Matthew Toseland wrote:
> We already do this.

Ah, cool. :-)

> I presume that the latter tactic would be best, since we don't know the 
> node's 
> location: send a swap request with our location set to m, and all our peers 
> at (m + 0.5) modulo 1.0.

I see what you mean. Perhaps a node should only attempt to reduce its 
own stress (in Robert's terms), rather than the product of its stress 
and the other node's stress?

Cheers,
Michael

Reply via email to