Comments

Paul Andersen | March 31, 2003 12:43 PM

I know this is an old article, but I'm going to comment anyway. I've been keenly aware of the "time" issues around employing classification schemes through previous projects and found it interesting that you were trying to formulate some representation of layers of IA to illustrate the relative difficulty of changing the various layers. Definitely provides a nice way to illustrate to clients the importance of identifying pre-coordination functions vs. post-coordination functions.
Have given more thought to this? Refined your thoughts a bit more?

Peter Morville | March 31, 2003 2:53 PM

I haven't had time to dig into these "pace-layering" issues in more depth, though I hope to soon. In the meantime, check out the bad Peter's thoughts on layers:

http://peterme.com/archives/00000320.html
http://peterme.com/archives/00000321.html
http://peterme.com/archives/00000323.html

David Locke | December 11, 2003 5:14 PM

What I'd like to see is the order of the decisions. Which decision do you make first? Which decision do you make last?

The decisions will fall out into the layered shearing model. We do go back and modify decisions. We will change not only that decision, but all the dependent decisions in the faster shearing layers.

The figures have omitted how long it takes to do the work in each layer. It seems to me that the temporal geography would be complicated by a process that take longer to do, than the element created by the process. So how long do the processes take in industry average numbers?