[benjaminjosephw]

> if you truly want a 'low-code/no-code' thing, you're stuck with no-logic no-interaction software

I think this used to be the case but the landscape is changing quickly. Deepnote[0], for example, looks like a really interesting programming interface that's not quite traditional programming and lowers the barrier to entry significantly and is (arguably) in the low-code space.

Low-code products for data engineering show that we're not too far away from these sorts of solutions in a more generalized offering. Spreadsheets aren't the only answer.

[0] - https://deepnote.com/

[sidlls]

The landing page does not show a low/no code solution. It shows what is effectively a (possibly) more sophisticated IDE wrapping a Jupyter notebook. One might try to argue it's low code, but if their landing page is showing something that isn't low code, I have doubts.

I think better examples might be Databricks, Tecton, H20.ai, or Domino Data Lab. These aim to provide "drag-n-drop"/no-code implementations for the "boring" infrastructure related work in most typical data engineering/analytics applications, while offering a set of tools for more sophisticated uses (e.g. the ability to supply one's own python functions for ayptical or custom data engineering).

Even these fall short, in my view: they're more like platforms on which to build a data science/ML infrastructure--which requires a substantial amount of engineering effort.

[oneplane]

The landscape can change all it wants, but as soon as you put enough complexity together, you end up with programming and even 'real' development and engineering at some point. This isn't even software-specific, it happens to any system that is sufficiently different from a generic catch-all solution.

In some cases it might be beneficial for some people to 'program' using pictures and arrows etc, but if you're in a situation where you need some complex process modelled and automated, the complexity doesn't just 'go away' because a vendor said 'low/no code!'.