Perhaps this is one for John Herbster...
Question: does Val(aString, anExtended, ...) produce the same numeric result
as anExtended := StrToFloat(aString) ?
Background: StrToFloat() uses DecimalSeparator whereas val() uses a hard
coded decimal point, and since I want the latter, I think val() is faster
for my purposes -- but I need to be sure that it is also as accurate
numerically.
Alternative A
val(aString, anExtended, dummy);
if dummy <> 0 then raise...
Alternative B
tmp := DecimalSeparator;
DecimalSeparator := '.';
anExtended := StrToFloat(aString);
DecimalSeparator := tmp;
AndrewFG
