I spent the whole day breaking through a problem with sending events and information from the client to the server. Everything flowed smoothly from server to client, but in the other direction... Still, I managed it. The mechanics immediately get tangled up with players, their controllers, and other complex things, though. So the next step is to create basic login and accounts to simulate players properly. Or rather, not merely simulate them, but allow proper network access from several clients...
Now that's something. The little ship is flying to Mercury, as its information shows: hull number 156, speed in km/s and acceleration in m/s. The green triangle marks the ship itself.
I've finally managed to work through almost all the interface complications. Planets in the star system now have labels, and the little ships have various useful interface elements, including drop-down menus.
I continue to debug the autopilot algorithms. I introduced a mirrored thrust vector for braking and am testing it. In the image you can see the planned trajectory, Earth and part of its orbit, Venus's orbit, and Mercury's tilted orbit. Ignore the gray sphere above the Sun )))
What are we looking at here? The Sun, Mercury's orbit and Mercury itself, Venus's orbit, and, behind it on the left, a bit of Earth's orbit. The little ship near the Sun is drawing its current planned course to Jupiter.
Today, while refining the ship's flight algorithm, I threw almost all of it away and rebuilt it from scratch. I made a lot of improvements as all sorts of things surfaced along the way: errors in calculations, bad timing for the start of various game modules, and smaller issues.
The navigation aim point is now set not on the planet but on where the planet will be. To do that, the ship calculates the planet's movement along its orbit, including all its nonlinearity.
I sketched out the basis of the ship's modules: life support, power, maneuvering engines, and so on. I think I can bring them to life today and create the architecture for how all this stuff works.
The planets' visualizations have been reduced by a factor of ten. At true scale they'd be only a few pixels across. So they're still much larger than they really are, and eventually they'll scale with the camera's position. The mathematics of flight, speed, and distance, though, is fairly realistic. To land, you have to fly mathematically to those very pixels, even though the visible surface is much larger.
The little ships didn't learn to fly today. I spent half the day working on our online store. I spent the other half chasing a bug introduced when I reworked some functions yesterday. Planet movement and orbit rendering suddenly broke. Clearly the functions were at fault, but where? Which one? Which variable was wrong?
The lockdown is certainly bad for our business, but the temporary closure of the shop has freed up time that I'm putting intensively into Exodus. I work on the project for at least eight hours every day.
A parade of planets. Modest. Plain. Without textures. BUT!!! The planets move at the game's simulation speed, even Pluto. Their motion follows real astronomical data precisely. Distances are to scale; the result gives us room out to 750 astronomical units. Orbits are drawn for the planets. Algorithms draw all of this automatically from six numbers in an Excel file. In other words, to draw a different system now, I only need to change one number in the loader: specify its number in the HYG catalog. If we have already added its data, a system with planets will appear. The next step is to add moons (most of the work is already done) and send little ships there with basic functionality for flights, landings, and so on. Expect an update soon!
