Open Carbrex opened 4 months ago
In some sense, our result is "more correct" though. Where do we draw the line between "GNU compatibility" and "correctness"?
I had some discussion with some friends about this. I would hope it's not the case, but I can image some people checking the output of seq to see if a number is a proper float. If they do, then I worry about their scripts in general.
But it would be solvable with an optional flag. For example, adding a --bigdecimal
for "more correct" output.
Otherwise, it would use floats with all the floating-point errors left in, thus being GNU compatible.
I think this can only be fixed by seq using float-128 Related #6235 #6182