Decimals of seconds in F3/debug mission times

[Gaarst]

I just had this idea of adding tenths and hundredth of seconds in the F3 menu (or in the debug as well) for better accuracy concerning any events that might happen during the flight (understand explosion).

While the current system shows events chronologically, it becomes really hard to see what caused the explosion if you have a 100+ parts rocket blowing up in an instant.

Also, I don't know if KSP already calculates times to that accuracy. If not, then this might decrease performance, and then I would understand why this is not in the game.

I don't remember seeing this suggestion recently on the forums, neither do I know any mod adding this; so sorry if this has already been suggested/made.

[SpaceK]

Using

, I was able to determine the physical timestep of KSP to be in the range of about 1/50 second. Maybe a bit more, but not less. Therefore, hundredths of seconds won't work because they don't exist for the game. Tenths of seconds should be possible, but judging from my everyday experience, that would be kinda odd... usually, whole seconds or hundredths/thousandths are used, not tenths.

Still, it would be a useful debug function to solve for the physical timesteps (i.e. in increments of .02 seconds if my assumption is correct)

Edited November 10, 2015 by SpaceK

[Gaarst]

Yes you're right, now that I think about it, there is a setting in the options menu which determines the time between each physics calculations (can't remember the exact name). IIRC, it is set at 0.04s by default.

Maybe you can't get time intervals smaller than this setting ?

Anyway, tenths of seconds would be better than whole seconds IMO.

[SpaceK]

The slider sets the maximum time between two frames. I'm not quite sure, but if I had written KSP (pray to whatever supernatural entity you deem appropriate, or just hope if you're not so inclined, that I'll never even come close to writing KSP code, but the argumentation should be valid nevertheless), I would have used the maximum timestep for boring coasting phases between planets or whatever, and adjust to smaller timesteps for more interesting events like landings, dockings and sacrificing rituals.

Since the rapid unplanned disassembly probably qualifies as interesting, the timesteps should be shorter, but, according to my experiences with my gargantuan and prohibitively useless first attempt at an interplanetary spacecraft, it can only go so far.

[pincushionman]

The max delta can be defined down to 0.03, but it would be interesting to know how small the delta can be under ideal conditions. It probably is somewhere in the high thousandths or low hundredths, so it would be perfectly reasonable to show event times in thousandths. I doubt you'd ever get samples every millisecond, but that doesn't matter. You'll get what you get, rounded to the nearest thousandth. In real science and engineering, and you get samples at whatever your sampling rate is, reported rounded to whatever the appropriate fraction of a second is.