How to NOT LEARN WebGL | Software engineering learning journey

An insight into how I learned WebGL and the mistakes I did along the way. Based of a comment from this case study video:

Video sources and examples:

👨‍💻 Follow me on social media for more projects and updates:

👨‍🏫 Get your WebGL 3d online course today at:

#softwareengineering #threejs #webgl

[alexisf]

Oh wow! I wasn't expecting a 12 minute long video at all haha, a short would have been great aswell :D.

You are exactly on point, by the way, this is what I was asking. I did follow Bruno Simon's course but I never really paid too much attention to The Book of Shaders, so that resource should help a lot! Although, to be honest, it has been like 5 months since I touched anything with it.

Your video kinda gave me that spark to try again, so I'm trying to finish a project I have that tracks outages (by using web scraping) in my city and displays them in a map in Blender (more like a shape of the map, rather lol) with different colours :D.

It's not the biggest project, but it will get me going! Thank you so much <3

[atalyy]

Operate on 'black boxes' when learning tough stuff. A good software engineer can abstract the problem, split it into solvable parts and sometimes use something that solves it without deeply understanding it. It's a combination of these factors. And the most important note - take your time.

[simonabunker]

Not so much calculus - trigonometry and matrices definitely! This is a great set of links - thanks for putting this together.

[CallousCoder]

Now shaders is similar how do you “learn it”. And for the record I suck at maths! And shaders is just maths but it’s applied maths. So just like programming, you do little steps. Work one tiny problem out at a time. Coming back to “understand the concept and where you want to go”. And sometimes it’s just experimenting and that experimentation actually taught me maths better than my two maths teachers did.
For the movie Dunkirk I had to code something for my former supervisor who was stuck and that was letting debris in the water (which were 2D images) Bob and flow along with the waves of the sea, whilst conforming to the camera track data.
The first and easiest approach was: “let’s make the water 3D and the objects that float on it we turn into 3D extracted planes that we warp along the vertices. And that water just didn’t look good with the contact of the ship. So I sat down thinking how we see and expect those 2D life jackets and debris to move. It was a step back into the psyche of us seeing things. Because it looked fake because it didn’t move the way we expected. After a day of analysis we had a plan of attack. We need to follow the tips of the waves which I could extract ik code but just crunching the image and raising the brights and then tracking those points over 3 or 4 frames. That would give a few motion vectors for each area. And those we applied to the 2D images and we had time that I could also see gradation in pixels (dark is deep bright is a higher part of the wave). So we added those to a mesh warp and you even had that subtle deformation. That was the most challenging applied maths for me. And I was seriously worried knowing how bad I am at trigonometry and linear equations. But with Google present we managed.