Postmortem

Well, this is my first postmortem, but seems like a great habit to have, so here we are!
Wish me luck!
Theme and first hours
The theme here in Italy is announced at Friday midnight, so I brainstormed for 1/2 hours before going to bed.
I started by writing down on Milanote whatever came to mind, then I got a list of the main game genres from Wikipedia and tried to apply the theme to any of them.
I ended up with quite a lot of stuff, and came back on first day (late) morning to trim my list and elaborate on the ideas that did pick my interest.

First day

After a little trim, this is what I had by the start of the first day:

Chosing an idea
Here is a list of some concepts worth mentioning and the reasons that led me to discard them:
1. Teleport
Understanding some past events through environmental storytelling while being teleported around every 10 secs
-> _Out of skillset and resource budget_
2. "change world layer"
Sidescroller platformer with objects on multiple layers that are switched between every 10 secs
-> _Was meant to work something like Hue, so I discarded it for the lack of originality. Thinking of it now it could have led to totally different puzzles through the 10 secs limitation, who knows..._
3. Vision
Being able to see every 10 secs
-> _Couldn't really think of a fun way to wait in front of a black screen for 10 secs, but I came across some great entries that used this approach_
4. State swap
Swap character or team every 10 secs
-> _This actually seemed like a lot of fun, but I wanted to keep it singleplayer and not having to write a decent AI in a rush. I'll probably try to extend more on this one concept._
I'm pretty happy with what I made, but I can't help but think many ideas I discarded could have led to a more interesting gameplay. Probably not as polished though.
First prototypes
At the beginning I thought to make the arena harder to navigate using physics based zero-g controls, and the main character had to be an astronaut wearing a jetpack.
After making a first prototype it was too frustrating to move around and avoid stuff, so I decided to stick with simpler, snappier controls and increase bullets number, speed and pattern variety instead.
Graphics and concept
While scouting the Web for assets and graphic ideas I came across these cool trees and knight asset packs that ended un in the final game. I then started making the arena and implementing the leaf particles when trees get hit (Yeah, I'm one of those bad guys that waste a lot of time on trivial stuff before they even have a playable demo).
Having a knight in a forest I thought the attackers would be druids, mages or something on this line, so I made some generic magic bullet shader and a flash of light of the same colour to announce the shot.
This is what it looked like by the end of the first day:

Second day

On day two I took it a little slower: the previous day I realized I couldn't submit the game as a compo because I used third party graphics, so I had one more day to finish.
Core mechanics and level editor
I abandoned the idea of having bullet spawners inside the arena because I couldn't think of a coherent look for them. It just seemed cooler to have some unseen foe hidden in the forest.
So I worked on the bullet variants and the spawning system (Before I only had a fixed source bound to an object) and did set up two splines that define the possible spawn points.
Then I wrote a draft of the level editing system. It uses Unity ScriptableObjects to store lists of bullet types associated with the seconds to wait before launch. Spawn position is a randomized point on the previously mentioned splines.
Music
To take a break from development I programmed a drum loop in Ableton Live and recorded guitar and bass for the soundtrack. The final version is pretty much unchanged except for a little editing and mastering done the next day.

Third day

On day three the core game was finished, more or less.
The Horse part
I found a vector graphics horse I didn't notice before in one of the asset packs I was using, and I wanted to turn it into an NPC, so I made a the little dialogue system you can see in the game.
The plan was to make both the horse an in-game character and the knight present in the dialogue, so that they could answer each other and react to the dialogue in the world, but I had to cut this for the lack of time.
Anyway I had time to make two little intro and ending scenes.
Health system and general polishing
I dedicated some time to animating the character and making win and death screens.
Then I decided to add a few lives to make the game less frustrating and made a new HUD to match the style of the dialogue panels. I also made sure hit feedback was as clear as possible (Flash on character sprite, camera shake, sound fx).
Bullets graphics update
I wasn't satisfied with the graphics of the bullets (Variants were just different sized color swapped versions) and with the intro gag the game had took a humoristic turn enough to think of some new bullets to reinforce that aspect.
The first idea was to throw completely random stuff, then I kept it fruit-only for consistency.
At this point it made no sense to keep the flashy light before launch, and it got replaced by a more cartoonish speech bubble. Actually I wanted them to preview the type of bullet, but hadn't that much time and they would have probably been too small to give any significant help, so I kept the danger sign.
Level design and jam end
My last jam hours have been spent making levels and testing them, until I got ten working levels and was happy and tired enough to call it a day and upload.
Some features that didn't make it into the final game:
Speech bubble bullet preview, banana boomerang trajectory, dialogue featuring both characters, in-game Horse NPC.
And here is the final version that you can play on my itch.io:

You might argue that blue bullets had better contrast with the environment, and I would agree with you, but having bullets of multiple solid colors in the scene (to differentiate between types) didn't look that great.

Conclusion

Ludum Dare is a great place to try new stuff, see cool ideas and get constructive feedback, and once again it has been a great pleasure to be part of this amazing community.
Thank you for making it this far in this long post, I hope this was useful to you in some way.
Take care and I hope to see you again in a few months for LD52!