On Tue, Nov 25, 2003 at 04:09:18PM -0800, Jason Thorpe wrote:
> 
> On Nov 25, 2003, at 4:06 PM, Jun-ichiro itojun Hagino wrote:
> 
> > 	you are voting for stop skipping random number of ids?  i can live
> > 	with that.
> 
> As long as it does not weaken the algorithm.

IIRC that skip is the only thing that makes inversion non-trivial.
For 16bit numbers (where it is trivial to ennumerate over two values, and
three isn't a big problem) you probably only need 3 or 4 numbers
in sequence to work out what the next one will be.

	David

-- 
David Laight: david@l8s.co.uk