.js", "platform": "web"},
      {"src": "js/lib/lumbar-loader-backbone.js", "platform": "web"},
      "js/init.js",
      {"src": "js/bridge.js", "platforms": ["iphone", "ipad", "android"]},
      {"src": "js/bridge-android.js", "platform": "android"},
      {"src": "js/bridge-ios.js", "platforms": ["ipad","iphone"]},
      {"module-map": true}
    ],
    "styles": [
      "styles/base.styl",
      {"src": "styles/iphone.styl", "platform": "iphone"},
      {"src": "styles/android.styl", "platform": "android"},
      {"src": "styles/ipad.styl", "platform": "ipad"},
      {"src": "styles/web.styl", "platform": "web"}
    ],
    "static": [
      {"src": "static/#{platform}/index.html", "dest": "index.html"}
    ]
  }
}

A router-linked module. When using backbone/thorax integration, this module will automatically load when the # or #hello routes are navigated:

{
  "hello-world": {
    "routes": {
      "": "index",
      "hello": "index"
    },
    "scripts": [
      "js/views/hello-world",
      "js/routers/hello-world.js"
    ],
    "styles": [
      "styles/hello-world.styl"
    ]
  }
}

This config file also defines the platforms that lumbar will generate, allowing for customizing the modules for different environments ("platforms": [ "android", "iphone", "ipad", "web" ]). At @WalmartLabs we utilize this to serve customized experiences for the native clients utilizing the same codebase.

Single stylus codebase styled for iPhone, Android, and mobile web platforms.

More dramatic customizations of the same codebase utilizing platform conditional code.

The final component of the configuration is the packages component, which allows for bundling specific modules into singular responses when a particular use case only uses a subset of modules, such as the native checkout implementations utilized by @WalmartLabs.

"packages": {
  "web": {
    "platforms": [ "web" ],
    "combine": false
  },
  "native-hello-world": {
    "platforms": [ "android", "iphone", "ipad" ],
    "modules": [ "base", "hello-world" ],
    "combine": true
  }
},

Package declarations defining normal loading for web and combined for native platforms.

On building this will generate the following structure:

$ find . -type f
./android/index.html
./android/native-hello-world.css
./android/native-hello-world.js
./android/native-hello-world@1.5x.css
./ipad/index.html
./ipad/native-hello-world.css
./ipad/native-hello-world.js
./iphone/index.html
./iphone/native-hello-world.css
./iphone/native-hello-world.js
./iphone/native-hello-world@2x.css
./web/base.css
./web/base.js
./web/base@2x.css
./web/hello-world.css
./web/hello-world.js
./web/hello-world@2x.css
./web/index.html

Which can be deployed as static resources to any web server, served via a CDN, or distributed by many other means.

While not a core feature, the lumbar-long-expires plugin takes this a step further by allowing resources to include expires tokens in their names automatically. When enabled the above may generate content like the following, allowing for the application resources to be served with extended Expires headers.

$ find . -type f
./android/cb188f8/native-hello-world.css
./android/cb188f8/native-hello-world.js
./android/cb188f8/native-hello-world@1.5x.css
./android/index.html
./ipad/cb188f8/native-hello-world.css
./ipad/cb188f8/native-hello-world.js
./ipad/index.html
./iphone/cb188f8/native-hello-world.css
./iphone/cb188f8/native-hello-world.js
./iphone/cb188f8/native-hello-world@2x.css
./iphone/index.html
./web/cb188f8/base.css
./web/cb188f8/base.js
./web/cb188f8/base@2x.css
./web/cb188f8/hello-world.css
./web/cb188f8/hello-world.js
./web/cb188f8/hello-world@2x.css
./web/index.html

The @WalmartLabs mobile team believes strongly in open source software which is why in addition to open sourcing Lumbar we are also open sourcing our Thorax framework. If you are a backbone user looking to ease common tasks such as data binding and linking
data operations to a particular route I urge you to take a look. It even integrates with Lumbar!

Quick: Firediff and Firefocus updated for 5.0 support

2011-07-07T02:05:15+00:00

Both Firediff and Firefocus have been updated to increase the target application support up to Firefox 5.0.

Both of these are available from the update stream on this site. New installations are recommended to use the Firediff and Firefocus projects on AMO as these will be more responsive to the march of Firefox releases.

Firediff 1.1.2

2011-03-22T04:21:57+00:00

Just in time for the impending Firefox 4.0 release, Firediff has been updated to version 1.1.2. This release contains compatibility changes for Firefox 4.0 and Firebug 1.7 as well as the ability to sort CSS properties alphabetically

As always, the latest version is available here and is currently under review at AMO.

Firediff 1.1.1

2010-12-23T04:24:35+00:00

Firediff has been updated to 1.1.1 on the incaseofstairs servers. This version is a minor bug release that fixes the Save Diff and Snapshot functionality.

Updated package is available here and is under review at AMO

Thanks to Borris and Joel for letting me know about this issue. Feel free to contact me with any other issues you may have found.

Firebug Extension Updates: Firebug 1.6

2010-12-03T00:00:00+00:00

As many of you are already aware, the recent release of Firebug 1.6 introduced some API changes that caused breakages within Firediff. To fix these some minor code changes within Firediff were necessary, thus the 1.1 release that has just been published.

Along with the updates for the new APIs, this version also includes:

Copy support
Bug fixes for Firebug detached mode
Change monitor context menu fixes
Slight adjustment to the colors used within the text diffs

There are a few support changes in this release. Rather than trying to support multiple versions of both Firebug and Gecko, official testing is limited to a single version of Firebug, 1.6, and as a consequence for Firebug 1.6 to a single version of Firefox, 3.6. This is simply due to the increased QA time that is required to test all of the permutations of Firebug version, Gecko version, and OS type. When updating for the changed APIs in Firebug 1.6 attempts were made to maintain compatibility with Firebug 1.5, but this configuration is no longer officially tested.

Since I was going to through the release process for Firediff I decided to release the 1.2 version of Firefocus as well. This version implements the uk-UA locale but is otherwise the same as 1.1.

Firediff 1.1 is available here on incaseofstairs and is currently under review on AMO.
Firefocus 1.2 is available here on incaseofstairs and is currently under review on AMO.

Thanks to everyone who notified me of the breakage when 1.6 was released. It’s great to hear feedback from users and to hear that the project is in active use! As always please do not hesitate to let me know how the project is working for you, good or bad.

Cross Platform Text-Indexer

2010-11-25T19:05:29+00:00

While working on a recent Facebook release we ran into a performance bottleneck within the Mojo.Format.runTextIndexer API. Performance on device was on the order of a couple of seconds to process the content for feed items 30-50. This combined with the other tasks in the formatting and rendering cycle led to poor performance in the news stream.

For those who are not familiar with the Mojo.Format.runTextIndexer API, this API scans human generated text for URLs, phone numbers and email addresses, replacing the plain text representation with HTML links to the particular object. Additionally this API will replace emoticons with image representations on supporting platforms.

With the news stream scene being the primary scene in the application it was vital that we make this as performant as possible so I spent some time profiling this and determined that the bulk of the time in this operation was spent in the generic C indexer logic. Since at it’s core this is just a text scanning heuristic that, for the most part, does not rely on any C-level constructs, I decided to investigate a Javascript implementation to see if this would be more performant.

The result was a Javascript implementation that was many magnitudes faster than the core implementation (multiple seconds vs. ~10ms for a variety of Facebook news feeds), allowing for much faster execution as well as the creation of the library being released today.

While this library is designed primarily for webOS applications, it has also been designed to work in cross-platform environments for all functionality other than the emoticon replacement which requires platform specific logic to determine the proper image to replace.

Implementation

There are two key components to this implementation, the cross platform link indexer and the platform specific emoticon processor.

The link indexer is a two-stage regular expression, using the quite daunting INDEXER_REGEX regex to extract possible linker tokens with Javascript being used to determine the meaning of each token (or throwing it out just text content).

The emoticon processor is implemented by scanning for known, or possibly known, emoticons using the EMOTICON_REGEX regex and then passing these tokens to the platform’s text indexer implementation. Doing this allows for our own custom implementation while retaining the look and feel of emoticons used by the rest of the platform. For non-Mojo platforms this logic is disabled and emoticons will be left unmodified, with minimal change.

Warnings

HTML Content

Like the API this aims to replace, this API does not handle HTML context when replacing content. As a consequence this algorithm can break HTML content. In order to prevent this, the input and outputs for the TextIndxer.run API should be considered text or minimal HTML that will not match any of the replacements.

It is still possible to use this with content containing HTML by processing in a manner similar to the following although the input still needs to be filtered for XSS and other security concerns.

var srcText = $(this).val(),
    womb = $('<div>' + srcText + '<div>');

womb.contents()
    .filter(function() {
        return this.nodeType === Node.TEXT_NODE;
    })
    .each(function() {
        var text = $(this),
            indexedText = TextIndexer.run(text.text()),
            womb = $('<div>' + indexedText + '<div>');
        womb.contents().each(function() {
            text.before(this);
        });

        text.remove();
    });

$('#previewContent').html(womb.html());

Framework Override

By default this library overrides the Mojo.Format.runTextIndexer API (when used within the Mojo framework). Care should be taken with each OS upgrade to ensure that this override does not break any expected behavior. If uncomfortable with this override then it can be removed by removing these lines from the library and using the TextIndexer.run API directly.

// Mojo Framework override. Unused on non-Mojo platforms and may be removed if undesired in Mojo apps
if (window.Mojo &amp;&amp; Mojo.Format) {
    // Override the Mojo API if it exists in this context.
    Mojo.Format.runTextIndexer = TextIndexer.run;
}

Code

The code is available on Github within the text-indexer repository. A cross-platform live demo is also available here.

jsPerf

2010-08-01T22:06:35+00:00

Recently heard about the relatively new jsPerf service that Mathias Bynens has created and after spending some time playing around with this on my own machine I can honestly say that I love this service. Rather than a few gushing tweets on the subject, I felt like this deserves a blog entry covering it’s coolness.

The thing that I found the most striking about the site (and most likely the reason that I am writing this now) is that it broke many of my previous assumptions about performance of specific statements and relative performance of different platforms. Is innerHTML faster than DOM manipulation when generating a DOM tree? Not for a specific (relatively simple) construct under Chrome (OS X v5). Is array.join(‘’) the fastest way to concat a string. On my test platforms (OS X Firefox, Safari, Opera, Chrome release versions), nope.

Being a public playground, the community benefits from the collective knowledge. Scanning through the test list I saw many of the problems that I have run into before as well as some I had never considered such as the multiple methods of performing a Math.floor operation. While some of these honestly scare me and remind me of Kernel code, it’s good to know they exist.

jsPerf has full support for mobile browsers as well, meaning that you can test performance test the various algorithm options on device and have a realistic view of how it will behave on the device. In running some of the tests on a webOS device there were some quite distinct differences between comparable environments on a desktop class machine.

As with any performance testing I would recommend running these tests multiple times on your target device(s) before selecting a particular path to use as there are many things that could cause a single test to provide inaccurate results for a given execution, particularly on the more resource constrained environments.

Great job Mathias and can’t wait to see what features are added to this as well as the content that the community creates!