... noes not define any operations called max and min. It has MinNum, MaxNum... would not call them 'min' or 'max'... For an extra complication (for comp.lang.fortran), what should MAX1(NaN,0) return? Note that C, or at least C89, doesn't have max and min, but they are commonly ..., and likely not symmetric. #define MAX(a,b) ((a>b)?(a):(b)) I don't expect MAX(NaN,0.) to be the same as...

..., I found that it assumed max(NaN,0.) = 0. We thought long and hard... was for it to return NaN -- many programmers seem to think of max() as the kind of arithmetic operator which should return NaN if any input is a NaN.... noes not define any operations called max and min. It has MinNum, MaxNum... the language. Personally I would not call them 'min' or 'max'... mfc

hi Had the inspiration when doing data bases a while back, that as well as null, void is also needed. null=unknown quantity void=no quantity max(null,0)=null max(void,0)=0 NaN appears like a null so max(nan,0)=nan ;-) curry shop ahoy!! NaD would be Not a Datum cheers. jacko

... the conflation of two |> meanings for NaN, so it wouldn't be at ...no* definite right answer for max(NaN,0). There IS a definite right answer, using the meaning of NaN that is implied by IEEE 754,...answer, you need a meaning of NaNs that is not currently in IEEE .... Firstly, IEEE 754 requires max(SNaN,0) to raise the invalid exception,... that would imply that QNaN+0.0 = 0.0 and, thirdly, the only...

...this case it is clear that setting max(NAN,0.)=0. will cause problems. It is less clear why setting max(NAN,0.)=NAN will cause problems. It is... is the conflation of two meanings for NaN, so it wouldn't be at all ... *no* definite right answer for max(NaN,0). Now, I probably have less experience ... if I may be indulged a little, what is wrong with... max(QNaN,0) = 0 max(SNaN,0) =...

...>After tracking down a bug in my Fortran program, I found that it assumed max(NaN,0.) = 0. We thought long and hard about this and decided that least surprise was for it to return NaN -- many programmers seem to think of max() as the kind of arithmetic operator which should return NaN if any input is a NaN. But no competing compiler did this, and it ...

... undefined and should be NaN. max(NaN,0.) = NaN Why? Well, because a NaN *could* be plus infinity, or a number ... > If one wants to *implement* NaNs *at all*, one's _reason_ for ... to the trouble of bothering with NaNs, and then deciding that treating them...about the consequences of setting max(NAN,0.)=0 (and John Savard appears to have none) how do you...

...as the outcome of the operation is undefined and should be NaN. max(NaN,0.) = NaN Why? Well, because a NaN *could* be plus infinity, or a number too large... represented. If one wants to *implement* NaNs *at all*, one's _reason_ for doing so... going to the trouble of bothering with NaNs, and then deciding that treating them pessimistically ...

... undefined and |> should be NaN. |> max(NaN,0.) = NaN |> |> Why? |> |> Well, because a ... If one wants to *implement* NaNs *at all*, one's _reason_ for ... to the trouble of bothering with NaNs, and then deciding that |> treating them ...You are wrong on three counts: max/min are not part of IEEE...he has no idea of why NAN's are used and is incorrect...

... |> I.e. doing the opposite which would be to require max(...,NaN,...) to |> always be NaN simply discards everything we know about the representable ... even in the case where the NaN simply means Not |> Applicable, i.e....By providing ways to lose the NaN state quietly, you are rendering them ...agree with you, specifically that dropping NaN information is a _very_ surprising behavior...