Bringing the big brain out on me! This is off the cuff.
I was not aware of Universal Design principles but a quick look suggests Principle 5: Tolerance for Error is most applicable to Software Engineering. Why not 1-5 bedrock? Because, in my opinion the general state of software is that it is more functional yet just as unreliable in decades past. What is the first thing a little experienced user to do when an error occurs? Yeah, they quit. No access. The micro-service paradigm has made the situation emminently worse as even finding the blame/responsibility for “no access” is fruitless.
2a. With anything of this type, the most obvious risk is to my own reputation. Security is a field burdened with responsiblity, people come to rely on it, what if they get hacked using my repo? I only took on things I spent months understanding and testing absolutely everything by hand. I limited myself to only distributions I could juggle, use daily, so I could be responsive to needs.
2b. Risk is competing objectives. FreeBSD and thus it’s reliants, Ghost and Dragonfly, are in a strange position right now. FreeBSD is Linuxifying itself and adding more Corporates Sponsorships than ever in a path away from traditional BSD security. This presents itself a potentially competiing ethos situation for me, but not yet.
2c. OpenBSD is used by world security intelligence agencies and I hear the DoJ. Am I without my knowledge picking sides here and favoring some entities over others? Famously DARPA and FBI backdoor right? I researched the OpenBSD Sponsorship list carefully and asked around. The OpenBSD availability (at least of the version we use!) is equitable and I purposely put out an OpenBSD honeypot to see which entities would try to compromise it! Results: Fair.
This is a can of worms because what we are really talking about is the Linux-Effect. Started out community home-grown to now be a Corporate Globally Mega-Corp sponsorship vehicle estimated to be worth $100 Billion. Even Apple is now a Silver Linux sponsor. What am I saying is Corporate dominance is think-tanking and policy making. Data selection is inherently profit focused instead of Humanity Progression focused. Bodies like the UN, EFF, et al. are wholly ineffective.
The paths forward on that are gruesome to be honest as what would be best would be something like a randomly selected group of High School Science Fair finalists and Waitresses to form a Governance body with teeth to dissolve Corporations completely for profiteering off populace private data, genetic data, financial data, and the engineering decisions that are ubiqutously driven by them when determined that a Corporation or other Government body is acting against out future.