I started with a blank simulation, into which I placed a black hole of 10 million solar masses (the best kind). Around this black hole, I inserted one of every included star, starting at the top of the list and working down. I inserted an extra Proxima Centauri by accident. Auto orbit was enabled by default, and the timestep was around 2 min/sec. Smaller stars were generally closer to the black hole, while larger ones were farther away, but there was no real structure to the simulation. After I had started the simulation, I decided to get rid of the duplicate star. Naturally, I decided it should be pulverized in the maw of the ultimate destroyer it was orbiting. I zeroed the velocity, and that is when things went downhill.
Let's examine the picture with the UI visible. Note Rigel's velocity; in the actual simulation, it isn't moving at all. Its trailis, though. In fact, all of the trails are slowly continuing in an outward spiral. All of the bodies are stationary. The stellar animations are working fine and when the ghost of Rigel collided with another ghost seconds after the screenshot was taken, the Rigel preview window and the visible Rigel both showed the collision normally. Also, notice the number of bodies. That should be 23 (or 24).
I have recreated the bug countless times. As far as I can tell, the necessary factors are the timestep and the star in the collision. It only worked when the star was much smaller than the black hole. Deneb actually consumed it in a glorious display of power and then continued as normal.
The log is attached. I recreated the bug twice in this log, and I think that is all that I did.
Topic: Black hole vs. assorted stars bug (Read 2370 times)
