TODEPOND TIMES: TODEPOND PONDCAST

Welcome to the final ever edition of the weekly...

🐸 TODEPOND TIMES 🐸

Yes, you read that's right. This is the final ever TODEPOND TIMES.

But don't fear! Next week, it'll be the first ever edition of the weekly TODEPOND PONDCAST.

I'll let you in on a secret... The TODEPOND PONDCAST will be exactly the same as the TODEPOND TIMES. The only difference is its name. Some might say that this is a simple "renaming" from one name to another. But those people aren't thinking in terms of virality and clickbait!

Anyway...

What's new this week?

Tode PondPond Cast

As you may have heard, I've decided to rename the TODEPOND TIMES to the TODEPOND PONDCAST. This is for a couple of reasons.

Firstly, Jimmy Miller suggested the name "PONDCAST" to me when I was on the Future of Code podcast, and I find it quite funny. By chucking an extra "TODEPOND" on the front, it's easier to google, and I enjoy how redundant and hard-to-say the name is.

Secondly, I'm now consistently recording myself 'speak' these updates, and I want to show off about that!

Some people might not realise that they can hear me blabber on about code each week, and they might want to (for whatever reason).

Instead of the weekly-update being "some text that you can listen to", it'll be "some speech that you can read".

The gap between my videos has gotten very large. Although I want to improve this, I do want to demonstrate that I'm putting out some audio-visual content in the meantime (just, without the visual part).

Portugal

I've booked my tickets + flights + hotel for SPLASH con in Portugal. Last week, my submission was accepted, so I'll be demo-ing CellPond there in October.

And this week, I've been unpacking some of the feedback I received for it, which has been very helpful.

Normally, the kind of response I get is something like "this is cool" or "this is dumb". But this time, the feedback has been much more precise and considered. I'm not used to having such a detailed lens held over my work! It's great! It's made me realise and discover a lot of new things already.

I'll continue to explore some of the more complex feedback over the coming weeks. But first, let's take a look at some of the simpler notes.

Here's one:

"It's hard to tell from the videos whether the language is Turing-complete or not."

My response: It is turing complete. However, it would be impractical to try to create certain programs in it. This is similar to how the Game of Life is technically turing complete. In theory, you can make any program with it. But is it practical to do so? No.

For this reason, I find it strange that there is often such a focus on 'turing complete-ness' in programming. It's not a hard requirement to meet, and it doesn't mean much about the real-world capabilities of a system.

I poke fun at it in both ✨ New Cellular Automata and ☁️ Define Define.

"it probably can at least represent class 3 rules"

'Class 3' is a nerdy way of saying that it can end up in interesting chaotic states.

CellPond can also represent class 4 rules, which is a nerdy way of saying that it can be a combination of chaos and order, like the Game of Life.

Speaking of the Game of Life...

"I would suggest preparing a more formal presentation where you show ... how you can program Rule 110 or Conway's Game of Life in CellPond"

Rule 110: Sure, no problem.

Game of Life: That would be really difficult. Game of Life is synchronous and deterministic. In other words, all rules fire in all cells at the same time, and they always result in the same thing. CellPond, however, is asynchronous and non-deterministic, meaning all rules fire in random places in a random way. It's a living, breathing, unpredictable system.

The Game of Life and CellPond are both turing complete. This means that, in theory, it's possible to emulate one of them in the other. But that doesn't mean that it's practical or easy to do so.

Even so, many people ask me to make Game of Life in CellPond. And I always give that same unsatisfying answer.

Often, they don't believe me. They think it must be easy. If CellPond is turing complete, then why can't you make whatever you want?

Well, the Game of Life is turing complete, so why can't you make CellPond in that? Ha!

But really... It's a good question.

I'm tempted to make a video where I do make Game of Life in CellPond. It could take months (or more), but I'd like to do it - just to demonstrate how hard it really is - and to prove that it is possible.

Components

This week, I've been building more of Arroost's user-interface (for the next video).

To do that, I built a fancy thing called an Entity Component System.

In the system, you can make things, and those things are called entities. And those entities can have certain characteristics and capabilities, which we call components.

For example, in Arroost I can make a 'button' entity. It has various components that let it do what it needs to do. It has a component that lets it appear on the screen. It has a component that lets it move around. It has a component that links it to the rest of the engine.

And some of these components have components of their own. For example, the appearance component has components that figure out its position and size.

And these sub-components can have components of their own, which can have components of their own, and...

It was partway through making this that I realised there was no difference between an entity and component.

Literally, my code for an entity was exactly the same as my code for a component. This happened by accident!

I guess that everything in my system is a component. Entities are just what we call the "top-level" components.

Or maybe everything is an entity? We'll never know.

Anyway, thank you so much for all of your support. You're allowing me to work on these strange projects, and to develop them, and to take them around the world! I can't thank you enough. Hey, I hope you have a great week 🐸

Days since tode fell asleep: 225
Days since bot went missing: 190