What are the possibilities of developing new Tools based on formal models of computation (automata, etc.) ?

Question

When reading the "Writing Your Own Tools" chapter of "Kyma X Revealed" earlier today, I was very inspired to learn that a tool is essentially a state machine. I'm interested in hearing about any existing or previous work developing tools that express structured processes like Markov models, finite-state-machine representations of content-free grammars, non-deterministic state machines, or hierarchical models like Augmented Transition Networks if possible. Any advice, suggestions and/or pointers to partial implementations or related works within the Kyma universe are enthusiastically welcomed with gratitude.

BTW - I hope to quickly evolve past this style of opened-ended questions (e.g. "what are the possibilities of x"). I've been reading over the various documentation sources and experimenting with Kyma 7 for about a month now, and I'm still finding it a challenge to keep all the possibilities I've discovered or imagined organized respectively within my memory, along with a firm idea of where to find any relevant materials for "possibilities" I've witnessed in action or otherwise know will work. Thus hearing from the community about "possibilities" that may have since been realized as tools or other solutions, as part of.. (1) the learning processes of other newbies, (2) the more purposefully directed efforts of advanced practitioners, or (3) anything in between :-) .. could help a beginner like me form a more definite mental picture of the existing potentialities of Kyma 7, and within that more specifically, which idea paths might be best to prioritize first for achieving both personal learning and practical results within a cultivated flow-experience (i.e. before any kind of frustration or apathy might appear as a result of repeatedly focusing attention: (1) on the wrong missing pieces of the puzzle for a given point along the natural growth arc of an idea or project, or (2) at the wrong level of granularity... for achieving satisfying or otherwise clearly discernable results after each iteration, and so forth...)

Answer 1

Since you can write Smalltalk in the Tool, and since Smalltalk is a general-purpose programming language, all things are possible.  Although that may be a bit like saying that all things are possible with a Turing machine, since it doesn't say anything about how one would do it.

To start your investigations, you could try searching the Q & A (using the search box at the top right).  For example, if you search for "automata", "Tool", or "Smalltalk", you might some relevant discussions.

Once you start diving into Smalltalk, my advice would be not to do so to the exclusion of the kind of "analog computer" style thinking offered by the signal flow in Kyma.  It's a different way of thinking, but in the real time signal flow evaluation, you can also find conditional logic, state and multiple threads (for example, in the Multigrid).