|
![](/i/fill.gif) |
Darren New wrote:
> Orchid XP v8 wrote:
>> Darren New wrote:
>>
>>> Now, how many bits do you need in your floats to represent that? :-)
>>
>> Depends. How many bits of precision do you want? :-P
>
> Well, that's the trick. It goes from "smaller than makes sense to talk
> about" to "bigger than the universe." That's telling you how many bits
> of precision you need.
>
> "Smaller" as in "it's nonsensical to talk about distances shorter than
> that because physics doesn't work that way", just like "there is no
> distance bigger than the entire universe."
Sure. I get that. And if you were asking about *fixed-point* arithmetic,
we could just take some logarithms and get an answer. But since you said
*floating-point*, all we figure out is how many bits the exponent needs.
That still leaves the question of how big the mantissa should be. ;-)
--
http://blog.orphi.me.uk/
http://www.zazzle.com/MathematicalOrchid*
Post a reply to this message
|
![](/i/fill.gif) |