I have a few.

partkart.com (now defunct) - it was supposed to be thingiverse but for cnc projects. I made it one summer when I was in school and had the naive notion that users would find it organically or through cnczone. I think one or two people used it total after a year. Learnings: have a plan for user acquisition.

partkam/makercam - browser based cam system, made during that same summer. Some people actually used this, as it solved a legitimate problem. I wasted a lot of time developing a custom polygon offset algo, and it was written in flash/actionscript cause it just had to be a web app. Learnings: don't develop for dying platforms, use libraries.

deepnest.io - nesting app for laser cutters. There are similar commercial apps so the market exists, just not sure how to reach the SMBs that are presumably the end users. Learnings: don't make an app that requires high touch sales if you have no idea how the industry works.

final learning: I actually enjoyed working on this stuff and learned a ton, even though they weren't ultimately successful. Having them contributed to getting a job after I graduated and later, more successful projects.

I wrote a new Java IDE for education called SnapCode: http://reportmill.com/snap . I've spent several years on it - and it's hard to move on. There should probably be a support group for this. :-)

https://www.failory.com/ collects startup founders' failure stories. For me personally it was luckily only domain and some hosting costs. The biggest investment was and will be time spent (opportunity cost to work or freelance on other projects).

Kind of ironic that a project to collect failure stories failed itself.

This is a good website. I hope you can keep it up.

Like Edison said, "I have not failed. I've just found 10,000 ways that won't work."

Sometimes, you learn the most through failures.

I've read this site a lot and enjoyed the content - keep going!

https://www.zipbash.com social network based on zipcodes. I was thinking it could be alternative to nextdoor where you can't post outside of your neighborhood. I also made registration optional, so you can post and comment anonymously. Also wanted to test out if people would pay $1/month for premium features I'd make in the future (private messaging...).

Anyway, most feedback I got was negative, so that's depressing. Maybe if I got proper feedback I could modify it for better.

I'm still working on it, there's a big change coming, not sure the outcome will be. Hopes.

What were the complaints?

Mostly about how the UI sucks, which I agree with but I don't have a creative UX skills. I'm mostly a data/business logic person.

Fix the UI, and start paying mturks to post topics from various locale's FB groups to get you off the ground.

I have gone down that rabbit hole one too many times.

I think finding an existing audience or building one through teaching are a better approach than building first.

Talk to potential users using open ended unbiased questions is important. When you get specific with questions or features, you end of creating a bias that will only give you bad data and set you up for potential failure.

$6000 coffee machine for a cafe. Well, the demand was there. It's just that I didn't like the people involved with that career path.