It's tempting, especially in the field of graphics, to completely ignore stuff from before 2000 or so.

It's also really stupid.  What we do now is just a faster version of what we did then (with the partial exception of shader language programming) - the underlying techniques haven't changed.

And people said really smart things nearly ten years ago which are, believe it or not, still very smart.  Steve Maddock is, i gather, a lecturer at the University of Sheffield now but when he wrote what i'm currently re-reading, he was a grad student.

This does not change the fact that it's a fine thesis (and well-deserved PhD unlike most, including mine) and says important things about the way graphics and animations are indivisible.

What i'm currently working on, of course, is a different way of constructing those graphics and you know what?  I've discovered that it's not that hard. Okay, it is that hard but i know how to do it.  However, adding bones without manually doing so, animating via inference, that's a whole 'nother bucket of squid.  That's so hard that i've had to throw away what i've done on the subject and go back to basics.

So finding Dr Maddock's thesis online has pretty much made my day. 

Woot.  Highly recommended.