on Mon Oct 01 2012, Steven Watanabe <watanabesj-AT-gmail.com> wrote:
AMDG
On 10/01/2012 03:43 AM, Andrew Sutton wrote:
Unless you mean to say that any program that uses a NaN in any way has undefined behavior, this notion of a "well-formed value" seems arbitrary and capricious.
The notion of "well-formed value" comes for Elements of Programming. NaN is given as an example of something that is not well-formed. It's defined in the first couple of pages, IIRC. Any state of an object that does represent a value (e.g., NaN does not represent a real number) is not a well-formed value. I can assure that neither Alex nor Paul would consider the definition to be arbitrary or capricious.
Well, I beg to differ.
Yeah, I have problems with the concept of "well-formedness" as well, because it adds an exception to every statement. It's not that it's an invalid way to look at things, but it merely complicates reasoning about code with no offsetting benefit.
A program that checks for NaN as the result of an invalid computation does not have undefined behavior because it is explicitly checking for that state. A program that uses NaN as if it were a real number should have undefined behavior.
That's no different from saying that a program that uses 2 + 3i as if it were a real number has undefined behavior. If you make an assumption about the value of an object, then it's likely to be undefined behavior if that assumption doesn't hold.
+1
It is easy to construct examples of programs where NaN breaks the postconditions of STL algorithms:
double s[] { NaN }; assert(find(s, s + 1, NaN) == s); // will fail
double s[] { ..., NaN, ... }; sort(s, s + N); assert(is_sorted(s, s + N)); // probably fails
Comparisons on NaN do not conform to the usual semantics of equality and ordering, and so the programs have undefined behavior.
Zero breaks the postconditions of division. So?
A NaN: - Can have various basic operations applied to it with well-defined results. - Can appear in program input and output. - Can be used safely in some generic algorithms.
Therefore, I maintain that a NaN is a legitimate value. Saying that it is ill-formed depends solely on the intent of the programmer in specific cases and is not a universal truth.
+1 -- Dave Abrahams BoostPro Computing Software Development Training http://www.boostpro.com Clang/LLVM/EDG Compilers C++ Boost