Notebook Adventure
A submission for Retro Platform Jam 4
The idea
I spent whole days trying to understand Pokémon Emerald's inner workings by watching no-cash's VRAM viewer with tired empty eyes. "It's no good..." - I said to myself. "How in the world are you gonna create an engine capable of doing even a fraction of what you see?". I have to admit: I'm bored of this question my subconscious keeps asking me for more than 6 years. "You know what? I'll do it! I'm no longer that 2016 guy who replaces the example DrunkenCoders logo with some other bitmap image and calls that thing «homebrew»." Genuinely speaking, what I'm afraid of? What could go wrong in starting a new project? Either it will become a piece of software art cummulating both my entire work and satisfaction, or it will turn out to be another one of my hundreds dead attempts floating around in my filesystem.
I watch the sunrise of a new day with a resonant promising feeling. Partially self-encouraged to undertake another one of my abyssal coding session, I make myself comfortable at my desk while opening the spell book and clicking a can of Dr. Pepper. There are my (the programmer's) mere weapons, along with a pen and a piece of paper full of unrecognizable notes from the precedent day. The holy devkitPro libraries, the classy-ish Notepad++ and the bizarre world of Makefile. And a -somewhat- experience turning -someway- into a curse which it seems that -somehow- I'm the only one I can relate to.
My whole activity as a homebrew-ist fits perfectly in 8 words: Long hours of coding, hard days of debugging. But at least, I'm giving it my all (time, luck, praises, swears, effort), then I desperately smash that Arrow up + Enter combo in the terminal in the hope of seeing any reliable progress. I wanna do it - and I'll do it. For the sake of my dreams, for the taste of my childhood.
After drawing myself repeatedly on the edge of exhaustion, I felt like I was touching the absolute. The engine is finally working. "Good job, N•I•L! Remember what you said? Pokémon Emerald, uh...? Not even close!" My so-called replica of the game is so different than my source of inspiration that I can call it an unique engine in itself. However, I felt no disappointment. I'd rather say that I felt a spark of revelation. It's the fact that technically speaking I was followed the wrong path from the very beginning, but I actually did arrive to a correct answer. My own correct answer. My buggy (correct?) answer.
Time passed, and meanwhile I managed to get caught up in the moment enjoying the enthusiasm of writing my very first paltformer. I was so absorbed in what I was doing that I commited the huge sin of losing sight of crucial springs of crashes and errors. Disable Backgrounds - it works. Disable Sprites - again, works fine. Put everything together - and watch how the CPU gets stuck into an infinite cycle with savor of bad alloc. "Now, I've done it. Where in those thousands of lines is the problem? This project wasn't mean to exist in the first place, was it?"
void * operator new(size_t size)
{
void * p = malloc(size);
DEBUG_MSG("%x. + %x : %i [%s]\n",cnt, p,size,dbg_ctx);
dbg_ctx="";
cnt++;
return p;
}
You can feel the desperation of making the new and delete operators write logs to the SRAM ("Uhm, N•I•L, you've heard of the great no$gba's debugger, right?" - Yes, I have, but I don't know how to use it, now shut up, I'm concentrating!). After three days of worthless brainstorming and bad code dianostics, I randomly started to change between static and dynamic arrays. It worked!, though many things still remained unexplainable to my humble knowledge.
While a bunch of other projects I worked on thought me about skill and patience, this one has a truly remarkable mention: I refused to abandon it, no matter what (and you won't hear this from my mouth very often). It has its flaws (still, but I think I should live with that) and I feel like me and my work are chiseling each other. I ultimately learned how to use the damn debugger, which spared me from a lot of struggle with discovering stuff like a strcmp-ed null pointer which could mess up the entire execution. Also, I was suprised that the ELF file is not just an intermediate form between code and the built ROM, actually it can also be executed (though the expansion "Executable Linked Format" was not enough for N•I•L to understand a blatant fact like this). And guess what? It contains symbols! Useful for - DEBUGGING! Phhoaaa... I'm grateful to my life that it showed me a thing like this only at the end of such a hardcore coding challenge.
All these being said, I have to add: Nevermind. Creating this game was fun. It really was fun. Nothing makes my blood pump more than waiting blindly changing a line and raising the phylosophical question: will it crash, or will it not? If this wasn't enough, this project also had me try MilkyTracker and made me laugh at my not-so-old attempt to mime a composer.
This game is my up-to-now best, and I'm incredibly excited to show it to the world (or the ~10 people who will play it, in the optimistic case).
And I din it... super !
How to build
Building process requires te devkitPro environment to be installed on your system.
git clone https://github.com/NotImplementedLife/NotebookAdventure
cd NotebookAdventure
make -f Makefile
Screenshots
Credits
- I should thank DogedomStudios for hosting Retro Platform Jam #4, the pre-text and also the factor without which Notebook Adventure wouldn't have come into existence
- BowersIndustry for discovering a ladder bug in the first level
- modarcihve.org for providing spectacular retro music-related resources
- libgba for the GBA development tools
- any GBA-related forum, article, post, source codes all over the internet, no matter I have read them or not. They are amazing and each one is a valuable source of answer for any of your possible questions.