In all science error precedes the truth and it is better it should go first than last.
The guy who knows about computers is the last person you want to have creating documentation for people who don't understand computers.
