Four steps to make meaningful progress on anything:
1. Get it.
Invest the time to understand the context for what you're working on. Where are you now? (That's point A.) Where would you like to be as a result of the thing you're working on? (That's point B.) Why is it important to move from point A to point B? What happens if you don't? How do you know?
Begin with humility. Acknowledge that it is statistically unlikely you are working on something new. Someone else, somewhere, has worked on something very similar. So why are you doing it? Who has tried to do this thing before? What did they do wrong? What did they do right?
2. Get it done.
First, do the worst possible version of your project. Second, make it a little better. Repeat that second step a few times.
The idea: spend less time on each version. Do more versions. Call it lean or agile or rapid prototyping or little bets. These concepts rely on a central insight: Pardon yourself from the shame of imperfection, but don't pardon yourself from the shame of inaction.
3. Get it out there.
It's not good enough to share yet. I know. That means sharing adds value to you. You'll learn from the people you share with, which adds value to you. It's ok, you deserve it.
Also, as a bonus: people generally love to be invited to co-create with you. Let them in.
It's an act of insecurity to perfect something before sharing it.
That terrible thing that you believe will happen if you share something half-baked with your boss/client/other-important-person? Run an experiment. Test your hypothesis. See if the terrible thing happens. What if you're wrong? Think of all the wasted opportunities for insight.
4. Get over it.
Work on the next thing quickly, to avoid the temptation of looking back in regret or pride. Don't try to learn from your experience. You can't help it -- you will. The greater danger is learning too many of the wrong lessons.
Move on to that next thing, and learn from the rest of world, not from yourself. Keep a scientific outlook, letting feedback teach you. Current feedback on a current project is a better teacher than pride or regret about the past masquerading as a "lesson learned".