Talk:The Bikeshed Anti-Pattern
This is a problem in any group of people, even large corporations. I don't think the answer is to identify useless discussion, because ultimately I don't think any discussion is useless. I think the key is to encourage the right kind of discussion. In the beginning, make the point clear that people should be brainstorming. When brainstorming in a group, that means working in rounds where you only accept ideas and don't allow discussion on the ideas. You want to collect the list of the ideas from everyone to A) provide everyone the opportunity to contribute, and B) provide a directed target for reducing the list. When all of the options are clearly labeled, then they can be discussed discretely and consensus or unilateral decision making can occur.
But projects need to get beyond brainstorming, and people will introduce new ideas late-term, which can have the same effect, "Even though we just started digging the foundation, I really want to go back to the color issue". You don't want to prevent them from doing that, you just want give the right avenue for it to go through. This is where running project meetings with agendas becomes important. Writing agendas is easy: Give 5 minutes to the floor to get additions to the agenda, give 15 minutes to discuss what happened at the last meeting and how things have gone since then, and give 5 minutes per topic on the agenda. Be vigilant about stopping discussion on time, table the issue, and add it to the agenda for the next meeting. Rands in Repose has a good article on this.
It's important to give people a forum, but it's also important to get work done. You can't unilaterally decide that such-and-such discussion is "pointless". You need a process, and the process is either proper brainstorming practice or proper meeting practice. Capn midnight