Because sometimes bad architecture is the correct architecture, and good architecture is the wrong choice.
You think we’d have Carmack’s DOOM with “best practice”? Hell no.
I’ve worked for incredibly talented people who choose the wrong technology, approach and design for problems they could have predicted or anticipated, had they seen it or a similar problem before, or had context about the organisation, customer or size of team managing the proposed solution. But they don’t.
AI is great at giving textbook and “Enterprise grade” solutions; but as code production cost reduces — I think the value of simple, digestible solutions will grow. TDD et al can also be helpful, but at a certain point it becomes too abstracted, too complicated, too hard for any one person to grasp.
At that point I think test managers who understand user journeys will be worth their weight in gold, and the real value in any solution becomes good, represented testing.
You think we’d have Carmack’s DOOM with “best practice”? Hell no.
I’ve worked for incredibly talented people who choose the wrong technology, approach and design for problems they could have predicted or anticipated, had they seen it or a similar problem before, or had context about the organisation, customer or size of team managing the proposed solution. But they don’t.
AI is great at giving textbook and “Enterprise grade” solutions; but as code production cost reduces — I think the value of simple, digestible solutions will grow. TDD et al can also be helpful, but at a certain point it becomes too abstracted, too complicated, too hard for any one person to grasp.
At that point I think test managers who understand user journeys will be worth their weight in gold, and the real value in any solution becomes good, represented testing.