[PRL] Joel on Software - Making Wrong Code Look Wrong

[Matthias Felleisen]

So I have read this blog. My reaction is ARGH.

1. Finishing this blog entry convinced me once again that blogs on the 
average are stupid. This one continues to propagate the good old 
anti-academic, anti-intellectual line of programming doesn't need more 
than a dummy mind, and academic CS is for bozos. To wit:

> if you write convoluted, dense academic prose nobody will understand 
> it and your ideas will be misinterpreted and then the misinterpreted 
> ideas will be ridiculed even when they weren’t your ideas.

I bet that he really didn't even want the comma between dense and 
convoluted. He wanted academic prose to stand for "convoluted" and 
"dense" and all the other implications that go with it.

2. It made me almost angry that you guys are spending time discussing 
it. But then it's your time not mine -)

3. Does anyone remember Jonathan Foster and the heated lunch discussion 
among Will, Mitch, myself and a few others in Church Hall? Foster used 
type qualifiers to track safe and unsafe types. Yes, injecting it into 
a flow analysis system will work, too.

4. The argument is incoherent and inconsistent. If you take it to an 
extreme, he argues against procedures as well as exceptions and so on. 
Just think: if you write x = f(x,y); you need to inspect f to find out 
what it does. Clearly that's B.A.D. according to Joel.

5. I won't even touch exceptions. I haven't heard of Raymond Chen, but 
the fact that he supports Joel's rant on exceptions immediately 
disqualifies him as "the best programmer in the world". Exceptions are 
necessary, useful, and good.

-- Matthias