What every C Programmer should know about undefined behavior #2/3 [LWN.net]

What every C Programmer should know about undefined behavior #2/3

Posted May 17, 2011 12:11 UTC (Tue) by dgm (subscriber, #49227)
In reply to: What every C Programmer should know about undefined behavior #2/3 by jmalcolm
Parent article: What every C Programmer should know about undefined behavior #2/3

C, Unix or Ada, none of them can ensure the system is completely free of errors. Nothing can. Some languages can help reduce "certain" kinds of errors, often trading-off execution speed and/or generality... and introducing subtle, new kinds of errors.

to post comments

What every C Programmer should know about undefined behavior #2/3

Posted May 18, 2011 12:57 UTC (Wed) by marcH (subscriber, #57642) [Link] (4 responses)

> C, Unix or Ada, none of them can ensure the system is completely free of errors.

This is really a pointless comment. Here is a similar one: Nothing can prevent the best burglars to break into your house (so why buy an expensive lock?)

> Some languages can help reduce "certain" kinds of errors, often trading-off execution speed and/or generality...

Yes.

> and introducing subtle, new kinds of errors.

No.

What every C Programmer should know about undefined behavior #2/3

Posted May 19, 2011 15:22 UTC (Thu) by dgm (subscriber, #49227) [Link] (3 responses)

>> C, Unix or Ada, none of them can ensure the system is completely free of errors.

>This is really a pointless comment. Here is a similar one: Nothing can prevent the best burglars to break into your house (so why buy an expensive lock?)

I would not call it pointless, but I agree it's rather trivial. Anyway, it's useful to keep it in mind when listening to vendor's preaching the latest silver bullet.

>> and introducing subtle, new kinds of errors.
>No.

The logical conclusion would be, then, that a "perfect" language that prevents any kind of error is possible, which is absurd.

What every C Programmer should know about undefined behavior #2/3

Posted May 19, 2011 23:45 UTC (Thu) by marcH (subscriber, #57642) [Link] (1 responses)

> it's useful to keep it in mind when listening to vendor's preaching the latest silver bullet.

Let's keep the trivial statements coming: every vendor is preaching the latest silver bullet. It's their job, they are paid for it. Their lies does not mean every product sucks.

> The logical conclusion would be, then, that a "perfect" language that prevents any kind of error is possible,

Your logic is really beyond me.

What every C Programmer should know about undefined behavior #2/3

Posted May 20, 2011 3:01 UTC (Fri) by viro (subscriber, #7872) [Link]

> Their lies does not mean every product sucks.

Their lies do not mean that water is wet either...

What every C Programmer should know about undefined behavior #2/3

Posted May 20, 2011 6:10 UTC (Fri) by dark (guest, #8483) [Link]

The logical conclusion would be, then, that a "perfect" language that prevents any kind of error is possible, which is absurd.

That doesn't follow. An alternate conclusion is that even the subtlest errors are already possible in existing languages.