4 ms·
I completely agree that "good design"/"bad design" depends on the problems we're trying to solve, rather than some abstract property of the entities involved (i
by chriswarbo 4y ago
I completely agree that "good design"/"bad design" depends on the problems we're trying to solve, rather than some abstract property of the entities involved (i.e. whether "Car" should extend "Vehicle" or not, and whether those classes should even exist, cannot be answered without knowing what we're trying to accomplish).
The suggestion of GUI widgets is certainly better and more concrete; although it's skating dangerously close to "Circle extends Ellipse" and "Square extends Rectangle" territory! (e.g. https://softwareengineering.stackexchange.com/questions/238176/why-would-square-inheriting-from-rectangle-be-problematic-if-we-override-the-set https://softwareengineering.stackexchange.com/questions/2381... )
- joshka 4y agoMuch agreement on this. It's concrete, visible and has behaviors that work well (click, resize, draw etc.) as well as providing a good entry into the thought process when designing modern user interfaces with components.