On 07.11.2010 15:10, Phillip Gawlowski wrote:
> On Sun, Nov 7, 2010 at 2:08 PM, Michel Demazure<michel / demazure.com>  wrote:
>> Phillip Gawlowski wrote in post #959912:
>> @Phillip : I am not sure I agree. When sorting on the second half, the
>> couple (["a", 3], ["b", 3]) falls under the "equal" case of the choice.
>
> "a" != "b" results in true, thus the couplet is *not* equal.

No, Michael is right.  Every sorting algorithm only ever looks at *sort 
keys*.  So since our sort key is the second array member (the number) 
both mentioned records are equivalent.  If you include the other field a 
stable sort isn't really interesting here since then all fields are part 
of the sort key - stable and instable algorithms will return the same then.

Kind regards

	robert

-- 
remember.guy do |as, often| as.you_can - without end
http://blog.rubybestpractices.com/