I tested and even with no color command I still got the glitch. This is not a color-related bug, but rather with non-ASCII tokens. I tried myself on my Prizm using OS 01.01.0200 and 01.02.0200.

Hmm, I only noticed it one time with one character.  Perhaps I didn't notice something...

Yeah that's what I thought about doing. Basically the chars won't display at the right location, if at all:

I wasn't aware that the glitch worked that way, thanks Dj. If it starts displaying oddly at least I'll know it isn't the program. Hopefully, Casio will fix this soon.

AH ok sorry to hear. I hope you don't quit afterward.

Hi everybody, I'm back again!

I could take the entire post explaining why I was absent and apologizing for that but instead I'll share some new developments with you I'm sure you'll like. It'll suffice to say I was preoccupied with a holiday job, thinking over something new to implement and attempting to program those ideas.

After some failed attempts I became interested in the subject of Randomly Generated Mazes. At first I just started it as a seperate project, just for the fun and challenge of coding it but when I had a first, buggy and extremely slow version running I thought I could implement it into my big project!

The idea is that my program generates a random maze of 19x5 because to generate one correctly it needs to check the locations next to the 'current' cell (more about that later) which will result in an error if the current cell is at the edge of the screen. Also, a wall around the maze looks slicker and it drastically cuts down the generation time (by around 25%!).

Now for what the function of the maze in the program will be; it will kinda act like a dungeon. While walking around in it you can randomly encounter monsters and discover treasure. The first part won’t be difficult since the fighting program is completely separate from the main ‘map’ program. The second part will be incorporated like this; when the maze is generated 1/8 empty places will be replaced by breakable walls. These look the same as regular ones and mainly serve to make navigating through the maze harder. But it will also replace 1/40 empty spaces with a different kind of breakable wall; one that will yield treasure upon breaking it.

The maze generates like this; it will first generate a matrix of 7x21. The first number represents the amount of rows=the height, so to keep the visual look of the matrix similar to the look of the final product Y goes before X. The maze will then fill up completely with the value 1. 1 stands for a wall. The program will then assign location (2,2) in the matrix to the starting point and begin by setting it to 0, which is an empty space. Then it will randomly check whether a neighbor is not a 0 already and is not on the edge of the screen. It will then proceed to check whether or not the neighbors of that cell are 0 as well. If they are all still 1 as well it will set the cell to 0 and set the X and Y coordinates of that cell as the new standard X and Y values.

At this point it will repeat until it can’t go on in any direction. At this point it will retrieve data from two lists that have been storing the coordinates of each visited cell. It will mark the one before the one it is in as the current cell and start at the beginning of the whole process. An important note is that if it starts backtracking like this it will start overwriting parts of the two lists, starting one after the cell the new side-route originates from. This is to ensure it will also check the side route for possibilities to branch off again, and to make sure it’ll check the cell the branch started for the possibility of having yet another branch.

This entire system will ensure that there are no walls left the maze can still bend into without looping on itself. The downside is that it takes around 40~80 seconds to generate a complete maze. One of my idea’s is to terminate the program when the maze has 60 empty spaces which will cut off the lag at the end when it’s backtracking all the way back to the beginning but will also leave some unused wall segments.

And that’s about it! Let me know what you think of it, I’ll either post a demo of the maze generator over the course of the next week or upload the new version of the game with the random dungeons implemented!

Welcome back you two. I am glad you didn't quit calcs or something. Also I'm glad this is still alive. I'm not as active anymore so I don,t check updates regularly but I still can't wait to see this finished.

Even if you reduce the time it takes to generate one of these dungeons it will still take a (relatively) long time to generate them.  I would suggest that you do use these dungeons in your game, but don't place yourself in a position where you have to generate a lot of them as that will take a lot of time.  Idea #3 seems like a good idea for this situation. 

More input on how the dungeons should look like would be appericiated! I'm currently doing a bit of fiddling around with the Casio SDK (aka C programming for the fx-9860GII). With more input and idea's I can do more work on the RPG!