ConwayLife.com

AlbertArmStain wrote: December 6th, 2022, 8:05 am What about in Symbiosis?

2017 me had no idea what he was doing, and stumbled upon an interesting rule.

silversmith wrote: December 6th, 2022, 6:41 pm The suggestion doesn't address the ambiguity of choosing how two rules would interact, which seems to be the larger issue with arbitrarily "hybrid rules"

Ah yes, that'd be a problem. I think discussion of that would better go in the hybrid rules thread though. Although I should say that a Turing-complete language would probably be required to produce any arbitrary hybrid cellular automaton, so something along the lines of John Walker's simulator (https://fourmilab.ch/cellab/webca/) plus a library of rule class constructors (i.e. those for each of Henselstring, MAP, Generations, etc) and wrappers (e.g Symbiosis, StateInvestigator, Wireworld, etc.) and then somehow use state names on those variables.

What I've noticed with Circuitry rules in particular is that the more experienced users typically create their own rules and don't use others' very often.
Factors for this might include paradigm shifts between rules, lack of documentation for some, and personal interest favoring their own ideas.
If this turns out to be true, that means there is a natural tendency for the Circuitry community to fragment themselves, making many creations one-man projects. This makes most of the larger-scale builds very difficult to build without prior experience in actual hardware architecture.

I could be wrong, though.

Could anyone provide good counterexamples other than the "classics" such as von Neumann, Codd, and WireWorld? Maybe Bliptile and NoTimeAtAll have had their fair share of activity.
What these rules also have in common are that they have a lot of derivatives, probably due to interest, and/or simplicity/extensibility:

• JvN - Nobili32 and Hutton32
• Codd - Devore and Devore2, as well as the loop rules
• WireWorld - WWEJ3, WireWorldModern, probably Trine as well
• Bliptile - Flow6, and maybe every vN transmission rule if you think those should be counted, including:
• NoTimeAtAll - LittleTimeIfAny, Shanghai, maybe more

Those are the ones I could name. Often, research tends towards making new rules, proving something about them, maybe building one thing in them, and maybe moving on to the next derivative. In my opinion it would be nice for people to pick up existing rules and collaborate on building stuff in them to eventually snowball a community, instead of making so many derivatives for so few people to use. I have no idea how to make this happen when that's how people work right now.

I have noticed the isolated nature of circuitry rules as well, and agree with many of your points.

One thing I can think of to address that issue is to pin a list of such rules to the top of the OCA board. It would give CA newcomers a reference for where they can go to contribute to existing rules, and provide a convenient reminder for existing users. However, I am hesitant to call it a solution, as it has many problems of it's own. A potential issue is the length of such a list. Many rules with circuitry have been made and a list of all circuitry rules could be overwhelming in size.

Another issue is figuring out which rules are considered "circuitry". There are numerous rules which sit on the border, and anyone interested in promoting a rule would be incentivized to extend the definition to include it. In addition, computation ability and construction ability are two separate and almost independent dimensions of circuity a rule could have.

Finally there is the issue of relevance. Specifically there are circuitry rules which are either too abstract or old where it doesn't make sense to include them on such a list. For example, a loose definition of circuitry might include a bitwise cyclic tag emulator, but it doesn't make sense to use such a rule for any projects. Similarly, a rule like Bliptile is historically significant, but not practical for projects when compared to modern rules.

Given all this, one option would be to use a standard similar to blah's list, something along the lines of whether a compact computer has been constructed in the rule.

i don't necessarily know about others but i personally find the lack of convenient unambiguous documentation of most circuitry rules the prohibitive factor. it would be nice if such things were just included in the ruletable.