Yes, log^2(x) seems almost always to be used to mean [log(x)]^2 in the routing theory papers I tend to read. I agree with you though that using log^2(x) for [l

Yes, log^2(x) seems almost always to be used to mean [log(x)]^2 in the routing theory papers I tend to read. I agree with you though that using log^2(x) for [log(x)]^2 is horridly ambiguous notation that should be avoided! (Which reminds me, I should go and check my thesis for this ;) ).