There are at least two open source planetarium programs that draw to Open GL. It would be a huge project to merge them, but it would be possible. The upside would be an incredible viewer experience. The downside? A slower, larger, viewer.
What do people think?
To begin with, I'm assuming you would extract only the parts of the planetarium program(s) that calculate position and magnitude (and maybe color?) and place an appropriate dot on the sky. All of the other things that those programs can do -- labels, constellation lines and borders, grid overlays, information pop-ups when hovering over an object, etc. -- aren't needed if the project's goal is only to make SL's sky look like the real one.
ReplyDeleteAnd -- to cut down on the number crunching -- you might limit the data base to objects of "naked eye" magnitude (for example, you could leave Neptune, Uranus and Pluto off the list). On the other hand, since Second Life provides thousands of locations with perfect "dark sky" conditions, perhaps you could do what Stellarium does with nebulae, galaxies, and the cloudiness of the Milky Way: display image files at the proper coordinates.
Other things I can think of to (possibly) limit its size and the effect it has on the overall viewer:
-- Make it optional, like Emerald's new and exhaustive list of Windlight presets.
-- Connect it to local (server) daylight settings and to the viewer user's Environment override settings, so it's not chewing up processor time drawing stars when they can't be seen.
Finally, if it works: offer it as a module to other "3rd-party" viewer developers and to individual users, with instructions for the latter group for incorporating it into their existing viewers. I, for one, would gladly learn how to recompile viewer code in order to use it!