From Duna to Mars - Realism Challenge

[DBowman]

I like your idea to have people be able to contribute / consume elements of / assets to the challenge. For those elements/assets I guess seeing your contributions used is reward enough and doesn't need points as such. I think thats a very good way of making the thing more approachable, keeping the interest of people who have 'done it' (or part of it), and adding another dimension to the thing.

Re scoring:

One idea for a scoring metric I'd been thinking about for a mission plan was 'something based on coverage area'. For example Apollo 11: 2 guys in eva suits, very short time, there is only so much ground they can cover - some fraction of the circle defined by their eva range. Later Apollos had the rover and more time and so had a bigger circle and more time to fill it in.

It sounds like Mars missions have their time span constrained by planetary orbital mechanics to be 'short' (30 days) or 'long' (500 days) - so we don't need to worry about day counting or anything, just short and long. Consider a short stay with no rovers - 2 eva guys could probably 'fill in their circle', or maybe it would take 3 or 4. With a long stay they would for sure 'fill their circle'. Ignoring rovers for a sec you could 'max out' your utilisation of a landing site either with lots of guys for a short stay or a few guys for a long stay - but there would be no point going for a lot of guys for a long stay.

Adding rovers under DRA 'Commuter' model is the same mechanic with different numbers plugged in for range: un-pressurised limited by suit eva and pressurised limited by life-support carried. One short stay un-pressurised rover couldn't max util their circle, but maybe could in long stay.

Pretend I'd worked out the numbers for real ranges and then simplified them to be easy to work with, then scoring might be like:

• 2 surface crew short stay no rovers = 2 * crew eva area * time fraction/factor = 2 * 1 * 1 = 2
  
• 4 surface crew short stay no rovers = 4 * 1 * 1 = 4
  
• 2 surface crew long stay no rovers = 2 * 1 * 4 = 8 ( they maxed their circle utilisation even though they were there for 16 times as long)
  
• 2 surface crew short stay 1 open rover = crew component + rover component = 2*1*1 + 1*16*1 = 18
  
• 4 surface crew long stay 1 open rover 1 pressure rover = 4*1*4 + 1*16*4 + 1*64*4 = 384
  

I think don't try to 'game balance' it - just simplify some realistic geometry calcs so it's not too onerous / confusing.

The DRA Mobile Home profile is more complex maybe because it can have 'sortie rovers' so it's coverage... - but if we don't like the 'Commuter' version above then there is no point thinking on it.

what I like about this scheme:

• non-arbitrary / natural / reality based (inspired?)
  
• has a built in upper limit per landing site - max circle util
  
• has a natural / motivated incentive for rovers and mobility
  
• maybe less susceptible to freaky min max strategies - but if it is susceptible then it's a consequence of 'reality'
  
• if we have launch mass per time / construction time limits then entries will cluster around 'quantitated local maxima' because there are only so many combinations of long/short * sites * rovers * crew. Clustering contenders is good for competitive spirit, also avoids the 'damn it piped by 0.0001% because...' effect. At those 'score hot spots' some efficiency measure (mass $$) would become important (damn it pipped by ...).
  

I'm still thinking re achievements and such but wanted to get the scoring idea written down.

[DBowman]

...

My big question is whether my proposed challenges seem at all interesting to you? My thought was to release a new challenge each week that was bite sized enough to get done, but also too hard to do right all by myself. Finding a compatible home for instance, that's a lot of work to find 3-5 possibilities. Let alone to do it in ro, stock, and stock hardmode. Same thing with the save file.

...

Oh - you want to do landing site selection 'in game' - thats interesting and could be fun. So it's kind of a challenge within the challenge? A quick google search didn't find any easy answer re 'how realistic is RSS Mars surface' - is there Olympus? etc? So send out some scan sats, scan, pick some landings, drop some robo landers and try to find 'interesting looking' places. Yeah that could be good - kind of a land grab dynamic, with always the outside chance that someone will stumble on something new. And presumably provides some learning curve (smaller lighter etc). Hard to avoid the hyper edit / out of game temptations though. I've seen a couple youtube vids of RSS Mars and it looked flat as a tack where they landed. How well can you see from the map view? Maybe we should hyper edit to look at a few 'iconic' sites?

So far for me RT seems a little imposing (but I guess I thought that about DRE before I installed it). You have to program in your de-orbit burn and rely on a pure chute landing (or know how to program KOS to burp at the last minute)? If the Mars end of it was doable I'd be content to 'magic' in the Kerbin end of the network. I'm not sure how save games work re it grabbing mod parts - I guess the idea is to make a minimal mod RO place the RT assets quicksave and then people can load into more modded system?

I like the idea of encouraging contenders to contribute craft files, so after a while there might be a few launcher/lander/rover designs one can grab 'off the shelf'. I know a proportion of people like replica craft, but I tend to want to work back from the requirements - not because it will end up a better launcher but because figuring out something to get the payload up is part of my fun. Though I'm sure after deciding that 'my' launcher is not heavy enough I might well be inclined to grab one rather than start from scratch. For me an open ended '110 ton to 400x400' sounds more interesting than 'Ares V HLV' - but some people are going to have the opposite view. It seems like we should accommodate Falcon Heavy fans, and even smaller launchers for the initial scan sat sections. We could have some launcher classes light / medium (Falcon Heavy) / heavy / humungo and also point to some 'authoritative' links (e.g. http://en.wikipedia.org/wiki/Comparison_of_orbital_launch_systems) and say making a replica is an achievement.

I feel the same way about mission profiles, everyone is going to have a different idea in their head or favourite 'Mars mission that never happened' that they imagine realising and the challenge shouldn't privilege any one above the other.

Just re using 'off the shelf' components for the challenge we should reflect that in achievements / medals somehow. 'soup to nuts' award, or some 'bars' along the bottom that are gold for 'self made' ( double bar / gold if someone else / enough others used it).

[vosechu]

Problems:

1. KSP units for things: it's all a bit mysterious. 1 unit of oxidizer looks like it weighs 4 liters of LOX. There are arguments over what liquid fuel is.
   
2. Figuring out reaction rates and energy required: I just plugged in very high numbers to make it easy to see it working, but I haven't figured out any real numbers for input/output. DRA5 says 300 days @ 23 kW, I didn't see a number for how much Oxidizer they thought to produce.
   
3. Power: if the US guys are right and 1 EC = 33 watts then running CO2 Decomp required about 700 EC 956 RTGs - 76 tons. Oh I see RO has 1 EC = 1kW, I've not seen what RTGs make in RO though.
   
4. Energy: I'm no chemist so I've no idea how much energy in and out the various reactions take - so we'd have to source info from somewhere. Seems like lots of mods make Sabatier require input energy (for 'game balance'?) when it's exothermic and generates heat.
   

None of this sounds insurmountable given some time and effort.

Agreed. Not insurmountable at all.

I'll add these to the front-page. I wish I could just give you access to edit the front-post at your whim!

[vosechu]

I like your idea to have people be able to contribute / consume elements of / assets to the challenge. For those elements/assets I guess seeing your contributions used is reward enough and doesn't need points as such. I think thats a very good way of making the thing more approachable, keeping the interest of people who have 'done it' (or part of it), and adding another dimension to the thing.

I'm glad you like the extra dimension, and I also think you're dead on that we do need some kind of primary score. It's just too important to not have it and it's a thing that a lot of people are driven towards. It's important for me to remember that my desires are not the same as everyone else's!

Re scoring:

One idea for a scoring metric I'd been thinking about for a mission plan was 'something based on coverage area'. For example Apollo 11: 2 guys in eva suits, very short time, there is only so much ground they can cover - some fraction of the circle defined by their eva range. Later Apollos had the rover and more time and so had a bigger circle and more time to fill it in.

It sounds like Mars missions have their time span constrained by planetary orbital mechanics to be 'short' (30 days) or 'long' (500 days) - so we don't need to worry about day counting or anything, just short and long. Consider a short stay with no rovers - 2 eva guys could probably 'fill in their circle', or maybe it would take 3 or 4. With a long stay they would for sure 'fill their circle'. Ignoring rovers for a sec you could 'max out' your utilisation of a landing site either with lots of guys for a short stay or a few guys for a long stay - but there would be no point going for a lot of guys for a long stay.

Adding rovers under DRA 'Commuter' model is the same mechanic with different numbers plugged in for range: un-pressurised limited by suit eva and pressurised limited by life-support carried. One short stay un-pressurised rover couldn't max util their circle, but maybe could in long stay.

Pretend I'd worked out the numbers for real ranges and then simplified them to be easy to work with, then scoring might be like:

• 2 surface crew short stay no rovers = 2 * crew eva area * time fraction/factor = 2 * 1 * 1 = 2
  
• 4 surface crew short stay no rovers = 4 * 1 * 1 = 4
  
• 2 surface crew long stay no rovers = 2 * 1 * 4 = 8 ( they maxed their circle utilisation even though they were there for 16 times as long)
  
• 2 surface crew short stay 1 open rover = crew component + rover component = 2*1*1 + 1*16*1 = 18
  
• 4 surface crew long stay 1 open rover 1 pressure rover = 4*1*4 + 1*16*4 + 1*64*4 = 384
  

I think don't try to 'game balance' it - just simplify some realistic geometry calcs so it's not too onerous / confusing.

The DRA Mobile Home profile is more complex maybe because it can have 'sortie rovers' so it's coverage... - but if we don't like the 'Commuter' version above then there is no point thinking on it.

I absolutely love this.

From the DRA5 I recall that the walking range was considered to be 20km and that an unpressurized rover was not considered a range extender, just basically a robot suit that helped them get their jobs done. I wonder how much ground you could cover in 30 days. Maybe it's not even ground coverage though, you might find some really fascinating spot and want to drill on it for the whole time?

The pressurized rovers still had a max of 20km, but what they would do is spend a week before the mission driving out in a direction, dropping caches (and/or other survival gear) every 15km, and then driving back to base to get the equipment they need. Then after the mission they spend another week driving out and picking up the caches again so they can put them elsewhere.

Also, I think they always go in pairs, so there's 2 in each pressurized rover, two rovers, and then two back at base coordinating and doing other stuff. Maybe they're driving robotic rover missions or something? I thought the timing aspect was interesting about laying caches though.

Maybe in addition to area we could have something about how many biomes can be touched and how long it takes to get there? Or is this aspect not interesting? Could we just say that the range of pressurized rovers is like 10 missions at 100km, or 100 missions at 10km? I also don't know if biomes are interesting enough or if we should count in other monuments like named craters, landing sites of other real rovers, in stock we could use anomalies in addition to biomes or something like that.

what I like about this scheme:

• non-arbitrary / natural / reality based (inspired?)
  
• has a built in upper limit per landing site - max circle util
  
• has a natural / motivated incentive for rovers and mobility
  
• maybe less susceptible to freaky min max strategies - but if it is susceptible then it's a consequence of 'reality'
  
• if we have launch mass per time / construction time limits then entries will cluster around 'quantitated local maxima' because there are only so many combinations of long/short * sites * rovers * crew. Clustering contenders is good for competitive spirit, also avoids the 'damn it piped by 0.0001% because...' effect. At those 'score hot spots' some efficiency measure (mass $$) would become important (damn it pipped by ...).
  

I'm still thinking re achievements and such but wanted to get the scoring idea written down.

"[*]maybe less susceptible to freaky min max strategies - but if it is susceptible then it's a consequence of 'reality'"

Mission freaking accomplished. You've min/maxed by min/maxing your access to awesome science.

But yes, I agree completely! I love how natural it is.

In the DRA5 they talk about launching 110t every 30 days, could we use this as an arbitrary limit? I didn't get far into KCT, but I think we could probably build a config around it and magico was nice enough to offer their services in configuring. What do you think?

- - - Updated - - -

Oh - you want to do landing site selection 'in game' - thats interesting and could be fun. So it's kind of a challenge within the challenge? A quick google search didn't find any easy answer re 'how realistic is RSS Mars surface' - is there Olympus? etc? So send out some scan sats, scan, pick some landings, drop some robo landers and try to find 'interesting looking' places. Yeah that could be good - kind of a land grab dynamic, with always the outside chance that someone will stumble on something new. And presumably provides some learning curve (smaller lighter etc). Hard to avoid the hyper edit / out of game temptations though. I've seen a couple youtube vids of RSS Mars and it looked flat as a tack where they landed. How well can you see from the map view? Maybe we should hyper edit to look at a few 'iconic' sites?

Oh, good question. I'll have to investigate; I would hope that it would be fairly realistic, I assumed NASA has good altimetry maps but I really have no idea.

I'll definitely hyperedit and investigate as soon as I can find some computer time. Duna is probably a little easier because we know that it has anomalies, but I don't know how the biomes look or what features are available. Are there scarps or canyons? I don't really know. This is probably my weakest point in the game, I never actually explore the planets, just a flag and go; something I can surely improve on.

So far for me RT seems a little imposing (but I guess I thought that about DRE before I installed it). You have to program in your de-orbit burn and rely on a pure chute landing (or know how to program KOS to burp at the last minute)? If the Mars end of it was doable I'd be content to 'magic' in the Kerbin end of the network. I'm not sure how save games work re it grabbing mod parts - I guess the idea is to make a minimal mod RO place the RT assets quicksave and then people can load into more modded system?

RT isn't really bad; it's mostly just one more thing you can forget to activate when you're getting to orbit.

There's a couple mods that greatly help in landing. Arming parachutes and setting their altitude is pretty critical, but there's a couple separatron-like engines that activate at the last second to help landing that RT recommends. On the other hand, when you have kerbals on-board it doesn't matter.

I would be okay if people didn't want to do RT, but I think it adds a really interesting dynamic for me. I also know that setting up all those geosats is a pain in the ass and is most likely the thing that prevents people from using it; having that pre-setup would be a huge help.

My experience of savegames is that they pretty much just work as long as you have the mods that were used to create it. I transfer my saves between computers all the time with no problems.

I like the idea of encouraging contenders to contribute craft files, so after a while there might be a few launcher/lander/rover designs one can grab 'off the shelf'. I know a proportion of people like replica craft, but I tend to want to work back from the requirements - not because it will end up a better launcher but because figuring out something to get the payload up is part of my fun. Though I'm sure after deciding that 'my' launcher is not heavy enough I might well be inclined to grab one rather than start from scratch. For me an open ended '110 ton to 400x400' sounds more interesting than 'Ares V HLV' - but some people are going to have the opposite view. It seems like we should accommodate Falcon Heavy fans, and even smaller launchers for the initial scan sat sections. We could have some launcher classes light / medium (Falcon Heavy) / heavy / humungo and also point to some 'authoritative' links (e.g. http://en.wikipedia.org/wiki/Comparison_of_orbital_launch_systems) and say making a replica is an achievement.

That is beautiful. I hadn't thought about making replicas an achievement in addition to just a general lifter but I really like the immersion of it. "Reality immersion" seems to be one of the guiding values for both of us (even if we don't always choose it) so I like the idea of having either situation and rewarding people who put extra work into making it more immersive.

I feel the same way about mission profiles, everyone is going to have a different idea in their head or favourite 'Mars mission that never happened' that they imagine realising and the challenge shouldn't privilege any one above the other.

Just re using 'off the shelf' components for the challenge we should reflect that in achievements / medals somehow. 'soup to nuts' award, or some 'bars' along the bottom that are gold for 'self made' ( double bar / gold if someone else / enough others used it).

Oh! I really like that! I love the idea of getting to claim bars if someone used your solution, but also having an equally awesome looking thing for people who really do the soup to nuts thing.

I have some graphic designer friends that I can talk to, but they're going to want some specifics before they start. Actually, one of them makes mission patches for our rocketry club so he'd be great; but I want to make sure that we have our ideas nailed down first. What elements do you think should be present in our patch? Red seems obvious, or something mars looking, but what other elements? Or is it a mission shield that adds components based on the mission parameters? Like maybe it's a red rock or a grey rock depending on whether you made it, with a satellite above, a lifter below, a robot rover, a kerbal'd rover, a lander, and a tent icon? And if you submitted a craft that someone said they used you get to use the gold version of those icons? It occurs to me that most of these are actually present in the J5JS flags pack and the icons below my signature, so I might be able to strap something together. Maybe I could even get the ribbon guy to create a little thing for us!

[vosechu]

If you haven't seen the ribbon options, here's a screenshot. I'm guessing we could use any of these to assemble something, we'd just need to give them meaning if it was different from the meaning already in the generator: https://www.evernote.com/shard/s49/sh/9804f0ec-5b5d-4a66-b06c-2b9eefbeabe6/08f13a1ff991f7c41c94a4e58f3958b0

Here are some of the challenges I saw. I haven't posted them because it's a chore to do so, but I will eventually.

Challenge #1 - Find candidates for landing sites

Challenge #2 - Build a savegame w/ existing assets in place

Challenge #3 - Create a model of the Ares V Heavy Lift Vehicle

Challenge #4 - Create a model of a cargo transit vehicle (Cargo MTV)

Challenge #5 - Create a model of a crewed transit vehicle (Crewed MTV)

(needs work) Challenge #6 - Create a model of a common descent vehicle

(needs work) Challenge #7 - Create a model of a CEV and the Ares I lifter

(needs work) Challenge #8 - Create a model of an ascent vehicle

(needs work) Challenge #9 - Create a model of an in-situ fuel and oxygen generator

(needs work) Challenge #10 - Create a model of a robotic rovers and their delivery and unloading system

(needs work) Challenge #11 - Create a model of an unpressurized rovers and their delivery and unloading system

(needs work) Challenge #12 - Create a model of a pressurized rovers and their delivery and unloading system

(needs work) Challenge #13 - Create a model of a on-ground habitat module

(needs work) Challenge #14 - Create a beautiful mission badge which represents this awesome challenge

(needs defining) Challenge #?? - Plant a flag on {{x}} super important anomaly

(needs defining) Challenge #?? - Using models or your own craft, fly the mission (wow, this is like 16 other challenges or possibilities...)

(needs defining) Challenge #?? - Using waypoint manager, define a mission of interest on the surface of Mars

Now that I'm confronted with how many challenges this is I'm wondering whether the badge is composeable.

I had intended to release one or two challenges each week so that people had a feeling of progress. What do you think about that idea?

-- Update --

Just took a moment to throw a couple extra challenge descriptions on the front page if you have a moment to look at them. I also put a sub-challenge to make the Ares V as opposed to just a ship capable of lifting.

Also, I realized that almost all the challenges are craft creation, but it seems like there's some pretty cool opportunities that are not just development. For example, I'd love to see some waypoint manager missions that have you trecking around and planting flags on important monuments. Or maybe I just need to propose a Contract Pack for this. Anyways, point is, there's more to the mission than building things, but I've been buried in the DRA5 so it's hard to extract from that.

Edited April 23, 2015 by vosechu

[DBowman]

I like the ribbon symbols - I'd like the 'meteor' for crew aerocapture at Mars

I know what you mean about being able to compose lots of items. There are a few ways around that:

- multiple ribbon patches: I mean the ribbon gen has a section for each body - we could have one for each phase, and each might have it's own 'bling'. Any Mars mission will have some common phases: LEO (Green and Blue), TMI (Black), LMO (Black & MarsRed), EDL (Pink? & MarsRed), Surface, LMO, ... etc each with it's own unique bling.

- extra ribbons for hanging 'global' stuff: DREHard, RO, NoNukes, ISRU, ...

well I guess that's really one way - 'use more space'

I wouldn't worry too much about you having been vehicle focused - sounds like that's what you are reading up on right now.

I'll try to put down something more concrete re the 'area coverage' scoring model: re-read DRA re that, and then synthesise that + 'what I think makes sense'. Then I guess I'd need to RO install so I can check Karbonite & hyper edit to 'check out Mars model'.

[Kiwa]

Somehow, the entire proposal feels a bit... (a) missing the beauty of simplicity, and ( railroading. Like... "build X, then build Y, then Z", instead of "achieve X and Y under the conditions Z". Which ends up being a totally different vibe than the Duna Permanent Outpost Challenge.

Also, maybe it is cramming too many separate things into one challenge? Example (1) - the RT part: building a "perfect" RT system for full Duna/Mars coverage is a challenge of its own and doesn't relate to a realism-based "go to Mars" challenge where it would be insane to waste money/resources on a full-fledged comms network when instead comms can be piggy-backed onto other required exploration assets and some comms outages are acceptable. Example (2) - the replicas: building replicas is something different than "going to Mars together" and exploring/achieving something there.

Edit: just to clarify, I think building a perfect comms network would be a good challenge, especially as it is also simple to score based on actual coverage (percentage, reliability under conjunction), redundancy and cost. And the results could form a good foundation for all the other Mars-challenges, so it has real value. The replicas are rather more limited to that particular special interest group, and scoring replicas is also really difficult; putting more than a token focus on replicas will likely limit the audience to that same SIG; trying to accommodate other people by adding options to skip the replicas will dilute the challenge and make entries non-comparable. In a general audience challenge, replicas might be better limited to a bonus point to the score here and there.

Edited April 24, 2015 by Kiwa

[DBowman]

I'm with you Kiwa in not wanting the challenge to be 'on rails'. I'd been thinking of replicas for example as being non scoring 'badges' or something - not everyone wants that badge, and the entry would be scored like all the others and be comparable with them.

I have put some more thought into a scoring system. I thought we needed it to be 'natural' and 'reality based' not 'game balance based' - but somehow be satisfying, competition driving, & game-worthy.

I'll introduce the ideas and rationals and then there is the simple table summary at the end. Have a read, consider, and give me any feedback:

• Due to orbital mechanics Mars missions are either Short Stay (30 days) or Long Stay ( 500 days ). So a LS is about x15 a SS.
  
• Score by exploration / science gathered by 'surface exploration elements' (SEE) operating from some 'centre of operations' (COO). Once a piece of terrain has been utilised it is exhausted.
  
• A guy on foot can 'utilise' 500x1000 m/day - visually scan / surveil, identify worthy science / sample targets, and 'consume' them. Some times there will be 'nothing to see here, move along' and some times will be intensive in depth work on a small area / feature - this tries to be some kind of average expectation.
  
• Call the area a suit EVA can consume in a 30 day Short Stay 1 point.
  
• Suit EVA range is 10km; 10 out, poke around, 10 back. This fits okay with KSP 6hr suit life support and walking at 5 km/h. Lunar EVAs were as long as 7.5 hours. 10 km range is 314 km^2 - so lets say 600 of the 500x1000 terrain patches, thirty days => 20 points are available in suit EVA range of the centre of operations. Two exploration elements on a short stay would score 2, on a long stay min( 20, 2 * 15) = 20.
  
• Rovers extend the radius around the centre of operations, the number of terrain patches, but don't effect how much can be consumed a day. You can go further quicker, but it still takes time to grok the terrain patch.
  
• Unpressurised / open rovers range is limited to 15 km to allow the crew to 'walk back' in the event of a breakdown. (per NASA DRA). This x2 the available area / score.
  
• Pressurised rover range is limited to 50 km radius (50 out, 50 back, 30 tooling around) over 15 days. They anticipate having to service & prep these rovers so you might only get 10 trips during a Long Stay even though you are there x15 as long. This is x25 the available area / score.
  
• Mobile Homes move the 'centre of operations' periodically, exposing new terrain patches. Driving 20% of the time at 5 km/h (I know but NASA thinks thats fast due to powering the things) means a 90 km track for Short Stay and 600 km for Long Stay. So in a Short Stay x4 the area, in a Long stay x30 the area.
  

Short Stay Scoring:

Suit Eva ........1 per SEE, up to .20, or up to ...60 for Mobile Home

Open Rover ......2 per SEE, up to .40, or up to ..120

Pressure Rover .25 per SEE, up to 500, or up to .1500

Long Stay Scoring:

Suit Eva .......15 per SEE, up to .20, or up to ..400 for Mobile Home

Open Rover .....30 per SEE, up to .40, or up to ..800

Pressure Rover 250 per SEE, up to 500, or up to 10000 (note x10 per onto x15)

• Robotic only exploration elements score x0.5 (oops dust obscured my camera)
  
• Including a drill rig x2 ( needs humans (oops the bit is stuck) and a rover (1000 kg) )
  
• Including a Science Lab at a centre of operation and ability to do sample return will x2 any points from that COO.
  
• SEE & COO are minimum 2 man units, buddy system. Using a non pair will put you in the 'thats not safe' division, with possible 'merit badges' 'lone prospector' and 'crazy hermit'.
  

Scoring examples:

Apollo style short stay: 1

A short stay 8 crew with 2 open rovers and one pressure rover with a drill: 2 * 2 + 25 * 2(drill bonus) = 54

A short stay 6 crew with 2 open rovers and a Science Lab & sample return: 2 * ( 2 * 2 + 25 ) = 58

A long stay 10 crew with 2 open rovers, 2 pressure rovers, 1 drill, Sci Lab = ( 40 + 250 + 250 * 2(drill bonus) ) * 2 (sci bonus) = 1580

A long stay with mobile home, Sci & sample return, and 8 lone prospectors on foot = ( 15 * 8 ) * 2 = 240

• Maybe a pair of open rovers could count as one SEE with a longer range e.g. up to 80 or something since the crew wouldn't have to walk back.
  
• there is something funny about mobile home vs pressure rover - a PR with half the crew dragging a trailer full of life support is a mobile home...
  
• Once we check out RO Mars some locales could be xX points.
  
• if you think the terrain patch is too big, I don't think that matters since everything will scale with that the same relative to each other.
  
• we could dial back the power of pressure rovers by moving back to 10 day range (cf long /short range in DPOMAC)
  
• I couldn't see how to 'rate' actual rovers speed so went with nominal 'everything is 5 km/h' - maybe you could imagine using EC as a proxy for speed (power = speed) so that rover design played into things.
  
• the orbital mechanics would imply a different long to short multiplier - but thats easy to work out.
  

Edited April 25, 2015 by DBowman

[DBowman]

I installed RO.

I used ckan with default RO recommendations. It hung and I stripped everything out but the 'essential RO' and it ran, I'll have to add things back by halves to see what the issue was.

ohhh, everything is huge and wonky - is the real solar system so inclined? maybe phobos & demos were throwing me off...

I added Hyperedit. Lets check out Mars terrain - explode. Ship landing works ok, but orbit velocity adjust - explode. JKR adjusting physics just before it blew, removed it - explode. um FAR and forces? remove - explode. DRE? remove - Mars! Did a couple of spot checks:

that chasm & Olympus (big)

Javascript is disabled. View full album

Seems we should be able to find some landmarks and such for 'waypoints' exploration achievements or whatever.

[vosechu]

Wow. I just checked out these two threads about running this exact challenge, but only doing it in stock and only doing it for one location. We're definitely undertaking something significant, but ultimately I think we're doing something vastly cooler than has been done before.

Stock version, great imgur album: http://forum.kerbalspaceprogram.com/threads/92434-Duna-Constellation-mission-pack-0-90-broke-some-parts-1-0-rebuild-coming-soon

Modded version, great video: http://forum.kerbalspaceprogram.com/threads/40559-All-craft-from-my-Constellation-mission-vid

It's fascinating how similar they are. I don't know if that's because of the DRA5 or the stock version was inspired by the modded version.

- - - Updated - - -

Somehow, the entire proposal feels a bit... (a) missing the beauty of simplicity, and ( railroading. Like... "build X, then build Y, then Z", instead of "achieve X and Y under the conditions Z". Which ends up being a totally different vibe than the Duna Permanent Outpost Challenge.

Also, maybe it is cramming too many separate things into one challenge? Example (1) - the RT part: building a "perfect" RT system for full Duna/Mars coverage is a challenge of its own and doesn't relate to a realism-based "go to Mars" challenge where it would be insane to waste money/resources on a full-fledged comms network when instead comms can be piggy-backed onto other required exploration assets and some comms outages are acceptable. Example (2) - the replicas: building replicas is something different than "going to Mars together" and exploring/achieving something there.

Edit: just to clarify, I think building a perfect comms network would be a good challenge, especially as it is also simple to score based on actual coverage (percentage, reliability under conjunction), redundancy and cost. And the results could form a good foundation for all the other Mars-challenges, so it has real value. The replicas are rather more limited to that particular special interest group, and scoring replicas is also really difficult; putting more than a token focus on replicas will likely limit the audience to that same SIG; trying to accommodate other people by adding options to skip the replicas will dilute the challenge and make entries non-comparable. In a general audience challenge, replicas might be better limited to a bonus point to the score here and there.

This is really helpful Kiwa and I think you're right on.

One of the ideas we had talked about was the idea of various levels of completeness for a particular challenge. For instance, for the comms network savegame challenge, a bronze might be a couple sats that can technically reach Mars and one satellite at mars, silver might be a group of geosync satellites at Earth with the right number of satellites at Mars, and gold might be replicas or something.

I guess the point I want to get to is:

• Bronze: technically allows the mission to go forward, but may be deficient in a number of ways
  
• Silver: some additional care towards the challenge that will make things more pleasant in some aspect
  
• Gold: clearly this is something that people care about and are willing to really go for in order to make people's missions more educational, pleasant, or realistic.
  

It was also pointed out that we need to accomplish this for both RO and for Stock-hardmode (and maybe Stock pure if we want to do that). Having a gold in RO doesn't really help the stock people at all, so they're different savegames and different challenges.

Does that make it seem any more fun Kiwa?

- - - Updated - - -

I'm with you Kiwa in not wanting the challenge to be 'on rails'. I'd been thinking of replicas for example as being non scoring 'badges' or something - not everyone wants that badge, and the entry would be scored like all the others and be comparable with them.

I have put some more thought into a scoring system. I thought we needed it to be 'natural' and 'reality based' not 'game balance based' - but somehow be satisfying, competition driving, & game-worthy.

I'll introduce the ideas and rationals and then there is the simple table summary at the end. Have a read, consider, and give me any feedback:

I wanted to put two things in writing: first, I love this. Second, I nominate you for officially being the head of primary score research and development.

I'm having trouble groking the numbers, but it all seemed pretty reasonable to me. Right now I feel so mired in thinking about how the hell to get things off this damn planet; I can't even think about the scoring aspect. That said, I know that it's HUGELY important, and I can't thank you enough for doing this research and work! If you're willing to take it on, I'd love to put your words in the top post when you think you have something ready to go up there. (/me wishes, yet again, that you could edit the top post)

I didn't understand the bit about a PR + LS is a mobile home? What did you mean by that?

On Thursday I took some time to get a basic RO savegame working and I'll upload it today. I don't have any of the Mars orbiters actually in orbit yet, but I'll have one in orbit by the end of tonight.

Also, I wanted to throw something by you for formatting advice. What do you think of this for the scores section (check out the spoilers for challenge #1):

[table=align: left]

[tr]

[td]Challenge[/td]

[td]Realism Overhaul[/td]

[td]Stock Hardmode[/td]

[td]Stock Pure[/td]

[td]Links[/td]

[/tr]

[tr]

[td]

Requirements, links, etc.

[/td]

[td]1023 by DBowman[/td]

[td]1002 by vosechu[/td]

[td]300 by Chuck's Cat[/td]

[td][/td]

[/tr]

[tr]

[td]

We need to find a home, but we also need to understand how much RO's version of Mars is like the real Mars. So we need to evaluate sites using things like ScanSat and compare them with important features on the real mars. Ideally, we'll be able to use these sites to go out and actually put flags down on Duna in the right places!

Your solutions will be used by almost everyone who attempts this challenge!

Requirements:

• Broad flat area, 10-20km diameter
  
• Kerbal rover access to as many biomes and monuments as possible (within 100-200km)
  
• Robot access to areas of biological concern, recent craters, mid-latitude gullies, "pasted-on" terrain, thermal anomalies, very young volcanic rocks.
  
• Plentiful in-situ resources
  

[/td]

[td]Bronze[/td]

[td][/td]

[td]Silver[/td]

[td]

Solutions:

Stock-hardmode

• xxx.xxN by xxx.xxW - username
  

Realism Overhaul

• xxx.xxN by xxx.xxW - username
  

[/td]

[/tr]

[tr]

[td]

Requirements, links, etc.

[/td]

[td]Silver[/td]

[td]Bronze[/td]

[td]Gold[/td]

[td]

links, links, links

[/td]

[/tr]

[/table]

- - - Updated - - -

I installed RO.

I used ckan with default RO recommendations. It hung and I stripped everything out but the 'essential RO' and it ran, I'll have to add things back by halves to see what the issue was.

ohhh, everything is huge and wonky - is the real solar system so inclined? maybe phobos & demos were throwing me off...

Yes. It really is that wonky and inclined. Since space is relative, it's really earth that's the problem though, but I suspect that Kerbin's equator cannot be inclined, so everything else has to be. Phobos and Deimos are indeed funny little bastards though. I hear that you can achieve orbit by jumping; it's like 2m/s

I added Hyperedit. Lets check out Mars terrain - explode. Ship landing works ok, but orbit velocity adjust - explode. JKR adjusting physics just before it blew, removed it - explode. um FAR and forces? remove - explode. DRE? remove - Mars! Did a couple of spot checks.

Seems we should be able to find some landmarks and such for 'waypoints' exploration achievements or whatever.

Lol, sounds like you played Kerbal. "Did a thing, exploded"

Interesting that DRE was making you explode, I wonder if that's because of a hyperedit bug or if DRE is going to break us and nobody has ever noticed because it's so damned hard to get there.

I'm glad that the landmarks are there. The screenshots are a little underwhelming though; I wonder if KSP just isn't able to draw proper scarps and meteor impact sites?

[cadaverific]

I love where you guys are headed for this challenge. I've seen a few people complaining about the complexity and 'railroading' and i'll have to respectfully disagree. I personally enjoy the complexity as well as trying to fit within the constraints provided. Just look at the constellation mission you linked. Every portion of the mission had multiple requirements and it became one of the most popular challenges for a while. Plenty of people still took part in it(myself included. It's too bad i had to delete the imgur albums). Moving on, i think the idea of separating this into separate mod options is nice, but i feel it would be very hard to compare stock vs. RO. If you intend to go that way i would think having a separate post for stock options vs RO options. I also quite like your proposed scoring system. People should be awarded for making the effort to complete the mission in a simple yet pleasing or accurate manner. Granted i haven't read all of every post, but i though i'd add my 2 cents. I'll certainly be keeping an eye on this and would love to participate. If you need a hand or need input on a mod/idea i'd love to help as well (i have a decent amount of supersadistic RO experience).

EDIT: i completely forgot to mention why i started. I saw your min payload requirement for the ariesV and it got me thinking. It would also be interesting to try and place tonnage requirements (or perhaps goals sound better) for each module on the spacecraft. It would be nice if it was possible to take other peoples models and combine it into one complete mission. I think one could be awarded extra points for falling withing some kind of constraints for that. It could be as simple as designating a portion of the ~110t payload for each module and try to remain within those limits. Just something i thought that would be fun (if you hadn't been thinking along those lines already).

Edited April 26, 2015 by cadaverific
forgot why i started posting

[Kibble]

There should definitely bee bonuses for minimizing mission complexity, like having as few purpose-built rocket launches, and as little purpose-built hardware as possible. Of course, in the context of this challenge everything would bee purpose-built, but in real life certain kinds of launch vehicles, spacecraft modules, rocket engines, et cetera, would bee in production.

[DBowman]

yum bacterial protein from methane http://www.economist.com/news/science-and-technology/21649441-feeding-farmed-salmon-protein-made-methane-gas-guzzlers less than $2 per kg by 2018.

Should be between 10 - 20% of caloric intake.

It needs nitrogen. Mars atmosphere is 3% nitroge, so far that is the only known source (https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=4&cad=rja&uact=8&ved=0CDMQFjAD&url=http%3A%2F%2Fchapters.marssociety.org%2Fwinnipeg%2Ffiles%2FISRU%2520Atmosphere%2520Harvesting.doc&ei=YU08VbXxL6S8mAXRooEY&usg=AFQjCNFljVL2znikJjIpWx307MV6E9zn8w&sig2=_cfX_TbQLwny5svodtKCkw&bvm=bv.91665533,d.dGc not sure when that is dated).

[DBowman]

I didn't understand the bit about a PR + LS is a mobile home? What did you mean by that?

If you get a pressure rover and give it a trailer of life support and a tent/caravan then its kind of the same as a mobile home in that it can keep setting up new 'center of operation' and do pressure side trips. But my scoring scheme wouldn't 'notice' that.

I'm happy to keep thinking and refine this scoring system and I'd be pleased to write it up and have it used.

I'll think out your scores/results section - I like what you did with the spoilers as an expandable table!

Interesting that DRE was making you explode, I wonder if that's because of a hyperedit bug or if DRE is going to break us...

I think it's just that when the craft pops out of hyper edit there is some physics state that is not 'zeroed' and DRE is interpreting it as '1 sqwillion G' or something. I hadn't come across it ever except in RO so something in the RO 'adjustments' is outside what hyperedit can cope with. A feature request for Hyperedit maybe. On the other hand it means it's hard to 'cheat yourself' except to suss things out in RealSolarSystem.

The screenshots are a little underwhelming though; I wonder if KSP just isn't able to draw proper scarps and meteor impact sites?

I know what you mean, if you squint then you can appreciate the scale of the chasm for example - but a higher rez terrain model would be great. Not sure how to even start there:

• find out what model they used
  
• find higher rez data & put it through whatever process they did
  
• probably some engine limits meant they didn't 'go higher' - but if we just want HiRezMars then perhaps the engine can do
  
• if 'just' plugging in an 'all mars high rez model' is not doable - maybe some High rez patches are possible
  

Is there a 'top ten locations to see on Mars'?

cadaverific - I think we should put Stock and RO in separate 'divisions' that are 'not comparable' but we can use the same scoring system. That way people can kind of compare what it takes stock vs RO to get X done. I think it's important to have them both in the same post / on the same page so any community of contenders gets concentrated rather than separated, and to make a smooth continuum of challenge so contenders can 'progress' from one to the other.

cadaverific & Kibble re simplicity & tonnage - I think my scoring proposal will end up with groups of entries at the same score - the only way to compare then is via some cost/mass metric - shich is at least related to simplicity.

[vosechu]

Whew, finally got my Mars Recon Orbiter in place tonight and uploaded the savegame.

https://github.com/vosechu/ksp-mars-ro-challenge/releases/tag/vosechu-ro

This has 4x GPS satellites in geostationary orbit around earth, 4x polar satellites at a very high altitude (probably unnecessary since I can't make them heliosyncronous), and one SCANSat around Duna in a 93degree inclination orbit. I've never tried this before, but hopefully it makes it easier for people to start with the challenge!

@DBowman: I'll have to check out the 8k textures for RSS. I think my machine couldn't handle them the first time, but I'll see if I can do it in opengl mode or something. If that doesn't work, I'll ask the RSS makers and see what they suggest (if it's even something we can deal with. I don't know anything about how the textures are done so it'll be a learning experience for me!

Also, thank you again for taking a look at the scoring system. Your response to cadaverific and Kibble was perfect and exactly what I would have said; it's a huge relief to know that you're helping and we're totally pulling the same direction!

@cadaverific / @Kibble: Welcome to the thread! We're still sorting out the details, as you can tell, so please let us know what makes the challenge interesting to you. If you have recommendations about challenges or their requirements, that would be very much appreciated! Delving into the DRA5 to get details is very time consuming!

[vosechu]

I looked at Mars surface yesterday by crashing my satellite into the ground and learned some things. First, there are huge swaths of Mars that really are just boring and flat. My understanding is the the northern plains are young and undisturbed. But the bits by the Arian(sp?) mountains is much more interesting. So at least that makes the challenge to find a landing site more obvious to me, "find a site that's not incredibly boring"

The second thing I learned in that we most likely have terrain scatter turned off which will make the whole thing much more interesting. It severely impacts performance so most people keep it off (at least that's my understanding). So tonight I'm going to try the 8k terrain textures, with ground scatter turned up to 11, and I'm going to land between a couple mountains. I'll let you know how it goes.

[vosechu]

How do people feel about the hullcam vds mod? I don't want to add it if others can't use it or don't want it, but I think it adds something spectacularly real to the game when it's used right.

I also ask because I'd like to add a hubble probe around Earth. I guess it'll just deactivate the probe if you don't have the mod, so that's not too bad.

Similar question around SCANSat, I realized last night that my MRO satellite had included SCANSat but maybe I need to have some Mars satellites that only require RT2 and nothing else. Thoughts?

- - - Updated - - -

Btw, I turned on scatters at max, and also the 8k textures for RSS. The textures do in fact make the planet more interesting, but I don't see any new scatters. So I don't know if that's a thing or if Duna just doesn't have terrain scatters or something. The only problem is that the 8k textures only let me boot about 1/4 of the time. It often runs out of memory for me. So, I can look at ATM or something, but I'm worried that will just reduce the heavy textures I just loaded in.

More experimentation needed.

[immelman]

So tonight I'm going to try the 8k terrain textures, with ground scatter turned up to 11, and I'm going to land between a couple mountains. I'll let you know how it goes.

If you want to save your RAM go for 8k textures for just and Earth and Mars and lowest for all the other bodies. I like your Apollo, but to save time you can use the FASA Apollo, it looks great!

Putting together a Mars mission on RO really brings home just how hard this all will be. For fun I'm using mostly Apollo era tech, if you have ever read Baxter's Voyage then ChrisPBacon's Ares mission shows just how huge these can get to...

Can I enter into this? Not having the patience to use eight Saturn 5 launches to orbit and fuel my mission I am going with the proposed Saturn C-8 variant, can get 200T to a 200k orbit, even then I will need at least four of them.

And Here is the orbital stack so far, why yes that is a Saturn SII and two Saturn IVB's for TMI, Mars insertion and Mars escape

Hardest thing is a realistic MEV, interesting paper in the challenges:http://www.4frontierscorp.com/dev/assets/Braun_Paper_on_Mars_EDL.pdf

You can break this down into three problems:

1. Slowing down in Mars atmosphere, using rockets lower down means being able to ignite then in a Mach4+ jetstream.

2. Parachutes, for a lander mass of 20-40 T it's not very likely you can just use parachutes at more than Mach 1, in reality they would be ripped off.

3. If you want to use a lifting body, the heatshield/ aeroshield needs to be more than 20m...

Edited April 27, 2015 by immelman

[vosechu]

If you want to save your RAM go for 8k textures for just and Earth and Mars and lowest for all the other bodies. I like your Apollo, but to save time you can use the FASA Apollo, it looks great!

Putting together a Mars mission on RO really brings home just how hard this all will be. For fun I'm using mostly Apollo era tech, if you have ever read Baxter's Voyage then ChrisPBacon's Ares mission shows just how huge these can get to...

Can I enter into this? Not having the patience to use eight Saturn 5 launches to orbit and fuel my mission I am going with the proposed Saturn C-8 variant, can get 200T to a 200k orbit, even then I will need at least four of them.

http://i.imgur.com/U3vwavg.jpg

And Here is the orbital stack so far, why yes that is a Saturn SII and two Saturn IVB's for TMI, Mars insertion and Mars escape

http://i.imgur.com/vPgVbyC.jpg

Outstanding! I hadn't thought about using the ships that come with RO, super great idea. I wonder why NASA didn't use this in the DRA5? But maybe that's just legacy of the DRA5 being slightly out of date. I read somewhere that NASA was thinking about updating it but didn't get the funding together or just decided that the funding could better be used elsewhere until plans were firmer and funding was a little more assured; so maybe that's why.

It is staggeringly hard though right? And if you look at the plans from DRA5, they wanted to do it three times in super different locations. But I had exactly the same reaction when I started trying to launch things into orbit.

Since that ship is already publicly available, if you endeavor to find a credit I'd love to mark you down as the first winner of one of the challenges. I think it'd be appropriate to share the credit with the person who originally made the craft for RO, but you deserve credit for doing the obvious thing and using what is already available!

Hardest thing is a realistic MEV, interesting paper in the challenges:http://www.4frontierscorp.com/dev/assets/Braun_Paper_on_Mars_EDL.pdf

You can break this down into three problems:

1. Slowing down in Mars atmosphere, using rockets lower down means being able to ignite then in a Mach4+ jetstream.

2. Parachutes, for a lander mass of 20-40 T it's not very likely you can just use parachutes at more than Mach 1, in reality they would be ripped off.

3. If you want to use a lifting body, the heatshield/ aeroshield needs to be more than 20m...

I hadn't really thought about this, what a fascinating problem. Does it matter if they're ignited before entering atmo? Or maybe the problem gets even worse because then you have hot exhaust being shoved back towards the combustion chamber. Wow, that's brutal. Thanks for opening my eyes! I'll post this on the front-page under the resources section.

[immelman]

Outstanding!

Since that ship is already publicly available, if you endeavor to find a credit I'd love to mark you down as the first winner of one of the challenges. I think it'd be appropriate to share the credit with the person who originally made the craft for RO, but you deserve credit for doing the obvious thing and using what is already available!

The heavy launch vehicle is all my own work, I got the proposed dimensions from Encyclopedia Astronomica:

http://www.astronautix.com/lvs/saturnc8.htm

If you haven't take a look, its fascinating seeing what might have been using the Saturn platform!

[vosechu]

The heavy launch vehicle is all my own work, I got the proposed dimensions from Encyclopedia Astronomica:

http://www.astronautix.com/lvs/saturnc8.htm

If you haven't take a look, its fascinating seeing what might have been using the Saturn platform!

That's even better! Can you post the craft somewhere I can link to? It's gorgeous, so much prettier than mine!

- - - Updated - - -

Or if you want I can give you access to my github repo.

Honestly, it's making me realize that I need a place where we can put compressed .sfs and .craft files. I'm not sure the best solution, but there has to be something!

- - - Updated - - -

I've updated the top post to note that the RO effort is officially under way. I've not been able to identify good landing sites yet, but I'm working on it still. I have great SCANSat maps now, but I need to study them a little bit more. It turns out that there are actually anomalies on Mars (not Duna). I'm not sure what they are yet, but it seems like a good start for locations!

Also, I poked around the RSS threads and it looks like terrain scatter has been specifically disabled because it doesn't work with planets so large. Unfortunately it's been disabled until rbray is able to work on it and he burned out recently. He's back now, but I'm guessing he's going to take it a little slower for a while. So we may just have to find interesting features in other ways for now.

I keep looking back at stock now and thinking how easy everything was. Maybe I'll also do this mission in stock as well.

Edited April 27, 2015 by vosechu

[DBowman]

I tried the 8K D** but the game would 'out of memory'. So we reverted to 2k and just pasted 8k mars over it. The game still had v high memeory usage, so not sure how it will go with more mods etc. Is Active Texture Management and/or the D** stuff something that compresses textures fairly loss-less-ly?

A few nice landmarks from looking at NASA topo as rendered in game at the 2 resolutions. For me the summary is 'looks great from orbit' and 'some will look ok as a large scale landscape'. Finding scenic spots and/or interesting traverses ("Up and out of Valles Marineris", "Olympus Ho" ) and "precision required landing spots" could make a 'land grab mechanic' - but inevitably there would be hyper edit finding so it would make sense to separate 'finding a site' and 'landing rank'.

There was a weird 'anomaly' you can see in the pics of Olympus - it was looking great from km up - drop a few 100 m and it rezzes down!

I know it's a high art building terrain engines, but I wonder what it would take to do better, just doubling the rez would not do it, it needs non uniform data - widely spaces heights in the boring flat parts and lots in the crinkly interesting parts.

Javascript is disabled. View full album

[immelman]

I tried the 8K D** but the game would 'out of memory'. So we reverted to 2k and just pasted 8k mars over it. The game still had v high memeory usage, so not sure how it will go with more mods etc. Is Active Texture Management and/or the D** stuff something that compresses textures fairly loss-less-ly?

A few nice landmarks from looking at NASA topo as rendered in game at the 2 resolutions. For me the summary is 'looks great from orbit' and 'some will look ok as a large scale landscape'. Finding scenic spots and/or interesting traverses ("Up and out of Valles Marineris", "Olympus Ho" ) and "precision required landing spots" could make a 'land grab mechanic' - but inevitably there would be hyper edit finding so it would make sense to separate 'finding a site' and 'landing rank'.

There was a weird 'anomaly' you can see in the pics of Olympus - it was looking great from km up - drop a few 100 m and it rezzes down!

I know it's a high art building terrain engines, but I wonder what it would take to do better, just doubling the rez would not do it, it needs non uniform data - widely spaces heights in the boring flat parts and lots in the crinkly interesting parts.

http://imgur.com/a/SVjKy

In the end I went to making a linux partition to play Realism Overhaul in KSP, it was the only way to manage the memory usage of all the mods needed.

The DTSloader sure makes the game start much quicker and saves some RAM. In addition I delete almost all the fairings and fuel tanks (since I use proceedural).

Regarding landing sites, I've wondered about making a "featureless" Mars with only a small section in great detail using NASA topo to generate a height map. Specifically for the landing area, not only save RAM but also to make it more interesting, should be possible? Thats going to have to wait a couple of weeks though

[vosechu]

I tried the 8K D** but the game would 'out of memory'. So we reverted to 2k and just pasted 8k mars over it. The game still had v high memeory usage, so not sure how it will go with more mods etc. Is Active Texture Management and/or the D** stuff something that compresses textures fairly loss-less-ly?

A few nice landmarks from looking at NASA topo as rendered in game at the 2 resolutions. For me the summary is 'looks great from orbit' and 'some will look ok as a large scale landscape'. Finding scenic spots and/or interesting traverses ("Up and out of Valles Marineris", "Olympus Ho" ) and "precision required landing spots" could make a 'land grab mechanic' - but inevitably there would be hyper edit finding so it would make sense to separate 'finding a site' and 'landing rank'.

There was a weird 'anomaly' you can see in the pics of Olympus - it was looking great from km up - drop a few 100 m and it rezzes down!

I know it's a high art building terrain engines, but I wonder what it would take to do better, just doubling the rez would not do it, it needs non uniform data - widely spaces heights in the boring flat parts and lots in the crinkly interesting parts.

http://imgur.com/a/SVjKy

Great work dbowman. Atm probably wouldn't even effect the planet textures if you get the dds version of them. But if you get the regular version then it'll certainly lose a lot (and take forever). At least, that's my guess, I don't really know anything about the internals.

I like the idea of pasting the 8k version in for just Mars and earth and leaving the rest as 2k, that would probably take me down to the regular limit. Hell, if pushed we could probably put an empty file for the planets we aren't visiting.

I think the real problem is the lack of terrain scatters though. Even if we had a 32k image I doubt the satellite has small enough resolution to render individual rocks. But I don't know what the problem was and whether it's something we could help with.

At any rate, I like your idea of just trying to do our best at finding challenging or interesting landing spots even if they look like orange pudding. Maybe we have to leave the looks to the stock part of the challenge?

- - - Updated - - -

In the end I went to making a linux partition to play Realism Overhaul in KSP, it was the only way to manage the memory usage of all the mods needed.

The DTSloader sure makes the game start much quicker and saves some RAM. In addition I delete almost all the fairings and fuel tanks (since I use proceedural).

Regarding landing sites, I've wondered about making a "featureless" Mars with only a small section in great detail using NASA topo to generate a height map. Specifically for the landing area, not only save RAM but also to make it more interesting, should be possible? Thats going to have to wait a couple of weeks though

That's an awesome idea. Maybe we could talk to the eve crew and see if that's a possibility. Right now I think the whole map is one image, so the resolution would still be huge, but maybe if dxt does the compression we could leave those chunks of the normal map empty and not suffer. I fear that wouldn't be the case though and we might have to wait for planets that load multiple images. :/

It would be interesting to ask the eve people what they think! Do you mind reaching out the them Immelmann?

[DBowman]

.. only a small section in great detail using NASA topo to generate a height map.

++ to this. Being able to overlay an area of high detail terrain over a low rez planet would be great. Do you know is it possible? the current stuff seems to be based on a 'uniform mesh' (the .pngs) ... who would know? the eve guys? the RSS guys?

If it was possible it's a natural game play mechanic. If you do the work of up-rezzing some particular area you will likely get others wanting to use it (kudos++), either cause it's and uber-cool site or because they don't want to take on something like that for themselves.

- - - Updated - - -

We talked a little about some landing site aspect to the challenge:

• flavor - just looks cool and/or the site has some personal resonance for the contender.
  
• achievements - Some sites probably require pin-point landings and/or 'ATV' landers etc. Historical sites for the re-enactors. Lists; Low Helas + High Olympus, The big 5 craters.
  
• finding somewhere that looks cool (or making it look cool via high rez patch) / has resonance and contributing it makes the challenge better for everyone else.
  

I had done some browsing through http://en.wikipedia.org/wiki/Geography_of_Mars which has a drill-able quadrant map with some nice topo maps.

I think we should at the least link this (or a better resource if anyone knows of it) to the 'intro' section of the challenge, get the imaginations going.

Here's the start of a 'landmark list':

Olympus Mons 27 km high, datum + 21.2km

Valles Marineris 4000 200 7 km long wide deep

Hellas Basin - largest impact crater deepest point -8.2 km

N pole Planum Boreum

S pole Planum Australe

quadrants:

Tharsis Quadrangle

Ascraeus Mons, Pavonis Mons, Olympus 18N 226E

Lunae Palus

22N 86W Feskenov crater nice circle and mount

24N 62W Lunae Mensa - kind of messa in a 'river valley'

26 56W Rongxar round crater with maybe rover access to big fossa system

Hellas

-32N 62E lowest point

-39N 113E nice crater 'Greg'

-32N 93E Dao Vallis

Aeolis

-8N 175E Apollonaris Mons

Elysium

25N 147E Elysium - Niiice conical volcano!

other things that seemed important / interesting

fretted terrain - what is this? probably need HiRez to look good

'river valley' / outflow channels

lobate debris aprons

special craters, historical vehicles