I think every suggestion you have there would fall under that very same lamppost: a linear model of computation going only one direction. Sussman details in this talk a different model he calls propagators that deal with information flow back and forth, allowing for some interesting ways of working with "fuzzy" and incomplete data, while still allowing for cheap backtracking if a contradiction arises. I haven't read the Radul & Sussman's paper "The Art of the Propagator"[1] yet, but I fully intend to now having watched the talk.
[1]: https://dspace.mit.edu/bitstream/handle/1721.1/44215/MIT-CSA...
Re:We Don't Know How to Compute (2011) [video]
by mgaunard on Monday May 15, 2023 @05:56AM
(#35945794)
Few of the things we care about need to be fast, and fewer of the things that need to be fast are held up by number crunching performance.
There are already a plethora of languages designed for numerical computing. Just use that if that's what you're doing.
Compilers for the most popular languages are pretty amazing these days. Yet it’s still on the developer to make the most of them. One aspect might be performance:
Another aspect might be data security rules, e.g. tagging and labelling with the ability to enforce rules for the developer.But there are literally hundreds of aspects to deal with. Our languages today look at the problem space through a toilet roll tube with the other eye closed.
The future of computation is guaranteed to be rich, vibrant and exciting.
[1] https://en.m.wikipedia.org/wiki/Streetlight_effect