Not many new and exciting things happened in April, apparat from my participation in a Google Design Sprint witch I wrote a whole blog post on.

As a spinoff from the Design Sprint we had a 4 day hackaton to make a new prototype. This was a very front end heavy thing, but I got to play a bit with Elastic Beanstalk and Express.

Both Beanstalk and Express are typical products of AWS and the NodeJS community.

Almost all the services in AWS great when you look into them and they normally solve an otherwise complex problem. But as soon as you start to scratch the surface you start to notices missing features or quirky implementations. I spent way to much time on figuring out that Beanstalk tries to restore NPM packages on deployment - even though my deployment artifact contained the resolved packages. It’s a bit of a mess if you have your on packages from your own server…

On the other hand I was quite impressed with Express. After running Yoman generator I had a fine sample site and it was very easy to hookup routes and APIs. The frontend was created in ReactJS with own routing, so I just had to setup my API and then make a catch all for the frontend. I worked great. But as always when trying out new Node frameworks you are always left with the thought: Did I use the best tool for the job, or just the tool that was most popular a year ago? This is the curse of NodeJS, there is a thousand choices for everything and it’s almost impossible to know if you’ve taken the right choice before committing a lot of time to test it.