Sleepy Trees

Sleepy Sand

I'm trying to get my different devices to communicate with each other (so that they combine their computing power to make a HUGE sand engine). This week, I managed to get my desktop to 'offer' sand atoms to the device 'below' it. I'm getting closer to having all 9 ways finished! I'm also starting to map out the video and write the script. I'll finish and release the video later in July.

There's one small thing I'm not sure about though. One potential way to speed up sand engines is to make certain atoms 'sleep'. ie: Sand atoms that are stationary don't need to be processed until a nearby atom moves. This can help to speed things up quite a lot in best-case-scenarios. In worst-case-scenarios, it makes no difference (because every atom might need updating anyway).

I didn't really want to include this, because I want my engine to process EVERY space at least once per tick (on average). I consider this one of the criteria of the engine. However, maybe the video won't feel 'complete' if I miss it out. What do you think? Should I include the 'sleep' optimisation? Or should I leave it out, because it's 'cheating'? Maybe a special mention?

TreeTode

I've renamed my Lisp-like language from LispTode to TreeTode (to fit more with the nature theme). And also... because I want to experiment with making it less like Lisp. Lisp is known for needing ((lots and (lots and) (lots (of brackets))). I find it quite hard to read. I want to try making a Lisp-like language where brackets are OPTIONAL. You can read the draft documentation for it here: l2wilson94.gitbook.io/treetode

This is what the 'bracketless' TreeTode could look like:

But... why am I making another language? What has this got to do with SandPond? Well... a couple of reasons.

Firstly, TreeTode is really helping me to test out MotherTode (which I use to make programming languages). It has shown me some of the problems and opportunities for improvement with MotherTode. This will help me to make more exciting languages and tools, so that I can make cooler stuff in the future.

ALSO, multiple people have recommended building a Lisp-like language. They tell me that there's a lot to learn from it (about language design). So far, I think they're right. I really like how you can just let code... be code. It doesn't have to do anything. It just ... exists. Does that make sense?

When TreeTode is a bit further along, I'll make a Deep Dive video about it.

🐸 Thanks for reading! See you next week!