I was somewhat confused by the structure of this paper. It started off by talking very simply about the problems of teaching someone to use the internet. From there it went into a very confused and over-complicated segment on mental models and constructivism. Finally, it ended with the same simplicity and also involving the internet.

Aside from the apparent poor construction of the paper (in my opinion) I think there is a good point or two lodged in between the fancy words and diagrams. The idea that laying new concepts upon old concepts can be very beneficial is important. But it also seems obvious. I mean, the first thing I do whenever I am teaching someone on hours, or whenever, is to try and understand what I can relate the material to that is comfortable for them. That way the student is able to translate what they know about a certain topic into the strange new world of the topic you are trying to teach. I have personally had much success with "teaching-by-analogy" and do not really see anything new or revolutionary in the idea.

Other than to present this idea I am not really sure what the author's goal was. Is it a good idea? Yeah, I think so. Done.

I'm not sure how obvious that teaching strategy is. It definitely makes sense to try to teach a subject in a way that people could relate to, but when you are trying to teach a person a subject like CS, I don't think it comes naturally people to try to find out what that person is familiar with, and try to tie CS in with that. I know it doesn't really with me. I think it is much more natural to try to teach something in the same way that you learned it, since that is how your brain works. Tryin to crawl into someone else's brain is tough.