It's nice to have some of the data on web pages be generated by software, instead of just typed by hand.
ASP, PHP, CGI, JSP, Servlets, etc.... It's common stuff.
But what if the code doing the generation is also on the web, as in WideOpenSource? How do you keep your head (and the software) straight about which files are which?
Makefiles give us one approach.
It would be nice to have bi-directional change propagation, sometimes. To have the "Makefile" know that two files are related in a certain way, and if either is changed, the other should be changed to maintain the relationship. Of course this isn't always possible, but it would be nice to have an architecture which supports it when the relationship allows it.
If we think of Wikis (and or filesystems, or the web) as just a bunch of slots where data can appear, ...., the question is merely one of how slots are related to each other, and who tries to maintain that relationship. With Wiki's trust model, things are somewhat simpler: we don't care who maintains the relationship!
... more to say at some point.