Really, at that time I was actually sort of a Debian fanatic. (Ssh! Still am!) But I also knew that my belief was nearly-religious in its fervor, and I could see others like me about on the internet preaching and being generally ignored in the way you ignore any crazyman yelling on the street. That is, I was such a believer I recognized I likely couldn't have had a normal conversation with anyone about Debian, and so I was careful to conceal it when people asked me about Linux distributions.
It's almost like when a Christian tells you to "just read the bible" (a Mormon friend of Meena did just this the other day) -- I have read parts of the bible, and while there are appealing aspects to Christianity I'd be far more likely to be converted by someone who makes a persuasive argument about its relevance today, not in reading some antiquated story. I think that to some people, the main points are self-evident enough that the only interesting parts to discuss are details like whether adultery warrants stoning or hanging.
In any case, all of that was lead-up to note that I see the same pattern with people who are into unit testing. At work, there's a large posse of 'em and they're always sticking up posters about it. And don't get me wrong -- I've written software that uses tests, and I think testing is great... but that makes it all the more strange to see how repulsed I am by the proseletyzing.
Above the urinals in the bathroom there are pamphlets elaborating seemingly irrelevant details of terminology (is it a test case? is it a unit test or a regression test?) and it makes me simultaneously roll my eyes and wonder what it is that makes communicating these sorts of ideals usefully so difficult.