Pituitary can take a code diff and check whether it conflicts with an accepted specification before that change is merged.
That detail changed how I think about researching a protocol like Dusk, because reading what a system is supposed to do only gets you halfway there.
Dusk built Pituitary after dealing with spec drift internally, where decisions, terminology and implementation can gradually stop agreeing as a codebase evolves. The tool indexes that recorded intent, can flag implementation changes that contradict it, and can trace which related areas are affected when a decision changes. It is deterministic by default too. For me, that creates a useful pressure test for protocol research: if I am forming a conclusion from an architectural claim, I want some way to notice when the code has moved while the claim has not.
A specification can describe the intended system.
The implementation decides whether that description is still true.
That gap is worth checking.
@Dusk $DUSK #dusk
That detail changed how I think about researching a protocol like Dusk, because reading what a system is supposed to do only gets you halfway there.
Dusk built Pituitary after dealing with spec drift internally, where decisions, terminology and implementation can gradually stop agreeing as a codebase evolves. The tool indexes that recorded intent, can flag implementation changes that contradict it, and can trace which related areas are affected when a decision changes. It is deterministic by default too. For me, that creates a useful pressure test for protocol research: if I am forming a conclusion from an architectural claim, I want some way to notice when the code has moved while the claim has not.
A specification can describe the intended system.
The implementation decides whether that description is still true.
That gap is worth checking.
@Dusk $DUSK #dusk