FIPS looks like a fantastical dream right now. They say that a general law of systems is that a complex system that works just evolve from a simple system that works. And a complex system that doesn't work can never be made to work. The success of FIPS will depend on how disciplined its developers are in keeping it as simple as possible. But with so much of the code being written by AI, I am not very enthusiastic about it. Pair that with the fact that it has to be private and secure puts even more pressure on it to remain simple. Software bloat will be its biggest enemy, in my opinion. Only the developers and Claude will understand how it works and how to fix it if it breaks. For a base-layer sovereign networking technology, it would be nice if it was independently maintainable rather than centralizing that risk on a monolith repository. Looking at the code, the node orchestration module is the most complex and the most intertwined. If there are problems down the line, breaking that down into smaller, more maintainable chunks would be prudent. Better yet, separate repositories for each target runtime with a shared core, and a separate e2e testing repo. @FIPS

No one has ever made a decision based on pure logic or reason. But it's comfortable to think so.

Maybe it's the friction and inefficiency that makes a thing secure. And making it easy and efficient makes it less so. And making it easy and efficient to improve its security is insane.