Personally, I find Tanya Khovanova's explanations (also linked to by the article) much easier to read: she's discussed this and related problems quite a few times.

http://blog.tanyakhovanova.com/?p=221