A good way to discover info about planets

[John FX]

Not absolutely sure which subforum to put this in but as I would like it to lead to a discussion I'm putting it into discussions rather than suggestions for development etc.

I just read a good article on the first probe to land on Mars and the thing that caught my eye was that the Russians had to send 3 rockets and the first was just a radar probe to enable the following rocket to more accurately target the planet and my first thought was I would really like to do that in Kerbal Space Program.

Here is the article

http://www.ozy.com/flashback/when-the-soviets-first-landed-on-mars/61124

What are all your thoughts on whether there should be some form of restriction of orbital data until probes are sent to nail down the exact location, mass, orbital period, radius etc of planets?

[Layereddiagram]

I think it would get annoying. It's in the same vein as some of the things that wouldn't go well in the stock game but are great in realism mods. Like the need for ullage burns and limited numbers of ignitions.

( you do mean that it would be hard to pin down in the tracking station and map views, right? )

[Luckless]

Personally I would love to have the option to have 'fuzzy data' and procedural solar systems as part of the stock game, along with more planning tools. I think it would be neat to have relatively vague data about some of the stuff farther out, and had to manage a science program to nail down more precise details, and also kind of neat to go 'out there' and see what the game generated this time around.

[xtoro]

Would be neat. But then what's stopping you from sending a probe with all instruments on it? In any case, you can look up complete stats on all planets anyways by googling

Personally I would love to have the option to have 'fuzzy data' and procedural solar systems as part of the stock game, along with more planning tools. I think it would be neat to have relatively vague data about some of the stuff farther out, and had to manage a science program to nail down more precise details, and also kind of neat to go 'out there' and see what the game generated this time around.

Procedural systems and planets would ruin the community's ability to learn from each other and have a common base to play on. Someone new wouldn't be able to ask questions about a planet's aerobraking or gravity assists because everyone would have different randomly generated planets. I think HarvesteR already mentioned this in an interview somewhere.

[Hagen von Tronje]

Probably worth pointing out that unless some randomization is thrown in at the same time, this is only going to affect new players, and only those who decline to look up the information online.

A "Kerbal Astronomy Program" sounds like a lot of fun, especially if you start with low resolution from land telescopes, then improve information from missions and orbital telescopes, but it would only be interesting after the first time if it involved some randomization, especially of orbits but maybe even of the celestial bodies themselves, and that sounds like a helluva lot of work to implement.

[Jett_Quasar]

I like the idea of having a probe that can detect a surface height map to locate flat landing areas.

JR

[Tex_NL]

...

Procedural systems and planets would ruin the community's ability to learn from each other and have a common base to play on. ...

Probably worth pointing out that unless some randomization is thrown in at the same time, ...

... but it would only be interesting after the first time if it involved some randomization, ...

First of all PROCEDURAL ≠ RANDOM !!! All it means is there is not pre-generated model for whatever you're trying to do. Instead the model is generated each time based on a mathematical procedure. Planet surfaces in KSP are already procedurally generated. Procedural generation CAN be based on a random seed but KSP uses a constant seed.

The idea of 'fuzzy' data without reconnaissance is great. (Hardly original but still great.) It even fits in very nicely with career mode. You'll have to actually visit a place before you receive contracts to do more detailed exploration.

Edited July 29, 2015 by Tex_NL

[KerikBalm]

Orbits should be calculated fairly well, and any body with a moon should allow you to determine its surface gravity...

I'd simply like that things like atmospheric density and scale height aren't available.

There were considerable unknowns about Mar's atmospher before a probe got there.

And of course, its surface geography was unknown.

Right now, you can get a pretty good look at a planet and where you might want to land by zooming in on it, I wouldn't allow that zoome until you've "revealed" it with at least a fly-by

In my career game, I send a probe to get basic pressure and temperature readings before I send a manned mission. No departure of the manned craft until the probe has sent data.

[xtoro]

I like the idea of having a probe that can detect a surface height map to locate flat landing areas.

JR

I know its not stock, but scansat will map out planets and show slopes. Using that data you can find flat landing spots.

[Hagen von Tronje]

First of all PROCEDURAL ≠ RANDOM !!! All it means is there is not pre-generated model for whatever you're trying to do. Instead the model is generated each time based on a mathematical procedure. Planet surfaces in KSP are already procedurally generated. Procedural generation CAN be based on a random seed but KSP uses a constant seed.

Are you aware that you quoted three different people, and two of those three used only one term or the other?

Speaking only for myself, I specifically avoided "procedurally generated" because I imagine it would be much simpler to randomize which actual date per orbital positions "Year 1 Day 1" represents. That would prevent you from consulting a launch dates table and circumventing the entire discovery system being discussed, as you'd have to at least do enough to figure out what time you're actually running on.

But I actually think most of this is better suited to a mod than anything. Most new players aren't going to volunteer to go in blind and revoke their ability to refer to a "cheat sheet" and most vets will probably want something more hardcore than what I imagine stock would go for (like actually randomizing planets).

[GarrisonChisholm]

Probably worth pointing out that unless some randomization is thrown in at the same time, this is only going to affect new players, and only those who decline to look up the information online.

A "Kerbal Astronomy Program" sounds like a lot of fun, especially if you start with low resolution from land telescopes, then improve information from missions and orbital telescopes, but it would only be interesting after the first time if it involved some randomization, especially of orbits but maybe even of the celestial bodies themselves, and that sounds like a helluva lot of work to implement.

Agreed^ - & I do "hamstring" myself about planets, by not looking them up in the Wiki until I have sent a first 'visitor', but as you point out, this is only a useful procedure on my first play-through.

[Hagen von Tronje]

Agreed^ - & I do "hamstring" myself about planets, by not looking them up in the Wiki until I have sent a first 'visitor', but as you point out, this is only a useful procedure on my first play-through.

You're not the only one! I definitely went this way the first time going anywhere, just so I could experience that wonder of landing on a planet I've never seen before.

[Tex_NL]

Are you aware that you quoted three different people, and two of those three used only one term or the other?

Speaking only for myself, I specifically avoided "procedurally generated" because I imagine it would be much simpler to randomize which actual date per orbital positions "Year 1 Day 1" represents. That would prevent you from consulting a launch dates table and circumventing the entire discovery system being discussed, as you'd have to at least do enough to figure out what time you're actually running on.

But I actually think most of this is better suited to a mod than anything. Most new players aren't going to volunteer to go in blind and revoke their ability to refer to a "cheat sheet" and most vets will probably want something more hardcore than what I imagine stock would go for (like actually randomizing planets).

You're right. I misquoted one person. That quote is now removed.

You're also right you did not use the words 'procedurally generated' but you did talk about random worlds. Random worlds are ALWAYS procedurally generated. There is no other way to create them.

[xtoro]

You're not the only one! I definitely went this way the first time going anywhere, just so I could experience that wonder of landing on a planet I've never seen before.

I did a combination of lookups and exploration with probes and found the latter much more enjoyable. Since I now know so much about all the planets, I went ahead and installed outer planets. Now I have a bunch more planets and moons to explore that are much harder to get to. I'm also doing it exploration style and not looking up anything about them. I dont even know if any have atmospheres, so it's up to probes to dip down and possibly get destroyed to find out. So much more fun and I wish I would have played like this from the start!

[Hagen von Tronje]

You're right. I misquoted one person. That quote is now removed.

You're also right you did not use the words 'procedurally generated' but you did talk about random worlds. Random worlds are ALWAYS procedurally generated. There is no other way to create them.

I guess what I'm asking is, why are you going on the assumption that I don't know this? Random worlds would be procedurally generated. The worlds already existing are procedurally generated. I didn't say anything about it one way or the other, because it plainly isn't the right term for what I was describing. You can see procedural generation already at play simply by flying far enough to see a patch of trees...

What I referred to was randomization. On a simple scale, of dates and/or orbital characteristics, on a larger scale, maybe of planetary traits, like ore distribution, atmospheric density, even sidereal period, etc.

Maybe I'm misreading this and you meant it as a friendly DYK, but it comes across like a correction to a mistake I didn't make. Indeed, it felt like you read the thread, saw inconsistent use of terminology, and forgot that one person didn't write the whole thing.

[Luckless]

Another point about procedural vs random that people often forget: Procedural isn't 'random', and because of that it can be reproduced.

So, the "Stock System" can still exist, and you and I can both play in that same system with the same planets and the same delta-v values and launch windows that all let the same crafts work to do the same missions. But once I get bored of that, which I mostly am, then I can go and recreate a new system, and enter seed 011101110110100001111001011011100110111101110100, and have some decent data on similar planets in slightly different orbits, but have new smaller moons and a few things way out on the edge of the system that I only vaguely know or haven't found yet. I can go out, and I can find some neat things that no one else has seen yet, because no one else has ever played that system's seed.

But,... I can pass that same seed number on, and then anyone else can type that in and go play on that same system as well.

And if it was an official dev supported feature, then I can be reasonably confident that things aren't going to break with the next update, and that they're not going to get bored of it and wander off and drop support because their lives got busy or they started playing some other game.

[juanml82]

This exists in the form of Scansat. I'm doing some biome hoping on Ike without it, and I'm landing on 20° slopes, 6° slopes, 10° slopes, etc, simply because I'm not using Scansat so I land blindly into what may, or may not, be flat terrain.

And considering that the stock game doesn't include a landing predictor, this is two steps ahead.

First, add the ability to predict the landing spot with the current trajectory, then add the ability to check the terrain features at the landing spot.

[Goddess Bhavani]

You're not the only one! I definitely went this way the first time going anywhere, just so I could experience that wonder of landing on a planet I've never seen before.

I agree that some orbital data should be masked in a difficulty option - both for realism and also, more instruments to carry to the planet = moar science options = moar fun = increased snack budget!

[cantab]

A concept I had along the same vein: For any body without a moon (including most moons themselves), you don't get the projected trajectory of your spacecraft until at least one operating craft has entered its SOI. (Perhaps you instead get a range of possible trajectories indicated somehow) That seems grounded in realism - if a body doesn't have a moon it's hard to know its mass until you send a probe past it, but rather easier to know its orbit from telescope observations. It wouldn't excessively hold the players back or make us jump through hoops or do gruntwork - if we want we can just go for it and sort things out when we get there at a probable delta-V penalty. And I think it would add some interest, and another reason to do a quick preliminary mission to a planet before sending Kerbals there.

[KerikBalm]

The moon itself would tug on its parent planet, and so we should still know its mass.

If the moon gets big enoug, its basically a binary system (pluto-ish).

Ike would certainly tug on duna enough to figure that stuff out.

Pol on Jool... not so much

Pol perturbing bop? perhaps?

Tylo/laythe on Jool? probably enough to get rough estimates

[Sharpy]

I think a mod - possibly an extension to RemoteTech? - that would provide navigation accuracy improvements would be nice. Say, instead of standard patched conics you're getting actual cones which denote where you might land after a maneuver. Deploy a GPS network, have radars orbiting the Sun at Lagrangian points relative to Kerbin, install cameras, radio-altimeters and such in the probes.

I really don't see this in stock, but it would make for an awesome mod.

[Hagen von Tronje]

I think a mod - possibly an extension to RemoteTech? - that would provide navigation accuracy improvements would be nice. Say, instead of standard patched conics you're getting actual cones which denote where you might land after a maneuver. Deploy a GPS network, have radars orbiting the Sun at Lagrangian points relative to Kerbin, install cameras, radio-altimeters and such in the probes.

I really don't see this in stock, but it would make for an awesome mod.

That does sound cool. I don't use MechJeb but I'd be interested in a mod that used GPS/comm relay networks to enable autopilot trajectories, say between two bodies that have GPS and comm networks in contact. That would feel sufficiently "earned it" enough to me and it would be fun setting up a network that lets you partly automate transfers. Throw in a few more tasks like a suborbital survey to allow landing autopilot. Autopilot is plenty realistic but I'd rather it be something you have to build, not just claim the keys to the kingdom, you know?

Basically remote tech + limited mechjeb + some new surveying parts to enable it sounds fun.

[Vanamonde]

This suggestion for the game's development has been moved to Suggestions and Development.

[Mr. Scruffy]

Another point about procedural vs random that people often forget: Procedural isn't 'random', and because of that it can be reproduced.

So, the "Stock System" can still exist, and you and I can both play in that same system with the same planets and the same delta-v values and launch windows that all let the same crafts work to do the same missions. But once I get bored of that, which I mostly am, then I can go and recreate a new system, and enter seed 011101110110100001111001011011100110111101110100, and have some decent data on similar planets in slightly different orbits, but have new smaller moons and a few things way out on the edge of the system that I only vaguely know or haven't found yet. I can go out, and I can find some neat things that no one else has seen yet, because no one else has ever played that system's seed.

But,... I can pass that same seed number on, and then anyone else can type that in and go play on that same system as well.

[...]

This so much. It would be an invaluable addition to the stock game, if it can be done*. As a mod? Well, if a mod can pull it off, the devs would be outright stupid not to implement it in the stock game, imho. It takes nothing away but adds to the game like a multiverse-theory. Ideally, on top of the seed, we could choose what type of star kerbol is.

*And it oughta be possible, considering old Elite-games, that came on 880K discs.

It kinda goes without saying, that we should be able to name the bodies of the generated systems in our games.

[tater]

There are only a couple reasons in game for any need to vet landing sites beforehand I can think of.

One, if there were ground features that were hard to see from orbit that were capable of damaging a lander (small-scale features in the terrain).

Two, if science returns were predicated upon picking areas of geological interest more precisely than the "biome" (should be renamed due to "bio") mechanism does.