@xj9 @rek @neauoire @xuv I think if it's good enough for civilian use and doesn't require satellite support then it would *definitely* be of great interest to the military. Even 100m accuracy - which would be disappointing for the average person accustomed to present-day cell phone GPS - with just an Arduino level of electronics would be a big deal for missile guidance.
@mattskala @xuv @neauoire @rek The SR-71 used to have navigation based on well-known stars, but the aweigh home page only says that it measures the sun's position. So that would be "celestial body" not "celestial bodies".
It also needs a real-time clock. I'm not sure if the SR-71's star tracker had that problem.
@CharredStencil @xuv @neauoire @rek A lot depends on the precision. If you need real time to one second, okay, no big deal although setting your clock accurately without using a reference that ultimately comes from GPS will be interesting. If you need microseconds, not so easy. Similarly, measuring the position of a star (including the Sun) to one degree is a lot easier than to one second of arc.
@CharredStencil @xuv @neauoire @rek Traditional celestial navigation (with a sextant and so on) can do star positions to one second of arc, which is what you need to even approach GPS, but only with expensive precision optics and a trained user. For doing it automatically, optical encoders good to that precision are *to this day* on the export-restricted-because-of-military-applications list, at least from Canada.
@xuv @neauoire @rek Personally, I'm keeping more of an eye on this stuff: https://en.m.wikipedia.org/wiki/X-ray_pulsar-based_navigation
@xuv I'm not sure to understand how does this work in practice.
Revel in the marvels of the universe.