Comment on Dijkstra on the cruelty of really teaching computing science (1988)parentComments−sqrt1716yA bridge that depends on all rivets being perfect is a bad thing. A proof that no longer holds outside of its domain of application is still useful.The trick is to recognize when you need one and when you'd rather use the other.−kiujygtyujik16ySame in software - an application that fills a business purpose but has bugs or "errors" is correct.An error free perfect piece of software hat doesn't is useless
Comments
A bridge that depends on all rivets being perfect is a bad thing. A proof that no longer holds outside of its domain of application is still useful.
The trick is to recognize when you need one and when you'd rather use the other.
Same in software - an application that fills a business purpose but has bugs or "errors" is correct.
An error free perfect piece of software hat doesn't is useless