If you can view this blog post right now you've got a front row seat to the future of the web. I launched this blog without using a bundler like Webpack, Rollup, or Parcel.
Here's why
- ES modules have broad support
- So does dynamic
import() - IE11 is almost dead and apps like YouTube and Gmail no longer support it
- 1-3 aside, I assume my audience is technical
The case for Snowpack
Wise guys: Snowpack markets itself as a development tool that you can hook your favorite bundler up to in production. It's a smart move that will earn them the highest adoption rate.
The big picture: Snowpack also optimizes production builds itself. It can accomplish nearly as much as Webpack, Parcel, et. al. I should know - I am writing a bunch of plugins.
Why I am bullish
- π₯ Active development: Snowpack is acquiring features at a clip.
- π Production-ready: Unlike Vite, Snowpack is capable of producing production-optimized ES module builds without a bundler. Vite uses Rollup for production builds and is itself just a fast development server.
- γΈ All-in on web modules: It's made by those who brought us Skypack
- β‘ Optimized for speed: Snowpack builds leverage HTTP/2 and HTTP/3. As a result, your app will download in parallel. Snowpack also code splits, minifies, and treeshakes - better than bundlers do.
Reasons to be bearish
- You (currently) can't accomplish server-side rendering without a bundler. Your app will have to be rendered client-side if you use Snowpack by itself. I have some ideas about solving this.
What do you think?
Are you ready to drop your bundler this year? Hit me up on Twitter with your questions, concerns, lamentations, aggravations, and libations πΉ.