Chris Knott 2025-09-17 16:55:47 Does anyone remember a paper about a low level data format that was a kind of structured JSON thing, but it was a binary format and gracefully simplified into a raw string or raw int if that was the only data. It was definitely posted here at some point, maybe 5yrs ago though…!
Paul Tarvydas 2025-09-18 02:09:56 This isn't what you are asking for, but the question and responses fired these neurons in my head: altscript.com. I see Altscript as kind of like JSON++ or Rebol++ or Lisp++. FYI, YMMV.
Hemi Edwards 2025-09-17 22:36:07 Paul Tarvydas A friend of mine and I were exploring this subject. He was a computer engineer trained at MIT back in the mid 1980s and had developed expertise in higher level languages like Go and Ruby, but was interested in going back to first-principles. He was interested in the possibilities offered by UML and generating domain specific language to solve problems. I actually thought about combining UML with a TLA+. UML could be used as a visual representation of a system's design and TLA+ could perform the role of specification and verification of the system's model and then generate Ada code to build mission critical system, like they do in aerospace.
Paul Tarvydas 2025-09-18 02:26:14 Interesting. Thanks.
@Hemi Edwards I tend to think of UML and TLA+ as being on orthogonal vectors branching away from simplicity. I would expect that pasting them together would not reduce complexity, but, maybe learning from them and inventing something new might lead towards something simpler. IIRC, someone who knew UML and knew my stuff (PBP/0D), told me to look at UML "Deployment Diagrams" (in addition, of course, to the most concrete part of UML - Statecharts). I think that nothing we have expresses asynchronous concurrency well enough (and, I see the need for something that expresses asynchronous concurrency better).
Hemi Edwards 2025-09-18 07:30:01 hi Paul Tarvydas I wish my friend Gerry were here. He'd be completely at home participating in these discussions. I'm not actually a software engineer or programmer, but I have a slightly naive, intuitive grasp of how these systems work. I learned a lot from our dialogue about the finer points of programming and computing system architectures and Gerry was able to bounce ideas off someone who knew somewhat about the higher-level abstractions and prepared to support his quest to question the conventional wisdom of current programming and software engineering paradigms.
