A somewhat ironic title; those in the programming arena who are serious about correctness are often viewed as "grumpy". EwDijkstra
is well known for being serious about correctness.
The problem is his style. There is an interesting give-and-take on this in http://www.literateprogramming.com/best/luminary.html
Perhaps the grumpiness is the result of having seen much buggy code that is buggy due to carelessness.
Another reason might be that for all the wonderful comfort correctness gives, there are few languages that support proof-of-correctness that also support development in the RealWorld
(tm). Correctness appears to have all the problems of BigDesignUpFront
thinking - the program, however useful it might be to the customer, is not good if it can't be proven correct. -- PeteHardie
(inventor of ForthLanguage
) is my nominee for GCSITW. He's a fan of small, lean software (mostly because it runs faster and there are fewer places for bugs to hide). He doesn't like the ANSI standard for Forth: he thinks you should write your own. And he does.
thinks modern OS/apps/processors are too big and bloated, and there should be as little as possible between you and the hardware. Hence he writes his own language and OS (ColorForth
), and designs his own asynchronous Forth MISC processors using his own homebrew CAD system. Heck, he even designed his own chording keyboard! Quite the iconoclast, though I don't think he would consider himself to be a computer scientist.
In the case of EwDijkstra
, it's not just a passion for correctness. Read his manuscripts. He seems to hate everything that has happened in the last 40 years. I spent a weekend reading through them and I got the impression that the world has simply passed him by.
His writing gives this impression. When he spoke, his tone of voice was not grumpy but gentle and wistful.
Given the bug-laden garbage that we all have to deal with daily, we can hardly blame him for being grumpy. Be careful not to fall into the "just doesn't get it" trap of the dot-com mentality when dismissing him.
He would definitely not like the code behind Wiki, unless you can show that you inducted the code mathematically and have proven it correct. He wouldn't like the language it's written in. He most certainly would object to ObjectAnthropomorphicPrinciple
is my nomination. Read TheThirdManifesto
is a good candidate. RichardStallman
as well, though not for purely technical issues.