Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: Reasons (not) to vote Motion 27: NO

On 08/18/2011 02:26 PM, Vincent Lefevre wrote:

On 2011-08-18 13:33:59 +0200, Arnold Neumaier wrote:

In any serious programming there is risk of mistakes, and one needs
to know some precautions to avoid the most common fallacies.


But here the risk is even more important, because the behavior isn't
intuitive and existing programs would need to be modified. Indeed,
existing programs do not use decorations (or at least, not the ones
defined in Motion 26), and the problem here concerns the decorations
only.
But the risk is easily avoided by following the instructions in the standard (which can be emphasized by giving your example). 





Note also that a good program would catch an emoty initial interval
before even trying to evaluate some expression. And an empty
interval created during a range evaluation doesn't cause problems.


A program should not need to do that. More precisely, a program
should not need to do anything to make the decorations correct.
To have correct decorations, the assumptions needed for proving correctness must be satisfied. There is no way around that. 


The only reason why a programmer would change his program to get
a different decoration is because the information given by the
decoration is not optimal. For instance, ones normally gets a
decoration saying that the function is defined, but from the
context, the programmer knows that the function is continuous,
so that he may wish to set the decoration explicitly. But if one
needs to catch the empty interval before evaluating an expression
only to avoid an incorrect decoration at the end, then the system
is poorly designed.


Decorations are designed not for arbitrary computations with intervals,
but for the range evaluation of a function with accompanying guarantees for this function in the domain on which it was evaluated.
You don't even define a function, hence are strictly speaking outside the domain of applicability of the decorations. 


Arnold Neumaier