Hi Jerry (and others), My apology for some blatant self-references. During today's NML discussion, you toyed with the idea of defining transfer functions as algebraic functions. We in fact did that, using the Identity function for links and regular functions for adaptation and inverse functions for de-adaptation. Here is the paper: http://dx.doi.org/10.1016/j.comnet.2008.02.013 or http://staff.science.uva.nl/~fdijkstr/publications/G805_Multilayer_Model.pdf A slightly revised version is found in chapter 4 of my thesis: http://www.macfreek.nl/work/ Although I do not see practical use in an OGF standard, it was a nice exercise. If you are interested, I recommend you contact Cees de Laat; the ideas in those papers led to the idea of using Prolog for path finding, and his group is still actively looking into the exact logic that underly path finding algorithms. Regards, Freek