At work I dabble in mechanical designs. Machines and systems to accomplish small tasks, like convey product, install o-rings or weld tubes.
We have a layered system that allows people to specialize. Our CNC machinists take a part blueprint and make that part, using just what appears on the blueprint. They don’t double-check the other components to see if they fit or if the product going through, with its inherit variations, will get stuck.
My engineering design might cost ten dollars, and making the machine might cost one hundred. If we need to change the machine after, the costs expound. We try to make everything work the first try. The most embarrassing moments are when I make two parts that bolt together, but the bolt holes do not line up.
Billy and I check for that using design rules on the computer. We have intricate solid models and designs that can be verified automatically using rules that check for fitment and tolerancing. Our boss can see how many of the checks we ran.
Nothing is worse than having an interference in your machine. You design two separate assemblies full of parts, and one of the parts moves into another while the machine runs. It is an error that can be seen blatantly on the computer screen, and yet, you gave the poor machinist the prints and he made the parts; the assemblers put them together and we tried to run the motion.
More engineering- more verification- on the front end helps. More ways to add details and add information into the computer let others review the solution for potential issues, and open the path for Artificial Intelligence checks.
I can’t imagine drawing parts on paper using a pencil and expecting them to ever work together, but in the 50’s people did that.
And that is how I see authors now-a-days. I don’t want to design like someone in the fifties. I want to leverage the computer’s infinite memory and let my thoughts flow, store them, let us review the later.
Parametric Writing is my answer to this call.