Incremental makes less sense to me when you want to go to mars. Would you propose to write the software for such a mission in an incremental fashion too?
Yet for software systems it is sometimes proposed as the best way.
We did start small with all of those things. We developed rigorous disciplines around engineering, architecture, material sciences. And people died along the way in the thousands[0][1]
People are still dying from those failures; The Boeing 737 MAX 9 crash was only two years ago.
> Incremental makes less sense to me when you want to go to mars.
This is yet another reason why a manned Mars mission will be exceedingly dangerous NOT a strike against incremental development and deployment.
[0] https://en.wikipedia.org/wiki/List_of_building_and_structure...
[1] https://en.wikipedia.org/wiki/List_of_accidents_and_incident...
In a different way that what is proposed in this thread. We don't build a small bridge and grow it. We build small bridges, develop a theory for building bridges and use that to design the big bridge.
I don't know of any theory of computing that would help us design a "big" program at once.