Rosalind - 9:40 pm on Nov 17, 2009 (gmt 0)

There are different versions of speed: the overall load time, and the load time per KB of viewable content. There's also page rendering time.

Personally I think overall load time is a flawed measure, because it fails to take into account just how much useful content you might be trying to display. So you may have an 80kb image, but is it an intricate and attractive picture or just some badly-optimised logo?

If you take the load time and divide it by the number of kb you might get a figure that has a little more to do with having optimised code and a good server, but I'm willing to bet there are more than a few flaws with that way of measuring speed as well. For instance, how do you distinguish between useful javascript functions and pointless code bloat?

Perhaps I'm becoming cynical in my old age but I suspect there's more to this than what the official line suggests.

Maybe the true agenda here is to get us to ditch all of those bloaty, useless nofollows.