JSFoo 2013

All about being creative with JavaScript

Frontend Dev Ops

Submitted by Aditya (@netroy) on Friday, 23 August 2013

Section: Full talk Technical level: Intermediate Status: Confirmed & Scheduled

View proposal in schedule


As the browsers are becoming more powerful & Webapps more complex, the way we deploy them is changing too.


I'll try to cover a few topics related to building & deploying rich webapps to production & automating as much as possible.

  • Automated Asset management (styles & templates)
  • Automated Localization
  • Automated Sprites & Icon-fonts
  • Linting & CI
  • Building using tools like grunt
  • Automated Continuous deployment
  • Caching
  • Network performance
  • Error tracking
  • Performance tracking


  • Experience building rich webapps.
  • Interest in deploying desktop-grade apps

Speaker bio

Aditya is a full-stack Developer at 6Wunderkinder GmbH.
He also works as the Front-end dev-ops guy for wunderlist.com





  •   Ankur Agarwal (@devilankur18) 6 years ago

    Is there anything more than the basic grunt / yeoman tasks ?

    •   Aditya (@netroy) Proposer 6 years ago

      Grunt just makes it easy to manage build-tasks, it doesn’t define what those tasks do.
      Grunt plugins do give plenty of functionality out of the box, but when you are building a large complicated app, you’d probably end up writing your own tasks.

      Grunt only takes care of some of the things I want to talk about, but there is more to it.. like:

      • faster deployments - right now, after we deploy our app, within an hour > 98% of open browsers have the latest code.
      • cache management - we (& our users) almost never have to clear our cache in production, never
      • & much more
  •   Ankur Agarwal (@devilankur18) 6 years ago (edited 6 years ago)

    Well I think this can be a useful session for many people. Though I have some questions

    Well I am not able to get the essence of ” within an hour > 98% of open browsers have the latest code” Can you please explain ?

    Also, I think you are talking about cache busting by adding a digest to the filename just like rails asset pipeline on each deploy? This feature is already covered by yeoman, or am I missing something ?

    •   Aditya (@netroy) Proposer 6 years ago
      1. ” within an hour > 98% of open browsers have the latest code” - currently most of our users leave the app open for hours/days. That would be an issue for a continuous deployment, as we might deploy a few times over the period the users are logged in (including critical bug fixes), so we reload the client side code without losing the state of the UI, but we do this when the user is idle or on another tab/window.

      2. Rails pipeline is great (with it’s own set of issues), unfortunately not everyone uses rails (or yeoman). Also getting a CDN to work with your app’s asset pipeline would be tricky. Adding a digest to the file name/path is a great idea, but things don’t stop there. It’s slightly more complicated than it sounds. You need to refer to these changed urls inside your CSS/JS, which are also on a CDN. It’s not a hard problem, & has a dozen solutions to it. Rails asset pipeline is just one of them.

      This talk is mostly about my personal experiences with building & deploying large web apps. A lot of it will be similar to what some standard tools do today, a lot won’t be.

  •   Ankur Agarwal (@devilankur18) 6 years ago

    Thanks for the details. Now its more clear. Looking forward to the talk :)

Login with Twitter or Google to leave a comment