ICAB for JavaScript Lab Brussels

Organized by

Obviously this should be experienced with a Cardboard or proper VR HMD : https://www.youtube.com/watch?v=ajBiOge0_HA There is a lot more to see e.g. stack (and even an Easter egg, hint you can look back!)

Going further

For participants

For instructors

  • ~30 participants
    • ~15 didn't try VR at all
      • they could ALL try a Cardboard demo during the beginning of the talk
    • 0 tried the Vive
    • 5 tried the Oculus
    • 3 programmed webGL
    • 0 programmed anything in VR
    • gave away 10 cardboards
    • comments via Meetup
    • 1 participants came after the talk and said very quietly "Wow... it was in 3D from the start, I didn't notice, I thought everyone around me did so I didn't know how to react but that was really impressive"
  • difficulty of making a meaningful presentation for both VR newcomers and VR experts, JS newcomers and JS experts
    • assessing with introductory questions help to better understand what could be done
    • code presentation at the end was probably too short and too late for people who wanted to integrate VR to their existing framework
      • yet also too long and too complicated for newcomers, presenting the ECS architecture and why it matters in a matter of minutes might not be enough
  • no video yet
  • to improve for next time if the audience is actually consisting of programmers (should be assessed quickly via introductory questions) with some knowledge of VR
    • re-factor to better highlight key principles
      • Custom element e.g. <a-box> to remove all veil of mystery
      • registerComponent() to insist on the importance of JavaScript at large and the benefits of three.js in particular
      • registerShader() depending on the knowledge of the crowd
      • scene.hasLoaded with total lack of magic
        • and its importance for getAttribute()/setAttribute()
      • using emit('myEvent') with <a-animation begin='myEvent'>
    • skip querySelector()/getElementById() but instead focus on querySelectorAll() as it shows a lot more potential
    • move all images to the asset section with meaningful names
    • don't just present mixin or <a-assets>, start with the pain point THEN show the solutions