Maybe that's the only possible way. More than a few systems have failed from over-planning and under-iterating.

And they do say cities are the greatest invention of humankind

Most successful systems are 1st & foremost practical. They do a job, and do that well enough to be carried forward.

Imho the "Keep It Simple, Stupid" (KISS) principle very much applies here.

It's even how we as biological systems work, there's plenty of evidence of old stuff and iterations still within every body and within every cell. This even was an explicit programming paradigm in the 60s and 70s, "extensible programming", for example driven by a lot of the designers of languages such as Smalltalk, aiming to enable systems that can grow not unlike living systems.

I'd say ruins are a precursor for cities to become beautiful--computation concerns similar strolling along age old paths and alleyways.

Nothing is built on stone; all is built on sand, but we must build as if the sand were stone.

Jorge Luis Borges

This captures such truth to me.

And it also defined the moral imperative for software to be malleable, for it to be soft and reshapeable. Needs change, and it's not just a brand new thing we slap on top, but something later to & extending the past. Whether we can do that, whether users have software that can be adapted to their contemporary needs, is a core liberty. https://malleable.systems

I feel like there is a duality here with:

"A complex system that works is invariably found to have evolved from a simple system that worked."

The personal computers start as toys and tools of the hackers. I couldn't find any other way to create it. It was definitely without much grand planning.

At the time of the writing there was actually a lot of planning going on before you started writing code.