+1

@ O.P: I read your entire post (thanks for relaying your lessons), and most of it is about internal development and personal "aha" moments.

What about customer driven development , i.e. user testing?

You have no sign up form ...

This boggles my mind - get users already!

This is a classic case of what I personally refer to as un-scientific thought processes. If I ask my my friend "what do you think of this business idea?" Really, what on God's green earth does it matter what he thinks? If he says "its great", does that empirically mean anything? If he says "i hate it", does that empirically mean anything? Who can empirically say anything is anything? In other words: test it. A joke I have with all of my friends; anytime they ask my advice about anything, they already know the answer... test it. You have not tested your product.

So your question: "What should we do?"

My answer: "Test it."

Thanks for the notice. Our problem is that I've tried to build lot's of features at the same time. The app became too big, it eats a lot of my time and it's still far from being completed - it's still not ready to be released.

Release early, release often. Once you have some users, they will be the best to guide you to what is working, whats not working and what needs adding. If you specify its beta then they can hardly complaing (espically if you offer some beta testers a discounted rate after release, positive motivation on both sides wont hurt)

also powerfull=powerful

Try http://spellr.us/

Also, my first thought was that this was like DabbleDB. Is it?

Yes, Integrity can be compared to DabbleDB, actually we have lot's of potential competitors, you can read the blog below to know more about the market

http://webappsatwork.blogspot.com/

DabbleDB is one of those products that I will probably never get to use despite being potentially very useful. Why? Because my company won't let me store proprietary information on somebody else's servers.

If you continue with this project, I would encourage you to consider letting people license the software to run on their company intranets. I know a lot of companies don't want to bother with installation support and such, but to me this seems like a pretty obvious market -- there are countless CRUD apps I could really use at work but just haven't had the time to build from scratch, and if DabbleDB had a not-outrageously-priced intranet solution, I'd buy it in a heartbeat.