2 ms·
Check it out if for no other reason than to read the rant about how poorly defined the candela is.
by btilly 16y ago
Check it out if for no other reason than to read the rant about how poorly defined the candela is.
- DrJokepu 16y agoAlso, it's really hard to resist the temptation to "rant back" on the kilogram question. Unfortunately / fortunately, HN comments are not the right place for that. (Basically, kg is the perfect unit of mass and if you disagree you are a heathen and should be burned)
- btilly 16y agoHis complaint is that if kg is the perfect unit of mass then make what we call the kg be the base unit and call that a gram instead. Makes perfect sense to me. Of course this is an imperfect world. In a perfect world I'd like a meter chosen so that gravity is, to a good approximation, 1 m/s at sea level, and I'd like all of the units to be in base 12. (Easy divisibility by 3 is often very useful.) Including time, so that one doesn't have the awkward 3.6 factor converting from m/s to km/h. And our number system should also be base 12, and...I think you can see that I'm willing to complain about lots of things which will never change.
- eru 16y agoI want my number system to be in base 11 or some other prime number. You see: The vast majority of numbers are never going to evenly divide our base, no matter what base we choose. So we might as well go for a solution that minimizes the special cases.
- btilly 16y agoThe vast majority of the time that people wish to divide something evenly, they want to divide it by a small number.
- eru 16y agoYes. I just wanted to give the standard mathematician's tongue-in-cheek reply to proposals like base 12. To be earnest, octal is quite a nice base. Or -2 (minus two!). Or 10, but with digits from -5 to +5, though that representation should be redundant, I guess. To speak of redundant representations: Fibonacci numbering is also interesting (http://en.wikipedia.org/wiki/Fibonacci_coding http://en.wikipedia.org/wiki/Fibonacci_coding). Or skew binary numbers.