Science is pretty much stupid. Just get rid of it.

[regex]

57 minutes ago, Mickel said:

That's what 'starting' a game means.  You might 'know' all this stuff.  Your new Kerbals don't.

Or we could just assume that people did some actual science beforehand so I don't have to follow the exact same path to space every damn time.

Honestly that's what we really need, a tech tree that can accommodate all sorts of playstyles, not some dumb rehash of American Space Program everyone thinks we should have.

[Mickel]

30 minutes ago, Kerbart said:

It's not super easy to fix that though. If your science instrument readings determine what you can unlock, you’ll still get the “go to crater X to unlock Y” mechanism. It’s just a bitless obvious but after unlocking half a dozen nodes it’ll be really obvious. Unless those readings are no predetermined. But then the game can become an exercise in frustration as non of your experiments is giving you that one reading you need to unlock docking ports.

There are two threads here which get you around your problem.  Docking ports are a new gadget.  You can work out how to building them and test them on the ground.  You can make them with the same materials you already use to build the rocket.  You just need to do that in space to prove you can, ie Soyuz 4 or Gemini VIII.

Going somewhere and finding something new is what gives you big step changes in capability.  Perhaps a new material that allows a more efficient or lighter engine, or a different fuel compound that means you can fit more dv in a given volume (yep, sorry, variations on a theme.  others will have better ideas).  Maybe you could get there without leaving the Kerbal SOI, but it would take a heap longer.  This is an issue in KSP1 - the power of the science labs means you don't really need to.

This sounds a lot like 'science points'.  However, following PDCWolf's train of thought, you don't exactly see the tree, just the outputs from your general intended direction.

7 minutes ago, regex said:

Honestly that's what we really need, a tech tree that can accommodate all sorts of playstyles, not some dumb rehash of American Space Program everyone thinks we should have.

On this point, we can agree.

If nothing else, all this talk of science has had me doing some basic reading into what did happen, when and how in the '50s through '70s.  I know I've had to rethink some of my ideas on tech progression.

[hatterson]

13 minutes ago, Mickel said:

There are two threads here which get you around your problem.  Docking ports are a new gadget.  You can work out how to building them and test them on the ground.  You can make them with the same materials you already use to build the rocket.  You just need to do that in space to prove you can, ie Soyuz 4 or Gemini VIII.

Going somewhere and finding something new is what gives you big step changes in capability.  Perhaps a new material that allows a more efficient or lighter engine, or a different fuel compound that means you can fit more dv in a given volume (yep, sorry, variations on a theme.  others will have better ideas).  Maybe you could get there without leaving the Kerbal SOI, but it would take a heap longer.  This is an issue in KSP1 - the power of the science labs means you don't really need to.

This sounds a lot like 'science points'.  However, following PDCWolf's train of thought, you don't exactly see the tree, just the outputs from your general intended direction.

The counter point to "and finding something new" is that  the act of getting to other places (LEO, the Moon, Mars, etc.) has given us very, very, very few actual hard scientific or engineering advances related to spaceflight and spacecraft design. And frankly, there are few true *scientific* advances required at all for humans to go to the moon.

There are, however, a *ton* of engineering advances that were required and those problems, and all of the new tech that came out of them,  were predominantly solved on Earth and with only knowledge we had on Earth. It's not as if NASA went to orbit and suddenly said "oh hey, we learned something fundamentally new that lets us build an F-1 engine." Obviously there was a ton of testing that required real world data to validate, but the big space program advances were much more along the lines of "spend a ton of time engineering a new solution and then go test if it works" as opposed to "go to the moon and suddenly we have a lot more knowledge"

The counter point to that counter point is that KSP is a game where we launch little green men into space as they humorously scream. Abstracting the whole engineering and testing process into "go place and collect progress points" is a fine abstraction for the theme of the game. KSP stock never has been, and never will be, a realistic space program simulator, it's just not the DNA of the game.

Edited January 2 by hatterson

[regex]

11 minutes ago, Mickel said:

If nothing else, all this talk of science has had me doing some basic reading into what did happen, when and how in the '50s through '70s.  I know I've had to rethink some of my ideas on tech progression.

Ultimately that's my real gripe about science gatekeeping in this game, it's keeping me from defining my own space program. There's not a whole lot of choice in the tech tree, certain nodes have much more value than others (the ones with experiments in them) and stuff like aircraft and probes get short shrift or have to be unlocked after. It tends to force me into a certain style of play, even if I unlock every part node as I go.

So yeah, learning about what atmosphere does by sending a probe every new game just isn't interesting to me. Let me send a damn plane on a pillar of fire where I want, when I want. Or whatever.

1 minute ago, hatterson said:

Abstracting the whole engineering and testing process into "go place and collect progress points" is a fine abstraction for the theme of the game. KSP stock never has been, and never will be, a realistic space program simulator, it's just not the DNA of the game.

Yes. Lately I've been considering each node in the tech tree being me directing industry where to ramp up first rather than "unlocking tech".

[Mickel]

18 minutes ago, hatterson said:

The counter point to "and finding something new"....

The counter point to that counter point is....

And this is the thing...  'real' science (which isn't just 'science' because it needs engineering and art and countless other inputs) is hours and hours of slog in labs by thousands of unwashed post-grads (with the occasional disrupter thrown in), some of which works and some that doesn't (sometimes in a news worthy way).  None of that makes a game.

Sending Kerbals out into space on some quest that a player determines worthy does.

Edited January 2 by Mickel

[Scarecrow71]

8 minutes ago, regex said:

Ultimately that's my real gripe about science gatekeeping in this game, it's keeping me from defining my own space program. There's not a whole lot of choice in the tech tree, certain nodes have much more value than others (the ones with experiments in them) and stuff like aircraft and probes get short shrift or have to be unlocked after. It tends to force me into a certain style of play, even if I unlock every part node as I go.

So yeah, learning about what atmosphere does by sending a probe every new game just isn't interesting to me. Let me send a damn plane on a pillar of fire where I want, when I want. Or whatever.

Agreed.  While there may be times I'm in the mood to follow the American or Russian or Japanese space program, the game is called KERBAL Space Program.  Not "History of Real World Space Flight" or some such.  Let me define how my program works.

[Ahres]

4 hours ago, regex said:

I went ahead and read it again and I think the only real thing I got out of it was having maps, which have a dubious utility at best unless they're high definition and have other features added to support them. Everything else is like, why? Every time I start up a new game I have go through the same tedious rigamarole of learning how an atmosphere works?

Well, no, I don't. I look at that gameplay and I wonder how providing the player with useful information isn't just going to end up being gatekeeping for the sake of gatekeeping.

I can't say I'm understanding fully. If you're worried about the tediousness of a new science save, you can just add extra science points at the start- or play sandbox, but it seems like you want some form of progression. And if you want to hurl a spaceplane through an atmosphere in a pillar of fire there wouldn't necessarily be anything stopping you. But for the sake of mission planning and science gameplay depth I'd appreciate being able to gather that data and make it useful.

To reiterate, I'm not saying the game should withhold parts from you until you've performed a very specific action like sampling an atmosphere. To keep up with the example, I'm saying you shouldn't know the atmosphere's characteristics until you've sampled it. No in-atmo trajectory display*, no parachute planning*, no resource survey, etc. But if you have played the game before and want to just go for it, then by all means do so! 

*Assuming these features will be in the game someday.

Admittedly, I haven't thought it through. Perhaps there's something obvious that I'm missing. Currently, we get Science Points for doing science actions. So would the list of things I just laid out also be purchased with Science Points? I suppose not right? We get resource overlays in KSP1 by using the orbital scanner, not by doing "experiments"... so perhaps the stuff I'm wanting is instead provided to the player once they've used the part that would yield that information? And those parts are unlocked using the science system that is now implemented? I don't necessarily have any problems with that. I suppose then my question to you, @regex, would be what your thoughts are on the resource overlay system in KSP1. You don't know where ore is concentrated until you've performed an orbital survey. I think that's a great mechanic. You still have to build and fly a rocket (core pillar of KSP) and on top of that you're incentivizing players to put a spacecraft into an orbit that they likely would rarely attempt (woo let's learn/do something new!).

[regex]

18 minutes ago, Ahres said:

No in-atmo trajectory display*, no parachute planning*

*Assuming these features will be in the game someday.

I have my doubts. But to the point, I learn all that through trial and error right now anyway so what you're offering me is quality of life, and at the point where you take away QOL in order to gatekeep I'm going to call that dumb. It's like the pilot class in KSP1, probably the biggest example of just how stupid QOL gatekeeping is. Like that Kerbal had to gain experience and level up like we're playing D&D in order to hold a point in another direction? What the hell am I actually playing?

18 minutes ago, Ahres said:

what your thoughts are on the resource overlay system in KSP1. You don't know where ore is concentrated until you've performed an orbital survey. I think that's a great mechanic.

That's totally fine, especially if resource concentrations are randomized between playthroughs. Hell, make it require multiple passes and have different resolutions like ScanSat, make slope and height maps too. The problem with maps though, is that they don't provide a lot of good actionable information unless they have high definition (resolution, zoom in) data and are also paired with a waypoint system where I can look at the flight screen and see where I want to land. Without those tools the map is nothing more than a pretty picture.

Edited January 2 by regex

[Ahres]

41 minutes ago, regex said:

But to the point, I learn all that through trial and error right now anyway so what you're offering me is quality of life, and at the point where you take away QOL in order to gatekeep I'm going to call that dumb. It's like the pilot class in KSP1, probably the biggest example of just how stupid QOL gatekeeping is. Like that Kerbal had to gain experience and level up like we're playing D&D in order to hold a point in another direction? What the hell am I actually playing?

I see where you're coming from, especially with the Pilot role of KSP1. I wonder where you draw the line between QoL and purpose of gameplay, however. If we take engines for example, one could say bigger engines are just QoL. I don't need them, I can just use a larger number of smaller engines. Is gathering science for the sake of getting bigger engines considered gatekeeping? I don't think you'd say that's the case. So if I'm saying I don't want aero-capture data available to me until I've sampled the atmosphere, I'd say that's reasonable. 

 

41 minutes ago, regex said:

That's totally fine, especially if resource concentrations are randomized between playthroughs. Hell, make it require multiple passes and have different resolutions like ScanSat, make slope and height maps too. The problem with maps though, is that they don't provide a lot of good actionable information unless they have high definition (resolution, zoom in) data and are also paired with a waypoint system where I can look at the flight screen and see where I want to land. Without those tools the map is nothing more than a pretty picture.

Agree with this completely. Again we're going to what you consider gatekeeping. You can't have those slope and height maps until you've gone through the rigamarole of sending a satellite out there to get it. You can't have detailed resource data until you've sampled the soil. You can't know atmospheric characteristics until you've pointed a telescope at it/performed a fly-by/landed a vessel/whatever else may yield that data. It seems like we're on the same page here. Or am I wrong?

[regex]

25 minutes ago, Ahres said:

So if I'm saying I don't want aero-capture data available to me until I've sampled the atmosphere, I'd say that's reasonable.

I'm presuming you have no intention of gatekeeping my delta-V expenditures or manner of mission execution and that basic planetary stats don't change from save file to save file. With that said, what actionable data are you offering me then? A new QOL feature that shows how I can aerobrake? I'll just slap down a quicksave and try a couple times until I know how it works. Or I'll leverage my knowledge from previous saves and just do the thing. At that point you're just keeping me from a time-saving feature. Meanwhile, I've been doing actual science learning how to aerobrake.

Edited January 3 by regex

[Pthigrivi]

3 hours ago, regex said:

I'm presuming you have no intention of gatekeeping my delta-V expenditures or manner of mission execution and that basic planetary stats don't change from save file to save file. With that said, what actionable data are you offering me then? A new QOL feature that shows how I can aerobrake? I'll just slap down a quicksave and try a couple times until I know how it works. Or I'll leverage my knowledge from previous saves and just do the thing. At that point you're just keeping me from a time-saving feature. Meanwhile, I've been doing actual science learning how to aerobrake.

For dV and interplanetary transfer planning I would say no, those should absolutely not be gate-kept. Just like SAS that should be given because these are not skills that map well onto trial and error. You need specifically built, in-game tools to solve those problems. Aerocapture is a bit different though. This is a learning experience. I think for players there’s actually a lot to be learned from many attempts at reentry and from quick loading an aerocapture at a new planet a few or even several times to get a feel for it and understanding with some trial and error what heat flux means and what the practical limits are in different atmospheres.  Having done just that many times though the question becomes can we get a visual estimator for late game? Can practice at aerocapturing around Eve or Laythe unlock a handy visual guide so I don’t have to do it 20 times on every mission?
 

This to me feels akin to supply routes. The first run is a cool engineering and traversal puzzle. Subsequent runs are boring and repetitive. This kind of dynamic really speaks to unlockable QoL. 

Edited January 3 by Pthigrivi

[regex]

3 minutes ago, Pthigrivi said:

Can practice at aerocapturing around Eve or Laythe unlock a handy visual guide so I don’t have to do it 20 times?

Why does it have to unlock? I've already done the science through play, I know how to aerobrake, I did it in my last save without the help.

9 minutes ago, Pthigrivi said:

Subsequent runs are boring and repetitive.

Yes, yes they are. Don't lock QOL features, just give them to us.

[Pthigrivi]

46 minutes ago, regex said:

Why does it have to unlock? I've already done the science through play, I know how to aerobrake, I did it in my last save without the help.

Yes, yes they are. Don't lock QOL features, just give them to us.

YOU have. Most players have not. All those new players have a lot to gain from manually encountering reetry heat and its repercussions. Try to understand: you and me are old—this game isnt for us. I don’t make broad game recommendations based on my own personal playstyle or what I happen to know from playing KSP1 for thousands of hours. I don’t really personally care about planes at all, but I recognize that a lot of players love that and there should be a path in the tech tree for them to find a home and experience the game on their terms. Maybe they want to do planes and probes to get that historical feel. Maybe they just like piloting and its planes and crewed landers—be damned preliminary probe data. Lots of folks have pointed out the tech tree is a bit—slender.  These key strategic decisions and broad gameplay opportunities are not being well catered to.

The key distinction Im trying to make is not simply between new capabilities and QoL features. There’s a fuzzy line here. High ISP engines are QoL but also gate-kept through tech tree progression. What Im pointing out is that some gameplay experiences benefit from trial and error—learning by doing—reiterating with different launch configurations and mission profiles and landing techniques. Other gameplay elements like estimating dV budgets and finding ideal transfer windows don’t respond well to trial and error. Nobody is going to intuitively deduce that they need to timewarp 230 days before departing for Duna. Nor should they be exiting the game pull up a dV map jpeg in a browser. Those are tasks that require built-in tools. Anything like that should not be gated. 
 

For many, many other tasks like designing launchers or learning basic orbital transfers or landing or docking players really need to learn by doing. The game should no-more land for you than it should design boosters for you. That exploratory process IS the game. Many of these learned processes though gain from the option of automation once they become old hat. We can save pre-designed booster and launch assemblies and re-use them. We should be able to automate re-supply missions to colonies after a proof-run. The sweet spot is taking the time to learn and achieve these kinds of tasks and then earn the ability to automate them so they don’t become mindnumbingly repetitive. Thats the specific category of QoL features that should be unlocked by doing it manually first. 

Edited January 3 by Pthigrivi

[regex]

25 minutes ago, Pthigrivi said:

The sweet spot is taking the time to learn and achieve these kinds of tasks and then earn the ability to automate them so they don’t become mindnumbingly repetitive.

Aerocapture/braking really isn't that.

In any case, I'll agree to disagree, for one because I don't really think it's a feature that needs to be in the vanilla game and for two because I think it's a really dumb thing to lock behind something that isn't mind-numbingly repetitive.

[KSK]

3 hours ago, Pthigrivi said:

For dV and interplanetary transfer planning I would say no, those should absolutely not be gate-kept. Just like SAS that should be given because these are not skills that map well onto trial and error. You need specifically built, in-game tools to solve those problems. Aerocapture is a bit different though. This is a learning experience. I think for players there’s actually a lot to be learned from many attempts at reentry and from quick loading an aerocapture at a new planet a few or even several times to get a feel for it and understanding with some trial and error what heat flux means and what the practical limits are in different atmospheres.  Having done just that many times though the question becomes can we get a visual estimator for late game? Can practice at aerocapturing around Eve or Laythe unlock a handy visual guide so I don’t have to do it 20 times on every mission?

Each to their own, but I disagree with this. Trial and error is boring. If you want to gatekeep a quality of life feature, gatekeep it behind something I can plan and execute, like putting something in a high orbit around Eve or Laythe to measure the extent and density of its atmosphere and give me some data (or better, a visual aerobraking guide) to help me plan a landing mission.  Please don't make me have to do a bunch of save scumming before my first landing, after which I'm deemed to have 'learned' how to aerobrake and can get the nice visual guide. Because I promise you that I'm not learning about heat flux or whatever, I'm just grinding through a bit of boring gameplay to avoid having to do it again.

More generally, I'm not really onboard with the idea that KSP2 has to be a learning experience and the notion that the players can start with in-game tools for these skills (like calculating delta-V requirements) but just have to git gud at these skills (like aerobraking) before they're deemed worthy of unlocking an in-game tool.

I say this as someone who a) doesn't have KSP2 so doesn't really have a stake in this discussion, and b) who found that the best bit of KSP1 was the learning to play part and that once I'd done that, the actual game was pretty thin. But just because I liked it, that doesn't mean that everyone should. And even then, the 'learning to play' bit was mostly about how I stopped fiddling around and learned to love the navball to do things like docking or powered landings.

[Periple]

What would the aerobraking tools even be like? It’s just so dependent on vehicle design — how pointy it is, whether it can be rotated to present more surface area to the airflow, how temperature-resistant its parts are, whether it has airbrakes, and so on. I think the best you could do is have a rough bracket for it.

[Sea_Kerman]

the trajectories mod for ksp1 was able to give a rough estimate, and you could tell it what you planned to do

[Pthigrivi]

6 hours ago, KSK said:

Each to their own, but I disagree with this. Trial and error is boring. If you want to gatekeep a quality of life feature, gatekeep it behind something I can plan and execute, like putting something in a high orbit around Eve or Laythe to measure the extent and density of its atmosphere and give me some data (or better, a visual aerobraking guide) to help me plan a landing mission.  Please don't make me have to do a bunch of save scumming before my first landing, after which I'm deemed to have 'learned' how to aerobrake and can get the nice visual guide. Because I promise you that I'm not learning about heat flux or whatever, I'm just grinding through a bit of boring gameplay to avoid having to do it again.

More generally, I'm not really onboard with the idea that KSP2 has to be a learning experience and the notion that the players can start with in-game tools for these skills (like calculating delta-V requirements) but just have to git gud at these skills (like aerobraking) before they're deemed worthy of unlocking an in-game tool.

I say this as someone who a) doesn't have KSP2 so doesn't really have a stake in this discussion, and b) who found that the best bit of KSP1 was the learning to play part and that once I'd done that, the actual game was pretty thin. But just because I liked it, that doesn't mean that everyone should. And even then, the 'learning to play' bit was mostly about how I stopped fiddling around and learned to love the navball to do things like docking or powered landings.

Okay so a) for anyone who hasn’t modded their game this how aerobraking and reentry has always worked. Its still how it works. Save scumming and learning by doing is how everyone learned to get to orbit and land on the Mun. Thats the kind of basic task that is going to require every new player a bit of practice so they understand the nature of it. Reentering Kerbin on those first several missions is similar.  And then b) you don’t have to save scum and do that on every new planet. Alternatively you could dip into the high atmosphere and take an atmospheric reading and voila: you’d have unlocked aerobrake prediction on all subsequent passes even with that first initial vessel. This conversation is in the context of Regex saying he personally wants to plow into the atmosphere in a pilar of fire and us explaining that he could absolutely still do that if he wishes. This is gate keeping with an a very simple, easy to use gate. Its no different than using a scanner to unlock topographic data. It’s certainly easier than unlocking parts or proving supply runs. Its simply a way of giving experiments actual utility and the game more depth in progression. 
 

Think about scanning or fog of war in any game you have ever played. Being able to see absolutely everything without ever exploring is certainly a QoL feature thats being “gate kept”. But thats kind of the point? You have to go there before you know everything? Thats kind of my whole critique, that science in KSP isn’t doing something thats very basic in all games that are about exploration: having exploration reveal new and useful information about the world you’re in. 

Edited January 3 by Pthigrivi

[regex]

It's funny to me that the one real example of a feature based on "actionable information" you can gather besides maps is something you can actually do science about right now, in the game. Probably the biggest use for an aerobraking assistant tool would be at Kerbin where we already should know pretty much everything and where it really doesn't matter where we land to do a recovery. MOST players will use it once or twice for every body outside of Kerbin with an atmosphere and they'll presumably have to unlock it for every body outside of Kerbin.

[Scarecrow71]

2 minutes ago, regex said:

It's funny to me that the one real example of a feature based on "actionable information" you can gather besides maps is something you can actually do science about right now, in the game. Probably the biggest use for an aerobraking assistant tool would be at Kerbin where we already should know pretty much everything and where it really doesn't matter where we land to do a recovery. MOST players will use it once or twice for every body outside of Kerbin with an atmosphere and they'll presumably have to unlock it for every body outside of Kerbin.

In the current iteration of KSP2, I agree.  There are only 4 bodies that have an atmosphere - Eve, Kerbin, Duna, and Laythe - so having an aerobraking tool would become useless rather fast.  Once you try to use the maneuver on a given body a second or third time you will probably never use it again for that body.  I could be wrong and someone else's mileage may vary, but I can't see it being very useful after the first time it is used.

Now, once we get to interstellar...could be a different story.  We have no clue how many other celestial bodies will have an atmosphere, or what kind of atmosphere they will have.  An aerobraking tool may be useful there...but again, the amount of use may be limited.

[regex]

2 minutes ago, Scarecrow71 said:

Once you try to use the maneuver on a given body a second or third time you will probably never use it again for that body.  I could be wrong and someone else's mileage may vary, but I can't see it being very useful after the first time it is used.

And that's going to hold true for every other solar system you go to, for MOST players (if they even go to every new solar system).

[Bej Kerman]

1 minute ago, Scarecrow71 said:

There are only 4 bodies that have an atmosphere - Eve, Kerbin, Duna, and Laythe - so having an aerobraking tool would become useless rather fast

It definitely wouldn't. It'd be about as useless as the dV tool, that is, it'd be extremely useful, especially if your vehicle designs (and consequently their aero profiles) vary a lot. And even if physics was broken and aerobraking worked the same way for every possible combination of parts, it's not a bad thing to have a quick reference.

There's also the matter that "only 4 bodies that have an atmosphere" is a statement which will not survive the entire planned development.

[Scarecrow71]

1 minute ago, Bej Kerman said:

There's also the matter that "only 4 bodies that have an atmosphere" is a statement which will not survive the entire planned development.

Which I actually mentioned in my post, had you bothered to read the whole thing.  I literally pointed out that we have no idea how many bodies will be in any new system, nor how many will have an atmosphere.  Please use the context of my entire post if you are going to try countering it.

[Pthigrivi]

18 minutes ago, Bej Kerman said:

It definitely wouldn't. It'd be about as useless as the dV tool, that is, it'd be extremely useful, especially if your vehicle designs (and consequently their aero profiles) vary a lot. And even if physics was broken and aerobraking worked the same way for every possible combination of parts, it's not a bad thing to have a quick reference.

Yeah I personally found once I discovered Mechjeb's aerobrake prediction tool back in the day I started aerobraking nearly every time I was encountering a body with an atmosphere, so basically any mission besides Moho or Dres. It's free real estate, as they say. I'll also point out that players are going to be taking atmosphere readings any time they go to a planet with an atmosphere anyway. There's no new or repetitive work being done here on the part of the player, just adding a dynamic that makes science relevant to exploration outside the tech tree and creates some incentive for diverse play styles.

30 minutes ago, regex said:

It's funny to me that the one real example of a feature based on "actionable information" you can gather besides maps is something you can actually do science about right now, in the game. 

Right so unlocking aerobrake prediction, topographic and biome overlays, resource prospecting, and unlocking simulators for testing rovers and ascent and descent vehicles. I'd say thats substantial, and there are probably others I haven't thought of. 

Edited January 3 by Pthigrivi

[hatterson]

31 minutes ago, Scarecrow71 said:

In the current iteration of KSP2, I agree.  There are only 4 bodies that have an atmosphere - Eve, Kerbin, Duna, and Laythe - so having an aerobraking tool would become useless rather fast.  Once you try to use the maneuver on a given body a second or third time you will probably never use it again for that body.  I could be wrong and someone else's mileage may vary, but I can't see it being very useful after the first time it is used.

Now, once we get to interstellar...could be a different story.  We have no clue how many other celestial bodies will have an atmosphere, or what kind of atmosphere they will have.  An aerobraking tool may be useful there...but again, the amount of use may be limited.

Also once we get colonies as theoretically if you wanted to put a surface colony on Duna, Eve, or Laythe, you'd want to land your stuff pretty close to attach it. I'd assume you'd need some sort of rover ability to move it around the surface a few hundred meters, but being able to land within 500m is a lot different than landing 20km away.