Comment on Dijkstra on the cruelty of really teaching computing science (1988)parentComments−kiujygtyujik16yThats the difference between maths and engineering.A proof that has a single exception is wrong, a brige that stays up with a broken rivet is correct−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−Twisol16y> A proof that has a single exception is wrong, a brige that stays up with a broken rivet is correctI really like how the mathematical part of that phrase is literal, and the programming part is metaphorical.
Comments
Thats the difference between maths and engineering.
A proof that has a single exception is wrong, a brige that stays up with a broken rivet is correct
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
> A proof that has a single exception is wrong, a brige that stays up with a broken rivet is correct
I really like how the mathematical part of that phrase is literal, and the programming part is metaphorical.