This is the last part; Part 5 of a series.

The Book

At some point I obtained the book "MSX - Programmeren in Machinetaal" (That's Dutch. In English that would be "MSX - Programming in Machine Language").

I still have it:

The cover said "Complete course on programming in machine language. Everything about the Z80 processor and the internal life of the MSX computer".

One problem with this book was that it only covered the MSX-1. But the bigger problem is that it was difficult to read. I hate to say this about a book that was the only book I had about a topic that intrigued me: it wasn't a good book for beginners at all.

The book consisted of a long section about the Z80 which was completely generic, and a shorter section about the MSX computer specifically. So no cool graphics stuff in the first section of the book at all.

As an example of how the book didn't really help, it quickly dived into topics on how to use Binary Coded Decimals (BCD). I didn't care about BCD. Nobody cares about BCD. I wanted to make games.

I'm sure I must have tried to read it. I was a voracious reader. But I must have gotten stuck pretty quickly. The book explained the various features of Z80 assembly extensively without actually placing them into context: why would you want it? So even though the book tried to explain the fundamentals, in reality it assumed far too much previous understanding that I didn't have. I think I would understand it a lot better today, but I still have little desire to read it.

The bug in the book

I showed my friend my previous blog posts about the MSX, and he related an anecdote to me that I had forgotten. Somewhere in this book there was an assembly routine for copying data. But when we tried it, it didn't work. We then identified a bug in it, and fixed it. Some form of off by one error.

My friend pored over the book step by step, but there it was: a bug in black and white, in a printed book!

PC

I still used a PC at times as well. At some point a PC, I think equipped with VGA graphics, became a fixture in the same home office where we had the MSX.

I'd read some book from the local library which had a description of Conway's Game of Life. I used the PC and QuickBasic to create a simple, slow implementation.

I also used the PC to play with Fractint, a fractal generation tool. I was amazed that this tool was available for free. I remember reading the creators' "Stone Soup" description of how they had created Fractint, which described the process of contributions by a range of people that we'd now call open source.

In the 1990s the PC was to regain prominence in my computing life. It had games like SimCity, and later Civilization and Railroad Tycoon, that really captured my interest.

The MSX upgrade

The MSX had a 3.5 inch 360 kilobyte built-in disk drive. But a friend in my village knew how to upgrade the MSX's disk drive. One evening he came over and replaced the disk drive with a double-sided 720 kilobyte model. It also required the replacement of a ROM chip. It was a scary surgery, but it worked. Now we had double the storage capacity!

There was a protection system so you couldn't accidentally format a disk meant for 360 kilobytes as a 720 kilobytes: there was an extra hole in the disk to indicate this increased capacity. And 720 kilobyte disks were more expensive.

My younger brother, a big user of the MSX computer as well, at some point obtained a special hole punching tool. He now had the power to turn a 360 kilobyte disk into a 720 kilobyte disk! Storage capacity for free! Unfortunately it was a game of Russian roulette; some disks just broke down entirely when the tool was applied. Many disks were destroyed in the process.

Things ended up well for my disk-destroying little brother. He's still a big tech guy. He was the first in our family to have a computer with more than a terabyte worth of storage capacity. He now has a tech YouTube channel and designs LED controller hardware.

The Power of Assembly

My friend and I did eventually somehow learn how to control MSX-2 graphics from assembly. After several attempts that went nowhere, but were undoubtedly very educational, we created a game we called Matrix. It was based on this idea I had inspired by the Game of Life.

The rules of Matrix are simple. It takes place on a grid of squares that fills the screen. Each square could be empty or contain a block.

There are two players, blue and red, so blue and red blocks. Each player controlled a cursor and could use it to move blocks to an adjacent square. The goal was to destroy the other player's blocks, though I don't think we ever programmed the detection of the victory condition.

You could create a new block by orienting four existing blocks around an empty square:

You could destroy an enemy block by surrounding it with at 3 squares. This was actually quite difficult to accomplish and not a very effective way to attack the other player:

Sometimes creating a new block would create a special block: a bomb block, a laser block and a trigger block. Each had a special symbol. A bomb block, when touching an enemy block, would explode and destroy a certain amount of squares around it. A laser block would destroy either a horizontal or vertical line depending on where the trigger block touched it.

I've wanted to come back to this game concept for many years, but I haven't yet. It's a good excuse to dive into Bevy again.

Beep!

We had done it. We'd written a game in assembler. It worked! It was copying the blocks graphics. It was implementing rules. It was fast! Even on a Z80, machine language is very fast!

It was too fast. The cursors moved far too fast. It wasn't a usable game.

So as a temporary hack we introduced a system call issue a system beep sound to slow things down. It worked, though we would have to turn the volume off while playing this game, otherwise it was a horrible racket. We didn't have any sound effects or music for the game anyway, so that was okay.

We never fixed the temporary hack.

Leaving the MSX behind

It was the early 90s. The Berlin Wall had fallen. The Soviet Union was falling apart. I went to university. I had discovered the Internet. It was a different world.

At university I learned about computer language expressions, which you could visualize as a tree.

In the summer holiday after the first year, I had an idea. Writing assembly language was really hard. What if I had some kind of expression subsystem to make life easier? I could model the expressions in memory, using addresses to compose a tree. I wrote a simple evaluator, on the MSX, in Z80 assembly. I used a memory monitor tool to help debug it. It was a really primitive implementation. I realized that besides simple arithmetic, I could model conditionals as expressions too. And for loops.

University started again. I had a course on the Lisp programming language. Some other students struggled to get it, but not me. I realized I had reinvented (badly) bits of Lisp.

I think it was the last thing I ever programmed on the MSX, and with Z80 assembly.

Conclusion

We're crossing into the early 1990s, in which I eventually left my teenage years. Therefore this concludes my series. on my early experiences with computers. The early Internet became important to me in my late teens, but that's another story.