January 7, 2024

After pushing update 0.0.30a, I immediately started working on 0.1: Towers and Temples. Therefore I had a decision: towers or temples? I decided to choose towers because I felt more inspired to do them.

My concept for a mad scientist tower from the start was to have a puzzle based dungeon. But which puzzles? My goal with puzzle design here is to have it be maybe one or two steps above a Wordle puzzle. Somewhat difficult, easy to get wrong if you don’t know what you’re doing, but definitely a less-than-five-minutes type of puzzle. Something like a full sudoku puzzle or even a crossword would be much too long. Another consideration is accessibility. When I say accessibility I’m not talking about arachnophobia mode or games-journo easy mode, but rather colorblind accessibility. 8% of the male population (so roughly 7% of gamers) are colorblind to some degree, making catering to the colorblind roughly the same as catering to the left-handed. This means puzzles where colors may be featured, but aren’t essential.

Ideally, I want a mix of “door puzzles,” which are encounter-driven puzzles produced by interacting with doors (much like hacking in many games), and “spatial puzzles,” movement or interaction-based puzzles based on object interaction or NPC movement (much like Zelda-style puzzles). Likely these puzzles will punish you for taking too long or entering an incorrect answer.

The first puzzle I made I called the “steam lock.” Basically it’s a lockpicking puzzle. The pins must be activated in a certain order in order to work, and this must be determined experimentally. For example, if the true order is {5, 3, 4, 1, 2}, and the player chooses to activate pin 1 first, the only pin that can be activated after that is pin 2. However, if the player chooses to activate pin 3 first, he can activate pins 1, 2, or 4 after that.

After determining it was a bit too easy, I split the pistons in half and added a middle state to the valves. This was not as serious as doubling the amount of pins to solve for, since it’s a guarantee that pin 1b comes after pin 1a. I decided to time myself to determine whether it was too hard or not. With the use of pen and paper, I was able to solve one such puzzle in 101 seconds. Without the benefit of pen and paper, I solved another puzzle in 147 seconds. Both times benefited from a mental algorithm I invented to solve the problem quicker, which I will leave to the community to figure out. Either way, this is probably a good amount of difficulty to puzzle through. The worst case here is I have to vary or reduce the number of pin sets, but that’s a problem for another time.

The next puzzle I decided to make is what I’ve always called a “light-switch puzzle” in my mind. I’m sure it has a formal name, but whatever. The basic idea is there is a series of lights which are controlled by switches. Each switch inverts the state of some lights and not others. The goal is to turn on all the lights. This would have to wait though, because I wanted to make this one the Zelda-style interaction type puzzle, meaning I would need an actual dungeon to be in. So the rest of the week was spent on getting an entirely new dungeon generating, as well as working on my car, which needed some maintenance this week.

I did manage to discover and squish a bug. This was quite an important one because it was a mathematical function that several pathfinding functions relied on. Oops.

Next week, I will finish up the actual towers and push out a release. I have reconsidered not releasing the update till it’s ready, since I want some other people to test out the puzzles and make sure they’re not too hard or too easy. Remember to email any questions, comments, and prayer requests you have to roguesofeuropa@protonmail.com.