Weldon is released to the submission pool of the 7DRL today, I hit my time limit, but I'm pleased with what I've got going for my first time joining the Jam.

Is it a successful entry?

Weldon was intentionally challenging to develop. I've been wanting to dedicate some time to do a polished game boy roguelite for a long time, and it would definitely take a while to build up the content based on how things are set up. What I present to you is a quick roguelite to play when you're stuck on the bus or in line somewhere.

In terms of scope, I achieved a bunch of my goals, though I obviously had to reign in my scope about halfway through the week.

What I Achieved:

• Procedural map generation, not endless variations but enough to keep it fun
• A Combat system that takes the place of dice rolls, and fits the console (insp. Kirby's Block Ball, Mario Golf)
• Randomized combat scenarios
• Item drops
• Difficulty Scales upon beating the first loop
• Full use of the Game Boy gpu and cpu available through GB Studio (more on optimization below)

What I didn't: Mostly due to time constraints, stay tuned to the page, I plan on making updates beyond the jam

• Passive Objective
• Multiple Environments
• Multiple enemy types
• Multiple types of combat mechanic

Development challenges:

The Game Boy, being 37 years old, is a creature of it's time. Mind you, GB Studio does take a ton of the edge off of working with a Z80 based system. Weldon was built using the beta5 release of GBS, and uses it's adventure core as the basis for the overworld movement and mechanics. As I stated on the main game page, I was looking to create something of a bite-sized roguelite like "Cave Noire",  a total hidden gem for the system. This was already asking for trouble, since the map generation in Cave Noire is impressive in its depth and variety of scenarios. Not to mention the expertly managed traps, monster placement, and light puzzling action. 

I have played through Cave Noire a few times, and I knew I wouldn't be able to approach it's level of depth within a week, so I focused on recreating its pick-up-and-play nature. 

GB Studio is a great toolkit, and is being expanded on all the time, even though it is primarily built as a classic top-down RPG engine, once you dig into the engine there's a lot of possibilities. The main challenges with development arose from how data is handled within GB Studio and the Game Boy In General. Truly random numbers are always hard to come by, especially on lower-horsepower hardware. Like with an Arduino, randomizing on initialization is a no go since the timer dictates the number you'll always get the same result.  I'm hijacking keypresses, interactions, anything not specifically timed to keep things from repeating.

Variables are treated as assembly-type variables, so essentially 8-bit mailboxes you push and pull from. The Game Boy is limited to 10 "Actors" as well, these are your GameObjects in Unity terms, not only do they act as characters, but they also house the scripts and determine execution order to a point. You cannot invoke scripts directly , but you can invoke an Actors scripts, meaning the game is littered with lots of invisible Actors doing the heavy lifting. In terms of art, each background may only contain 192 unique 8x8 tiles, which I am maxing out to make the forest in Weldon vibey.

I tried to nail down any bugs i could find, If you have access to a flash cart, or cart flasher, I implore you to give it a shot on real game boy hardware, the timing overall feels a lot tighter when running on real hardware.

7DRL definitely broke me out of a code rut, I'll be back next year!

B