Trickle-Down Dragonslayer

Title screen (emulated).
The first disaster.
Probing for the first repair.
Successful first repair.
The DIY null modem used for serial.

Inspiration

We had an Apple //e lying around. Got a shoddy Super Serial Card. Thought it would be funny to make an AI generated text-based adventure game, just like the GPT2 days. Unfortunately, as you will see, we ran into a a few hurdles.

What it does

Trickle-Down Dragonslayer is an Apple //e based cold-war medieval text-based adventure game straight from the 80s. Fight past the Iron Wall to slay The Red Dragon and free Ruskiland!

How we built it

With a lot of pain. See below.

Challenges we ran into

At 11:32 PM, the day before, the Apple //e had some complications. It's proud owner (Louis) decided to do his washing, drying, and run his desktop at the same time. His breaker box did not appreciate this. Nor did the Apple //e.

At 09:15 AM, the day of, Louis sent "We've had a nightmare, think my IOU [input output unit] chip is dead." to the team groupchat. Some quick home debugging proved fruitless. Once the Apple //e made it to campus, we frantically pulled integrated circuits off of our donor board. First the LS245, then the LS244s. No joy.

At this point (around 12), hope was running low. We had just replaced one of the most likely failed components (CD ROM) but nothing. One question saved us: what does an EF ROM do anyways? Turns out: we don't know. Also turns out: replacing it fixed our problem. Job's a good'un.

Beautiful. The next step in our plan was to hook up a Super Serial Card (SSC) so we could connect a modern computer over serial to communicate with a local AI model (which we were actively scouting). Easy, right?

By about midnight, we gave up. Turns out, our dodgy SSC clone from '84 does not work how we expected. Heath experimented with reading directly from the cassette port; maybe we could read data from the model through that? Peter started writing up some Applesoft BASIC to get at least something resembling a game running.

Unfortunately, neither of these proved fruitful. Cassette was too slow and unidirectional. BASIC had mystery syntax errors.

Fortunately, Heath discovered an online C compiler and emulator which might work. Peter got to work translating the BASIC program into C. It worked... though it was now 04:12 AM.

Over the next few hours Peter hacked away at the game and Louis storyboarded, wrote a manual, and made some beautiful marketing materials. A complete, working game on a complete, working computer. Nice.

09:56 AM. Disaster strikes. Again. One of the electrolytic capacitors in the power supply had released its magic smoke. The hunt for a 250V 47μF capacitor was on. Nothing usable in the hardware lab. We were tipped off about a broken microwave ready to be scrapped for parts. Toolbox in hand, we made our way over to the microwave.

We pulled it apart. Took out the main board. An unfortunate failure, the only 47μF capacitor was rated for just 25V. The only 250V+ rated capacitors were 4.7μF. Maybe the makerspace has the right capacitor? After running across campus and rummaging through their stock we found nothing usable. Even a quick dumpster dive did not suffice.

Last resort. Take the cable harness out of the original power supply, strip back the wires, connect them to four lab power supplies. All we needed were +5V, -5V, +12V, and -12V. Oh wait, the power supplies don't do negatives. Time to break out the breadboard.

We got a beep! Hopefully nothing goes wrong when we display it.