Inspiration

Well, I was thinking of something crazy , then a game 2d RPG game dev video popped up in my YouTube. Then I thought, why not make this. I mean, you explore and learn and test your learnings in an interactive game. Plus I can test codex for game dev.

What it does

It engages you into a learning game loop.

You start it , you go around and you see that your townhall is broken and you are broke too:) so, then you go to the study halls and meet with the wisdom wizard and he guides you through your lessons and you learn new things in the halls.

If you have enough knowledge gained up then, you are ready to fighting some Dungeons ^-^

In the dungeons there is loots and monsters. And the fun part is that you literally have to solve a math problem (or any problem , based on the dungeon type) to attack your enemy. It's like a turn-based attack mechanism. And your damage depends on your problem solving capability.

Damage Formula-
Damage = base weapon damage / (1/t)

Defeating the dungeon boss gives you reward and items that you can use for your future lessons.

Like exploring the wild to find rare item and using them for your next lessons and crafting items with them to restore/improve your townhall.

The game also have a merchant who sells food to heal you. (◡‿◡)

I might not have been able to push all the learning topics of the world in this game,,, but it's a concept that shows you how we can make gaming explorations into learning lessons.

How I built it

The entire thing was built in one codex session.Literally one codex session. And you will see the side effect of that in the next section :|

I used some free to use assets for the game, but I could use codex to generate the assets for me. but it would drain my credits.

But the windmill in the game was generated via chatGPT. and codex was able to generate assets with precise animation frames as well.

The sidechat feature of codex was awesome! Like, I was chatting in the sidechat about my preference idea, development phases and researched with it and made my prompts with it and then dropped the prompt in main chat.

This might costed me some extra credits but it was worth it. Basically its like I have a personal assistant who has all the resources and a an engineer to work for me.

Challenges I ran into

As I said , the long session and growing codebase drained credits faster at the last moments. But still I was able to ship an usable form of the game. Plus this is first time for me on game dev. I am not even a proper developer yet. Yet I had to learn about phaser, best practices, game dev starter and more.

I just spent my first few days on those research before I even started my first prompt.

Also I tried to oneshot it at first but it made something that was kinda 30% functioning and 0% visual. And I wasted 250 credits on that:)


Accomplishments

Umm, the thing I'm proud of is that I was able to pull it off in one single session. I dont know if it counts:)

But another thing is that , I learned a lots of things.

The amount of learning you go through in a hackathon is uncountable.

I learned:

  • basic gamedev tips
  • using codex properly
  • prompting (I finished a prompt engineering course from openAi for this hackathon)
  • networking with other people
  • learning how to organize things
  • and more and more mistakes , I cant all of list them.

Those may not be perfect lessons , but its atleast +1 in my knowledge base.

And I really love learning new things!

What More I learned

  • Learn coding first before you code with ai
  • Learn system designing first before you plan with AI
  • Learn prompting first before you drop "Build X, make not mistake"
  • Do not oneshot things unless you are just testing.
  • And many more lessons from my mistakes. Can't name all-

What's next for DungeonTopper

Upgrade, add good visuals, animations, lessons. making the townhall building useful, expanding the map, adding more features to it so you don't get bored just by fighting taking lessons:)

Built With

Share this project:

Updates