This is going to be a weird sentence. But as expected everything is harder than expected!
Doing a game alone is no easy task everyone knows that. But I'm actually just doing a vertical slice and present it as a game!
Still... Progress as been made. I'm just getting lost in my unlimited tasks in all fields. And I like that! It keeps things fun. But it makes the actual coding go a lot slower than if I was only focusing on programming.
​Still, let's present what has been going on.

Again. I feel like at some point. Some people are going to join me and help me. And to accommodate. I created my own hack'n'plan to use kanban and priority request I assign to myself in order to increase my workflow and allow anyone to easily join.

Also I'm using tags on top of the kanban to also do good debugging and also if I need to focus on specific types of tasks.

(I also did a tag for documents but I didn't bother to take another screen... Sorry!)

So now I have a debug map that is way bigger than what I showed last.

I also coded a lot. A lot of architecture and not actual features. But I'm trying my best to shape up a system and then get used to it and then surf on it like a beauty.

And I have a first sight of what the game will look like with that architecture. I'm trying to do a lot of things to give myself freedom. The game might have more than one controlled character for very specific instances. So I have to code the actual control managing in the controller but the character functions IN the pawn... Once the polymorphism is done, I'll have a lot of freedom.
Anyway, so far I just have the view of the game and I already feel glad about how it feels. As creative you always doubt. And I do... For sure. But it doesn't mean I can't keep going on and trying.

That should be all for this week. To anyone reading. Take care of yourself. And thank you for reading.

Next week is school week. We'll see how I'll do!

I'm sad I didn't code too much (even though to be honest I did some minor shaders... And this game isn't going to be shader heavy ).
But I did quite a lot of work which I'm happy about.
I'll do some quick points so it's much more visual and impactful!

Ready?
Set...
Go!

Going from just creating the project to do a little bit of prep work for a good debugging environment, coding materials around a folder architecture using only material instances in a scene deriving from a master material. And also git if I ever want to work online with someone, all of that in one day. Is pretty good to be honest!

If you pay attention to my git commit you'll see I write them like a patch note so it'll be easier to follow my work. If you want to see it more clearly here it is!

• Created a map
  
• Created a main material
  
• Created dev material instances
  
• Updated the map a bit

Have a nice day everyone!

Just so you know this is going to be a small article, this post is just a little update about my work.

Spatialisation! Have you heard of it? It's amazing. It is basically EPIC OPTIMIZATION!
The name seems to suggest that it only applies about objects in space and while it might be true. It can also simply apply to abstract and hard to understand when you don't work with it data.
Usually, when you search for data. The easiest and most common way when you are searching through all the data of your application. Like let's say in a video game, all the enemies on screen. You might want to check their health and you want like this big skynet IA to tell each of them what to do based on that health. 

​

So them IAs, they right now don't need to act because of position but because of health. And to be sure you have to check each and every one of them every frame and honestly it sucks. Well there are methods to make it quicker and basically it all come to one thing: sort them.

So these IA's, instead of going through all of them. Let's say we sort them and we sort them only when they get damaged, like we put them at the beginning of the list. Then once we find one of them that is full health, we just stop. Because we know there is no more wounded IA's that needs to be checked.

I know there are better ways to handle things like that than my made up example, but let's go with an HOORAY for now. I have a lot less tests to do, my thousands of AI are way less computed. Stuff is cool.

Now what did school ask me to do about that ?

Octrees...

Octrees is a way of testing things between them based on proximity. And my job is to prove that it is way easier to compute things between them using an octree instead of an array or other sorting methods.

Octrees works way better with space so I'm going to do a little project about that on... UE4 (for no real reason, I just wanted to and experiment with the engine, and since it's a school project well...)

​So anyway here is an amazing photograph of my thoughts drawn poorly on paper!

Don't pay too much mind into my amazing programmer's notebook but on the drawings and maybe what's written.
I'm going to use UE4's colliders and compute them inside the tick/the function executed every frame to detect if object are overlapping each others instead of using unreal inbuilt system. 
I'll do that using a simple array vs an octree.
I'll do a static project first and then a dynamic one with moving elements inside the limit of a box
.
Static project will have:

• Differently sized objects
• Stacked up objects
• Sparse objects
• Iteration system to do a number of tests
• Same tests with an array
• Stats

And the dynamic one will have the same but the objects will be moving wich means the octree will change and I will need to figure out how to dynamically update it without it being too costly.

So far I have just done a basic debugging environment but I feel pretty good about what I did. There's not much visual about what I did but here's a sample!

Also I failed horribly when I started and it's hilarious so here it is.