HN Theater @HNTheaterMonth

In order to be sold on webpack you need to understand the stack of problems it solves. I tried to do this during a talk last year; the relevant part is https://youtu.be/VkTCL6Nqm6Y?t=9m39s

⬐ sandGorgon

What do you think about jspm and systemjs - the way it separates front end vs back end dependencies and generates a config.js was very nifty.

⬐ peterhunt

Haven't used it, but I don't believe it has code splitting?

I wonder if they've looked at webpack?

Here's a great presentation about it by pete hunt of instagram : https://www.youtube.com/watch?v=VkTCL6Nqm6Y

I concur with him that it is the only tool that does these things Right.

I don't know what Github uses, but where I work, we do screenshot testing. We have PhantomCSS[0], which is a helper on top of CasperJs that visits pages, takes screenshots and compares them. There are a lot of other tools, Huxley and Selenium come to my mind, that do the same stuff. Even Facebook[1] tests this way.

[0]: https://github.com/Huddle/PhantomCSS [1]: https://www.youtube.com/watch?v=VkTCL6Nqm6Y

⬐ WickyNilliams

This talk really excited me.

I've spent a lot of time considering module systems, dependency graph optimisations etc.

Asset-graph [0] excited me previously, because it generated a dependency graph, implied by HTML/CSS/JS structure (e.g. script tags, link tags, image tags etc). This sounds great, but i can imagine it would be fragile. I'd been holding off on it until it matured, but I think now i'll skip that completely...

WebPack [1] adopts a similar approach. Generate a dependency graph of all assets for a web page/app and optimise from there. However, WebPack goes further, by pumping everything through the CommonJS module system. That way you have explicit dependencies between CSS, images and your JS. This will surely be less fragile and can be done without firing up a browser (headless or not).

This feels like the future.

[0] https://github.com/assetgraph/assetgraph [1] http://webpack.github.io/