I spent a long, long time working on an open source reimplementation of Diablo 1 (https://freeablo.org), as a clean room reimplementation. In the end, a new project called devilution cropped up that was based on decompiling the original binary. Because of that, it was playable day 1, and started to build real community and momentum. Seeing their success killed my motivation, and the project has languished since. I never actually cancelled it though. It's still kinda painful for me to think about.
I did learn a lot from it, and it contributed majorly to every job I ever got, so it's not like it was a complete waste of time.
If I had to share a lesson: don't invest too heavily in someone else's IP.
Built a platform that aggregated results from a variety of APIs, along with native results from the platform itself. The largest API provider we used decided to make their API private to everyone except for the partners they signed deals with. Those partners' platforms then were the only platforms that had the results the majority of users were looking for.
Although it was a completely different market, pretty much every social media platform does exactly this now with their APIs. First, they open their APIs up to everyone in order to gain marketshare. Once they've cornered the market, they make the APIs private or useless so that competitors can't leverage them.
However, the platform owners recognize the value that some companies using their APIs provide, so they hold those API consumers hostage and milk the best deal they can get out of them, and kick everyone else out. The only way to become a partner after that is to have enough funding to sign a nice deal with the platform owners, assuming those owners don't just copy potential partners' apps wholesale.
You can see this trend on Twitter or Instagram, where there are a handful of "blessed" companies that are allowed to automate and consume content on those platforms using the platform-supplied APIs. Companies that automate or consume content without being blessed are regularly blasted off of those platforms entirely, especially if they can be thought of as competitors.
Fwiw If you were to put together a book, site, or course describing the overall architecture, the design decisions you made, the rationale, possible improvements or alternatives, etc., I’d buy it. Something like pikuma.com, or handmadehero could be really interesting.
I’m not a game or graphics guy, but that kind of stuff has always fascinated me.
If it makes you feel any better, these projects derived from decompiling binaries infringe copyrights in the original and can be shut down whenever the owners decide to assert their rights. Your project is at least your own work.
I did learn a lot from it, and it contributed majorly to every job I ever got, so it's not like it was a complete waste of time.
If I had to share a lesson: don't invest too heavily in someone else's IP.