Apollonian wrote:...
given the mission goals of Celestia as primarily an educational tool.
...that's what our sub-community of educators likes to emphasize, but being a longtime Celestia developer and co-author I certainly disagree.
I rather see Celestia as a general, most versatile and precise
framework for simulating the Universe, rather than a specialized application.
For example, I have been striving a lot in recent years that Celestia's data base is becoming a 1-1 port of published professional catalogs of celestial objects. Since all catalog extractions are done by PERL scripts, any modifications of the original data (including elimination of typos) are documented in human readable form in the PERL scripts that are part of the official source code.
Being a senior scientist in theoretical particle physics, where most activities these days focus on astro-particle physics and cosmology, I also know what I am talking about in this domain...
++++++++++++++++
Altogether, my dream is to make Celestia a unique sythesis of precision and stunning 3d graphics, with a reliable asto-physical data base of scientific standards. A most important aspect being Celestia's modular design and the many possibilities of expansion at the user level.
++++++++++++++++
But I also have my doubts about using SPICE ...For a number of different reasons.
Incidentally, Celestia's accuracy has been tested quite intensively (by me and others) with very favorable results. Notably the very high precision of delicate mutual (eclipsing) events among e.g. Jovian moons, moons of Saturn all the way to precise occultation timings of the Pluto Charon system are most remarkable. The VSOP87 expansions are used as much as possible for calculating planetary orbits. There is certainly room for improvement as to the Moon's ephemeris predictions. But the reasons for this exceptional case are well known and would require something like VSOP2000...
Bye Fridger