seanpg71

I built a small mun return craft that was just a capsule with some parachutes and a heat shield, a monoprop tank and some maneuvering thrusters. And was a bit annoyed when I realized that meant I couldn't use a maneuver node for the burn to break orbit and return to Kerbin.

So... we have parts locked in a tech tree. This gates parts of the game and builds a sense of progression. I like that. However - the current tech tree leaves a bit to be desired. With an ideal tech tree, everything you unlock would grant some new capability that you didn't have before. Each would be something you'd be excited for and working towards that would let you do something you previously couldn't or something that was better than the previous parts you were already using. This tree isn't that. There are a bunch of techs that are just sorta strewn about that you unlock just because they're on the way to whatever else you might want or that are just a different version of something you already have. There's a bunch of things that are just longer tanks or different structural bits or trusses or what have you and those just don't seem like exciting things to unlock. This isn't necessarily easy to fix though. Part of the problem is that you need a good portion of the parts to be available pretty early. Tanks and engines need to be available basically at the start. There is a progression to larger capsules, tanks, and engines - and that works but doesn't necessarily make sense for early progress. So instead you have techs that unlock longer tank sections or structural trusses and stuff which just aren't especially exciting - the game should just give you those. You could start with *bad* parts - engines that are especially inefficient and tanks that are heavy - and then unlock better stuff. KSP1 did a little bit of that by starting with the solids and the engines that didn't gimble. But there's already enough parts - we probably don't want engines that are just bad. I could totally see something like engines being limited to 50% thrust until you unlock a tech though. KSP seems to think kerbals are an important part of the experience - so they want to start you with a capsule and parachutes. I understand that - but that's also a mistake for a good tech progression. It'd be much better if we had to start with a probe that had limited comms and battery that could barely survive a quick hop. And then we'd have to research to unlock batteries to allow orbital flight and solar panels to allow more than a few orbits, and then either comms and a comm network to get out orbit or would need to research capsules and parachutes and heat shields. Here too we're stuck with the progression just being to bigger parts, but not necessarily better. Unlocking better solar panels and batteries and more capable probes and whatnot would be nice - but instead we're often just sorta unlocking different sized copies of the stuff we already have. -- Now - related to collecting science in general. There's a bit of a physics problem in that you need to have a good portion of the tree unlocked before you can actually get kerbals to other planets and especially before you could get them back. So you end up with a ton of different experiments just collecting science around kerbin and the mun. And then once you get to actual far away planets where the exciting science should be happening, you've already unlocked all the stuff you care about. So, I personally would separate science and engineering. Have the tech tree be an engineering tree. You unlock parts automatically over time by using the current ones. You want bigger tanks? Fly some rockets with the current ones and apply various stresses. Maybe have a way to instrument the vehicle in some fashion if you want to speed up a particular sort of research. Use your current engines a number of times or fire them engine in a vacuum. Cycle your batteries. Build some smaller space stations and do some docking to get the bigger docking parts. There could be missions to hasten things or ways to focus on a particular tech. Science then, could be something else. You wouldn't have to worry about it at kerbin at all. It would be weird apparatus that you attach and take to other planets. Maybe this helps you locate resources on those plants. Or maybe it unlocks game info. Maybe the game won't do delta-v cals for planetary transfers until you've put a big telescope in orbit, or the surface of a planet is just a flat unknown that you're dropping your lander into blind if you haven't done some sort of orbital terrain scan or maybe maneuver nodes just give you a question mark about what sort of orbit you'll end up in until you've actually sent something there once and done some sort of gravity scan. Or maybe there are just science missions that want a particular sort of info. But if you split science from the tech tree, you could make it much more interesting and less front loaded and repetitive.

When the Apollo astronauts were on the way to the moon, they had to look out the window occasionally and line up a hand held device with the earth's horizon while pointing at various stars as part of the measurements to figure out where exactly they were. The accuracy of these numbers was entirely up to the skill of Michael Collins and as the mission progressed he figured out what worked and didn't in 0g and with the actual view out the window as compared to what had worked in the simulators with different optics and lighting conditions. When they landed on the moon, NASA wasn't entirely sure where they were landed. They knew the general area to within a crater or two but that's about it. On every orbital pass they had Collins searching a different (incorrect) area of the moon below based on their guesses from the length and direction of the various burns and thruster firings, and from the descriptions of the local features from Neil and Buzz. He never managed to find the lander. There is a reason that they put optional course correction burns in the flight plans for real missions. We don't always know exactly where things are, and they don't always perform exactly the same. So sure, if you point retrograde and burn the physics will always be the same, but a less experienced crew might not actually be pointing quite retrograde the whole time - or might not correctly plan for the shutdown transient and then would have to correct for that afterwards. We could model inexperienced Kerbals by having our exact orbital parameters be fuzzy, or by having our keyboard inputs get a bit mangled at times. Unfortunately - having our rockets fly crooked and not knowing exactly where they and where they're going would be annoying to play. So I don't see a big problem in having our craft always fly where we tell them to because it's fun, while imagining that the less experienced kerbals waste a bit more fuel in unshown corrections to make that happen.