Ideas are to be shared. Good, quality code (with an appropriate license for attribution) should be shared as well.
You might ask where I draw the line? I draw the line at whole applications (I would never give someone the entire source to a webapp I charged money for), or physical goods that I need/use (servers, hardware, etc).
I had put on the facebook page my idea and explained that I am doing it to learn how to program myself, as my day job is creating the best customer service in the world, and has never really touched on writing code myself.
I was nervous about sharing my idea because of the fact that anyone who has been learning to program for more than 5 days (I am on day 4) has a huge head start on me and could launch the idea way quicker than I could.
However, What I got back from the community was more encouragement and enthusiasm. And I even was politely asked by one of the guys if I wanted to partner up on the idea as he seemed to really like it, instead of him just taking it for his own.. as he is studying comp science in university i think.
This being said, I am a huge advocate on the fact that every idea has been thought of before and people are going to make the product whether I tell them or not.
So in conclusion, I think telling anyone who will listen about your idea is the right move. You don't have to tell every detail of course, but the feedback, encouragement and opinions of the people who would listen to me have been a big help in this project for me.
That being said, I really look up to them and hope to keep improving and becoming the best :)
The only reason I've ever been 'stealth' with an idea wasn't because I didn't want to share the idea; it was only because I didn't want to talk up an idea that I only ended up taking half-way.
And I agree with everyone here; almost more important than the idea is your ability to explain it - your pitch. The best ones are natural and conversational, and that takes telling and retelling. The only secrets I keep are the confidences of other people that share their ideas with me in private. Besides, you telling someone else your idea is a datapoint for yourself in regard to your own enthusiasm for the idea.
I'll even email a company, tell them I have this idea for X (that relates with Y which they already have) and ask them if they'd like to hire me to help them build X. Sometimes that will get you a nice contract job. Most of the time it will fall on deaf ears. I haven't had an idea stolen yet this way.
I have a lot of stuff marked "Later".
With regard to someone stealing the idea, I don't really worry about it. I regularly give ideas I've had away, especially if I think I'll never make them.
A few years ago a friend and I were at a restaurant and there guys were trying to figure out something to build on the iPhone, my friend and I had just come back from an amusement park in southern California and I suggested a custom map of amusement parks that has points of interest for all the restaurants and rides and gets you there within the boundaries of the pathways in the park. I had the idea earlier that day, but I would never build it myself, so to me it was fairly useless.
(Brain crack discussed here:) http://www.zefrank.com/theshow/archives/2006/07/071106.html
I'd have to have been in the park a lot to map out points where you could walk, write the algorithms for navigating the park from any random point of interest to any other random point of interest in a reasonably efficient way while following the boundaries of the pathways. I'd have to have kept the app updated when anything changed at the park, including things like "For two weeks, this one path is blocked off because construction is happening" which happens fairly regularly at Knott's Berry Farm.
Although those are all challenges I could have overcome given the time and resources, it wasn't something I would be very passionate about and it would die out as a result.
I've had a lot of reasonably good ideas, and some of them have worked out rather well in the past. Most of them I don't pursue more than as an idle thought. Those I choose to work on I do as a result of how much fun they will be, how passionately I feel about what I'm building, and how challenging I think it will be. A few weeks ago How I Met Your Mother inspired the idea for http://www.notanewyorker.com/ and that was something I said, "I'll build it tonight, I have nothing better to do." It's a cute little website, and the biggest motivating factor was I thought it would be funny, and using Python on Google AppEngine would be interesting because I've not used it to build anything like that. I found I hate Google AppEngine now though.
Works well, saves time.
However, the chances that your idea is unique and particularly special are small. The chances that you'll be able to protect your idea from someone who really wants to steal it are also small.
You are much better served by sharing your idea with a lot of people (might as well make them people you trust, no reason to go out of your way to have someone take your idea), gather as much feedback as possible and build build build.
Your time and effort is much better spent making progress on your idea and moving faster than any idea thiefs than it is worrying about idea theft and trying to protect your idea.
Ideas are generally worthless, everyone has them and on their own don't mean much. Build something of value, then worry about protecting that.
Ideas don't matter. Execution does.