« first day  last day (15 days later) » 

Malcolm
Malcolm
10:03
@imallett Essentially you are saying that good programmers will write safe code, and that is obviously true. No one questions that. But it doesn't say anything about C, you even said it yourself that it applies to any language.

Some languages are more likely to trip you up, some are less. Critical software will get written with people who did not receive enough training or do not have enough discipline, that's the reality of life. And the more the language helps to stay safe, the better.
 
8 hours later…
imallett
imallett
18:26
@Malcolm The quality of a tool should be judged by usage by people trained to use it; a power drill isn't something you give to a toddler, but that doesn't mean it's inordinately dangerous. Yes, people who are careless or incompetent get accidentally hurt--and sometimes they hurt others.

I think this speaks more to the sad reality of FizzBuzz-type debacles, and overestimation of self-qualifications in-general. That is, the problem isn't the tools, but the unqualified people who try to use them. I say again, no one else really likes this perspective because both causes have the same effect,…
Roy Tinker
Roy Tinker
I have trouble with the "C is powerful" vs "C is dangerous" arguments. To me, arguments for or against C on either ground are misguided. C's intent is to expose hardware features directly to programmers with some syntactic sugar, full stop. If you need anything other than direct access to hardware, I would suggest looking elsewhere first. Several other languages are "powerful" and "fast" (and tradeoffs include requiring runtime libraries and abstracting hardware away).
 
1 hour later…
Malcolm
Malcolm
19:51
@imallett Well, what I said about unqualified people still stands, no matter how much we wanted it to not be true. As for judging C when used by the qualified people, they make mistakes nonetheless because they are human, and then we have bugs like Heartbleed. @jameslarge made a great point about dynamite, the same thing applies to languages as well. Yes, we need people to apply best practices, but we also need the tools which do the same things in a safer way. One doesn't exclude the other.
 
1 hour later…
imallett
imallett
21:07
@Malcolm . . . right. That's why we *have* other tools, and it should be obvious choosing the right one for the job is important.

(Sorry; indulge me to rant a bit): I'm just saying, most people objecting to C/C++ being "dangerous" aren't really C/C++ experts qualified to make that statement--and the rest have bought into the culture of self-doubt they've created to think C/C++ is something dirty and unnecessarily hazardous. It's rare nowadays to find C++ programmers that don't have some kind of insane guilt over their work, as if they thought C/C++ are statistically more prone to bugs.

« first day  last day (15 days later) »