I don't mind using the existing Maneuver Node graphical widgets to rough in most burns, and this to fine tune it:īut I hate running the risk of screwing it all up be my fumble-fingers or missing the start/stop cues. We have this but I have not yet tried out the updated version: Mind you, KontrolSystem2 is pretty amazing! I'd just rather have a mod like K2-D2 for precision execution of nodes. I'll take whatever I can get, but I like the idea of a C# mod solution that runs in the game a bit better than a kOS script sort of thing that may or may not run as quickly. I'm poking around with KontrolSystem2, and it's exec_node function looks like it will do something like this - but the results leave something to be desired, possibly because it's running to2 code (that compiles to IL? and supposedly would be almost as fast as C#?). MechJeb 3.0?) any time soon, but darn it! We need a reliable way to execute maneuver nodes with precision! You can specify the precision required of the total dV added when you you ask MJ to control the burn through the maneuver node, and the code reduces the throttle at the end to try to hit that requested precision with good accuracy.īasically MJ is my gold standard. This is one (of many) things MechJeb does really well in KPS1. Adding a more sophisticated approach to throttle control in 4 may make solving 5 easier. So, at a minimum I think K2-D2 needs better warping control in step 2, and needs to use a better approach to determining when to end the burn in step 5. There needs to be more smarts here so that the end of the burn really does give the player what they've planned with the maneuver node! This could even mean updating the maneuver node pointing direction dynamically to keep the orbit evolution headed toward the node prediction, but at the very least is means doing something smarter than just switching off the motor when the burn time has elapsed. As the two get close toward the end of the burn, the mod may need to throttle down a bit (depending on the TWR and how fast the two are converging), and the mod may also need to control the end of the burn so that the actual orbit has just met the sim prediction. I'm hoping K2-D2 can help out here!įor a mod to be a really good maneuver node executor I think it needs to keep an eye on converging the sim predicted maneuver node result and the actual orbit of the craft. This is actually not ideal, and this is where we need a better approach. I suspect that most of the mods out there with maneuver node execution capabilities just set the throttle to 1 and attempt to end the burn on time. I think the biggest need for a precise node executor is in problems 4 and 5. I'm not sure how far past the intended orbit it would have taken me, but it certainly overshot the intention for the node. In fact, K2-D2 was burning right past the sim predicted orbit for the node and I had to hit the stop button on K2-D2. I used Maneuver Node Controller to set up a precise burn that would get me a Munar periapsis of ~40km, and when I had K2-D2 execute that node it did appear to start right on time, but it ended late and I was heading for a Munar impact mission instead of my planned Munar encounter with a prograde flyby at 40km Pe. Prior to entering the Mun's SOI I used K2-D2 to help me execute the transfer burn from LKO to the Mun. At that distance most likely all time warp speeds would have been available anyway. I can try again without Time Warp Unlock, but as that execution began shortly after entering the Mun's SOI I'm not sure if Time Warp Unlock would have had much effect. I happen to have the Time Warp Unlock mod running, so it's possible an interaction with that mod and yours resulted in the misbehavior. I created an issue for you relating to 2 as using K2-D2 to circularize my Munar orbit at periapsis resulted in shooting right past the node and ending timewarp far too late. That part works really reliably, and other mods are available to set up a node precisely (e.g., Maneuver Node Controller, etc.) You've got the pointing part of 1 nailed I think. I think this breaks down to five problems fundamentally, and you're well on your way to solving them!Ĭalculating the required burn and pointing the ship to it Thinking further about this, I think it's clear your mod serves a vital purpose and will be a huge success - Being able to reliably time and execute maneuver nodes is essential, and while using our own human reactions to time start and stop via the stock capability will do OK in many situations, there is definitely a need for being able to execute nodes more precisely. While he was very easy to use, the results were not ideal and so I've created some issues for you on your github repo. Took K2-D2 flying with me for a quick trip to the Mun.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |