I've tried to use Nix multiple times for things that it would be perfect for... I really want to use it (this is the 5th time I've given it a shot) ... and ever

I've tried to use Nix multiple times for things that it would be perfect for... I really want to use it (this is the 5th time I've given it a shot) ... and every time I put it down. You've built a great tool here, seriously nice work! ... but you built it on an experimental feature that isn't enabled by default, that requires understanding the extensibility mechanisms to add experimental features that the Nix project doesn't feel comfortable calling done yet, that may change in future ... except everyone says flakes are totally stable and fine to use cause. Like basically everyone seems 100% fine with flakes... to the point where I've even come across first time Nix user tutorials that assumed you had already jumped through the hoops to enable an experimental feature a.k.a. "flakes" and this was several months ago. But if flakes are so amazing and so stable everyone is using it and assuming everyone uses it and wants everyone to use it... after an entire year of wide public use ... why aren't they a stable feature yet, or even planned for release as a stable feature? Once Nix can get this sorted, I'll probably be using it within a week. But I'm not comfortable diving in and making heavy use (I've played with Nix at every layer, from my macOS user as a homebrew alternative, to Docker/Container building, to running entire headless and desktop immutable OS installs on bare metal using NixOS) ... but for now its still an experimental feature so once again I'm putting Nix back in the "check it out again later next time HN gives me a good reason" drawer... I'm not going to rely on experimental features and I'm not going to learn and internalise how to do everything the hard way when flakes is obviously better (I did actually RTFM on all this), I just keep waiting for it to be blessed as stable.