Comment by helgee
2 days ago
Oh, hi! Project creator here.
I am very happy that you folks are showing interest but I am also terrified because this thing is definitely not ready for primetime, yet. The discerning astrodynamicist and/or rustacean will surely find things which are less than optimal or plain wrong.
I am also preparing for a trip so I won't be able to monitor this thread closely but I'll try to answer question wherever I can.
Can I build a solar system simulator with this? Looking for calculating the absolute positions of planets and moons with the sun at 0,0,0 given a datetime. Kepler calculations for planets, moons etc would be fine, as long as I can insert arbitrary objects and calculate their trajectory using gravity. Can this project do that?
We already have the possibility to get positions of celestial bodies with very high precision via JPL ephemerides but you would have to write and integrate your force model yourself.
-
Hey, this is really impressive stuff, talk about a dream job! Two quick questions if you find the time+interest:
1. What does "safe" imply? Type safety, maybe? Or just a general synonym for "few bugs/well tested"?
2. What kind of industry is this primarily envisioned for use in...? I have experience in the satellite industry (GRE and TT&C) but I was surprised to not see the usual buzzwords on this page, namely "TLEs". I get that it clearly does way more than terrestrial orbital work, but AFAIK that's, say, 99.9% of the current space industry.
In other words: is this library for ESA missions??? That would be rad as hell, if so.
also P.S. I love AI art compared to most, but I would def recommend carving out that cool crab and putting on a background of real stars. As you may or not be aware, many scientists have been a tad radicalized about AI, and AI art in particular -- would be a shame if that slowed adoption of a cool project!
Ground Receiving Equipment, Telemetry, Tracking, & Control, and two line rlement sets. Apparently everything you need to predict orbits of a satellite are contained inside TLEs metadata and is considered a good way to share orbital information between disparate parties.
Sorry for explaining what the jargon in that comment was, i guess. i did typo "element" as "rlement", i guess i suck.
Heh, you actually highlight a funny frustration from my time there: TLES technically has a capital "S" too, but it's basically always discussed in the plural -- so I would very frequently forget to capitalize the "S" to the veterans' chagrin. Clearly, I haven't shaken the habit!
In hindsight, I guess it's the industry's fault for not pronouncing it "Tee Ell Eee Esses"...
EDIT: And I appreciate the explanation, personally! That's the best part of HN. To be clear to passers-by, a TLES is a string of characters that describes the orbit of a single satellite (artificial or otherwise) based on its position & momentum at some given start second. They have to be updated pretty frequently to account for gradual stochastic drift, which is both an interesting technical problem and a compelling philosophical metaphor.
1. Safe as in type-safe, memory-safe, and null-safe in contrast to the state of the art which is mostly C++, Java, and tons of ancient Fortran.
2. At the moment, we are using it for speeding up Python code for telecommunications constellation design.
I want to commission a real human artist to design a logo and header image in the future. The AI art is a nice placeholder.
P.S.: I wish this was my day job. I actually had to quit my job at a major aerospace contractor to be able to work on this project.