Declarative Object Graphs: programs as queries

If there are no physical constraints could a program look up it’s solution by firing off a set of queries into a solution space?

While taking a shower this morning I was pondering about some things I read recently.

First some history. Programs were created to run in constrained resources. As such, a program is just collection of small bits of data transforms. Yes, more complicated than this, don’t interrupt.

What if there were no constraints? In future we will have more memory and network speed will be much better.

With no constraints we could map inputs into outputs directly, no computations. A program becomes a query into possible solution spaces. A functionless approach.

This is already the case with database technology. For example, relational databases use a Relational Algebraic basis for declarative programming. In some data centers a whole database of terabyte size fit in memory.

What is required is a mathematical basis for non-constraint programming. Here is my intuitive notion. A program is a dot in a vast tensor state space. When a user clicks on a button, that dot moves to a different subspace and all it’s queries are evaluated in with new basis vectors. When the dots are related to other dots, we have a system.

Since a program is just a bunch of queries from an “instance” to a ‘space’, it should be possible to use big data techniques (“AI”) to generate that space. This would make it possible to automatically generate applications based on a transition graph of how objects move through space.

Well, I have no time or expertise to make the above work. It was probably some chemical in the soap. Doesn’t seem possible to create such a space.

20180207T0540: Added stuff about automatic program generation.

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.