The biggest mistakes I made with YouGame and P+ were honestly communication and a lack of consideration. So I want to apologize again to the P+ and Brawlback communities because of that. I announced a 4-week-old prototype as if it was a finished product instead of what it really was: An early alpha build in need of a lot more testing, refinement and community input to become great. And I didn’t publish the repo and make it clear before announcing that it was a fully open source, clean fork of dolphin that used a totally different rollback architecture to Brawlback (although this architecture and a few files of code have now been adopted by Brawlback as well.) Things probably would have gone a lot better if I had. I still believe that AI is an incredible tool for building games that will bring about a revolution in gaming, helping us build games and experiences we only dreamed of as kids, but its default creative output is slop and it takes a lot of human refinement, taste, vision and testing for it to become great. People’s feedback is huge in helping with development and quality, but it was wrong to not be more transparent and respectful of people's time and experience. Early build testing can be messy and frustrating, so the communication should be very clearly targeted to people who are interested in dealing with that messiness and contributing their time and energy to an unpolished project so that it can become high quality enough to bring everyone else in. So that's the main thing I'm changing moving forward. Very clearly labeling early builds, getting groups of internal testers to try things out before alpha releases, and only making big public announcements/events for builds that have already been well proven to work reliably.