dosync Archive Pages Categories Tags

JavaScript Modularity Shaming

06 January 2015

Pete Hunt of React fame recently got into an online discussion about the pros and cons of Webpack vs. Browserify. In the discussion he inadvertently coins a hilarious term for a form of rhetoric in some circles of the JavaScript community - "Modularity Shaming". React itself has been on the receiving end of "Modularity Shaming" due to its relatively large size for a JavaScript dependency, ~37K gzipped.

The wellspring of "Modularity Shaming" is of course the very real need for JavaScript applications to load quickly. This means trying to decrease the payload of the application through some obvious approaches like concatenation and minification. However the very popular combination of jQuery and Underscore.js alone add up to ~40K gzipped, and of course real applications need significantly more support than that and we haven't even gotten to your application code yet!

At this point there are a couple of options - break apart the code and only use what you need when you need it through asynchronous module loading. While this works OK for library dependencies, for your application modules you'll likely as not be fighting race conditions if you don't exert some discipline. Surmountable for sure, but certainly not as simple to reason about as a single concatenated file.

The other option is to use smaller dependencies. The problem with this is that the microlib of yesterday may very well become the macrolib of tomorrow. The fully internationalized version of moment.js is as big as jQuery & Underscore.js combined!

There is an alternative solution to this problem that's rarely discussed.

Dead Code Elimination

For suitably written JavaScript, the Google Closure Compiler to this day defeats all comers in producing compact JavaScript - this is due to Dead Code Elimination. Google long ago decided they would need an incredible amount of JavaScript functionality across their products so they created the Google Closure Libary. The code base including all tests is a whopping 300,000 lines of JavaScript!

Through the magic of Dead Code Elimination if I use just one single function in this behemoth of a JavaScript library I will only get the code that's necessary for that one function to do its job.

But just how good is it in practice?

Consider the following ClojureScript program. It includes DateTime formatting, DateTime arithmetic, and internationalization:

(ns practice.core
  (:import [ DateTime Interval]
           [goog.i18n DateTimeFormat]))

(def f (DateTimeFormat. "EEEE MMMM d, y"))

(let [week (Interval. nil nil 7)
      dt   (DateTime.)]
  (.add dt week)
  (.log js/console (.format f dt)))

If I want this ClojureScript program to emit a DateTime string in French the total cost of this entire program will be ~4K gzipped. That's 10X less of a dependency than moment.js with internationalization.

Use Google Closure

If you're a ClojureScript developer I highly recommend always first checking to see if there isn't a satisfactory Google Closure module that will get the job done. Just the other day I needed a pseudorandom number generator for test.check, Closure did not disappoint.

In some cases Google Closure doesn't provide the nicest of APIs or is poorly documented but it's generally a simple matter to present a lovingly crafted facade. There's no need to promulgate the disease of NIH you encounter so often in JavaScript circles.