Day 122: Today wasn’t super productive. It seems like every first day of a new sprint isn’t really productive. I think fixing it would require more resources than those wasted during those days. At lunch I had an ersatz of a salad. It was just corn and tomatoes with a bit of olive oil. That’s incredibly interesting isn’t it? I always knew that you scretly wanted to know what I’m eating.

Daily interesting stuff:

• Reddit post, freelance dev got scammed. Work was done, but no pay ever arrived. This sucks big time. How can someone can engage themselves to pay a certain amount and then don’t live up to their word like that?

• Game AI pro first edition is available for free online! This book is a big reference on game AI I wanted it for some time, but it was too pricey. I’m glad I found that. Oh and to those wondering it’s 100% legal. They just decided to give their book for free after two years.

• Reddit post, when is cloning and when is inspiring? Pretty standard post, but it was still interesting.

Tasks done:

• Planned the next sprint. This was way harder than usual. Mostly because on my beta draft, I only wrote “content” and “polish”. Yay, very useful past me, thank! So I played the game trying to find what content I should make. Pondering to see if it’s not better to wait for the graphic assets before doing the new rooms. I decided to still go with a “Content” sprint. I added some trello cards, but I’m not satisfied with the level of details. And now that I wrote this, I just remembered that I didn’t put a deadline at all. So I’m going to do that. Be right back. Alright, so this sprint will last three weeks and should be over by the 10th of mars. This leaves three weeks for the polish sprint. Oh and while I was playing the game, I realized it was very close to what I want. I think it only need some more content (rooms), be harder, have clear long term goals, and then tuning and polish, and it’s done.

• Fixed a bug where the TP enemies could get in the walls. It’s a simple thing that I forgot to do in the previous sprint. I also made that they can change room like the tracking enemies. But since I don’t like code repetition, I created a component for that and assigned it to both enemies type.

• Tried the new tileset assets that the graphists sent me. It’s pretty good. But I didn’t have a center tiles for the walls, and the floor has too much repetition. But it looks good with the ships.

• Had a very, very unpleasant surprise. It turns out the room collider automation during the Tiled import is completely fucked up. I had no idea why it didn’t work anymore. So I looked into it. It was pretty hard to track this bug. But I managed to find the problem. And, yes it’s super hard to fix. It turns out that when a tile map is imported, it’s set up as a mesh by Tiled2Unity. This is normal and frankly it saves a lot of perf. But the problem is that Unity “optimizes” the mesh vertices. And I read those to find the colliders. It works perfectly with the first tileset. But with the new it does nothing right. The reason is that when the wall tiles are different graphic tiles (like border and corners). Well the vertices aren’t ordered from left to right. Instead it sometime go back two units left or not. There’s no set rule. And guess what? I use the assumption that it goes from left to right as the core of my tool. So I looked at many things trying to fix this. I didn’t succeed yet. But I have good hope that using the triangles array might solve my problem.