[RUMBLE BLAZING] Program code cleanup and optimization

I’m writing this post while riding the excitement of finally feeling like I’m about to reach a stopping point in the current work.

I’m very sorry for keeping you waiting and for taking so long to provide an update. Thank you very much for all your comments as well, and I apologize for not being able to reply to them.

The work wasn’t progressing as smoothly as I had hoped, and I was having a pretty difficult time mentally, which made it hard for me to post any updates.

However, I have finally almost finished reorganizing the program code, and I will begin debugging tomorrow.

Once the debugging is complete, I plan to release an update. Ideally, I would like to release it by the end of this month, or by early July at the latest.

This update is mainly intended to optimize the development process, so there will not be any major changes to the game itself.

I removed and reorganized duplicate processes and unnecessary variables, so performance may have improved very slightly, but the difference will probably be negligible.

As with the previous update, I plan to upload separate folders containing the previous version and the development version for the DLsite and FANZA releases.

The Steam version will be uploaded as a beta branch.

I will explain the details again when the update is released.

The explanation of the work I have done this time will probably be quite long, so I have included it further down the post.

It is nothing particularly exciting, though.

Debugging Videos

Here is a CPU-versus-CPU test from two days ago.

It stopped almost immediately after the match began.

And here is the latest CPU-versus-CPU test.

For now, I was able to complete an entire CPU-versus-CPU match without the game stopping.

While watching the CPUs fight, I thought it would be nice if players could move the camera themselves while spectating.

In any case, I was unable to play through a full match until yesterday, so I’m relieved that it is finally working.

What I Plan to Do Next

First, I will finish adding the moves that are currently still in development.

After that, I plan to change the character display on the character-selection screen to use 3D models.

A much older version of the game used 3D models on the character-selection screen, but I later changed them to static images because of memory-management concerns.

However, it now looks like the memory-management issues can be resolved, and I think displaying both the character and their selected outfit will make it easier for players to choose.

For the immediate future, I will continue development according to this plan.

There are also many other things I would like to add afterward, so while working on the features above, I will think about which ones to tackle next.

Explanation of the Recent Work

I would like to explain the code-reorganization and optimization work I have been doing.

Much of the existing code—the stuff with lots of English words written everywhere, haha—was written while I was still learning how to program.

As a result, there were many unnecessary or duplicated processes, and large parts of the code had become extremely difficult to read.

The main problem with difficult-to-read code is that, whenever I want to add something new, such as a move or a character, it becomes difficult to figure out where the new code should be placed.

It also means that adding new moves and other features requires writing much more code than it should.

I had been gradually reorganizing the code for some time, but whenever I tried to make major changes, moves would stop working or the game itself would freeze.

Because of that, it was difficult to properly address the problem.

However, after considering all the features I would like to add in the future, I decided that now was the right time to reorganize everything.

It took a little longer than expected, but development has become much easier now.

I will do my best to continue working on new moves and additional features.

There are probably still some improvements that could be made to the final sections of the code, but they should be easy enough to fix at any time, so I am going to set them aside for now.

Bonus: How Unreadable Code Is Created

Ideally, I think the best approach is to plan everything from the beginning and then build the code according to that plan.

In my head, I usually have an idea along the lines of, “If I write the code like this, it should behave the way I want.”

But when I actually try it... well, it usually doesn’t work. lol

It might behave in an unexpected way, or an error might appear.

From there, I repeatedly make corrections until it behaves more like I originally intended.

During that trial-and-error process, however, I end up adding all sorts of temporary variables and pieces of code.

Before I know it, the code has become a complete mess.

“What happens if I write it this way?”

“Nope... What if I add this?”

“That’s not right either. Then how about this?”

“No, that doesn’t work...”

“There’s no way adding something like this will make it work...”

“Wait, it worked!? Why!?”

“Wait... why did that work!?”

That is more or less what the process looks like.

Surprisingly often, I end up making something work without fully understanding why it worked.

During that process, unnecessary things inevitably begin to accumulate.

And that is how messy code is born. lol

When I look back at it later, I can usually understand what happened.

But code written without fully understanding it tends to become messy after all.

I have become much more comfortable with programming recently, and my code has gradually become less chaotic.

Even so, it is still rare for everything to work perfectly on the first attempt. 😭

But the feeling when it finally works properly is really something special.