This contains my entry for the Caveman Challenge 1.12 (it's about completing the tech tree without upgrading buildings)
Below I will explain my missions and strategy to accomplish the "Nano-crystalline Diamond" (or NCD) level of the Caveman Challenge, it is the absolute hardest setting in stock KSP. It is just career mode with the following limitations
No upgrading any buildings
Custom difficulty, every slider on the lowest/harshest setting
10% rewards for science , funds , reputation
Max 120% re-entry heating
0 starting funds
...you get the idea, it is ridiculously difficult the first time you see it (Details can be found on the Caveman challenge topic)
You win when all Tier 1-5 tech tree nodes have been unlocked
Green Monoliths are easter eggs that spawn in random locations (1 for each planet/moon). Discovering them on location will unlock a random available tech tree node
The most recent challenge rules allow the usage of these
"available" tech nodes in-game are defined by having their parent node(s) already unlocked, but Green Monoliths don't care about building level limitation (they can potentially unlock Tier 6+ tech nodes without you upgrading the R&D building)
The Tier 5 nodes cost 90 each, and there are 3 Grenoliths in the Kerbin system nearby. Holding out until we unlocked all Tier 4 techs, means that we can save ourselves getting 270
Holding onto contracts for the parts they provide for testing, gives us some usage of those parts, at the expense of holding 1 contract slot closed
Use small steps instead of leaps to reduce risk. Funds, kerbal lives, contract slots are all things that can end NCD runs if they run out
Green Monoliths are not shortcuts to skipping science. In Caveman mode, it takes as much time to detect anomalies (and visit all of them), than it takes to just go and collect science on Mun
The Minmus Grenolith was the easiest: There is only 1 other anomaly on Minmus, so spotting the Grenolith is simple
Mun/Kerbin have lots of easter eggs, which look identical on the Kerbnet detection system. If you are not familiar with the easter eggs on Kerbin and Mun, be prepared to visit 5-15 locations on each before getting that Grenolith
The random tech might not be Tier 5 (helping achieve the Caveman challenge). It could be a Tier 6, which sounds great, unless it is something like "Command Modules", "Adv. Fuel Systems", "Heavier Rocketry"(what are you supposed to to with a Mainsail? )
"Rover" ships work great on Minmus! 4 Wheels are enough to farm 100% Minmus science in a short time frame (possibly faster than preparing launcher for hop landers)
Mun, not so great. It needs more wheels, reaction wheels, solar panels and batteries to keep up with the higher gravity
Robotics on an "Engineer Van" can hoist a ladder some +10 m up into the air
Build rockets in the VAB that are over part count, then remove the small bits (science equipment, batteries, antennae, ...) and click launch
From the Runway (or before the main rocket was built), launch an Engineer Van containing all the missing parts, and Bill (engineer), and maybe Valentina to help assist with the heavier parts
Ride out to the launchpad, hoist them up and add the missing parts
Launchpad construction vehicles can be built with a rectangular wing as the base, it save on part count, allowing for more wheels
HG-5 antenna (a little retractably relay dish) is vastly superior to the Communitron-16. The HG-5 is heavier, but has enough range out at Minmus, and even connects to other relays better because of the improved strength (try it out here: ksp-visual-calculator.blaarkies.com)
Before you start your own challenge run
The Caveman Challenge (on normal) is fun, it pushes you to accomplish what seems impossible to 90% of KSP players
The NCD level is different; it is a commitment, expect to play it for 40+ hours, doing 100+ launches
You need to be sure that you know how to launch on a shoestring budget
Be sure to understand the career system 100%
How does science work, where to find extra science that is not immediately obvious (KSC micro-biomes)
Which contracts are best for funds, how to get around bad contracts
Be comfortable with the CommNet system; losing signal of an unmanned probe launching eastward, or behind Mun, or out at Minmus, are real issues you need to consider with every launch
Correction burns halfway en route to Minmus can become difficult because of this
There is a tool for this : ksp-visual-calculator.blaarkies.com
How much DV is required to move around the Kerbin system (and possibly Duna/Eve)
You will be building rockets that cannot be refueled, and they won't be "cheap" relative to the amount of funds that you have; they need enough DV to complete at least part of their mission
You won't have fancy colorful patched-conics orbital trajectories. Returning from Mun low-orbit, means escaping Mun at a certain angle, at a certain velocity; do you know what those are before the burn starts?
Piloting skills; you will be landing planes on Kerbin, probes on Mun, and possibly Bob (scientist) on Minmus while he is hanging onto a ladder.
Any landing needs to be easy to you because you won't have proper SAS for most of these landings
Breaking things on landing could bust the whole mission and bankrupt your space program
NCD is arguably 10x harder than Normal difficulty, since you only get 1/10 of the rewards, but profits aren't that simple. The difference between contract reward and launch cost determines how difficult it is, making it potentially even harder
If any of the previous points had you thinking too long, it is best to try the easier levels first and see how you like it
Topaz[Hard] goes by much quicker than this, and introduces a fair bit of the challenges faced here
There are levels between Hard and NCD if you feel confident in your skills
After completing any level of Caveman Challenge, you realize how overbudget we have been building craft for simple missions all along
See the Diamond run for tips on how to get going with 0 funds
Starting important tip: Don't throw darts above 30km !
They tend to dive down, nose first at the speed of sound. You can't control them, and parachutes don't work at those speeds
There are many solutions, I chose to pick a "test the decoupler" contract and use the part for my own needs for a while
We need to build a backup supply of funds, we need it to unscrew ourselves if a launch fails. Grab 2 contracts that have a high chance of succeeding (things like "test parachute at 5km", or "test thing on launchpad"). Then do a tiny launch where you hop the vehicle onto some KSC buildings. The primary objective is making funds, so complete the contracts and recover the entire craft. The secondary objective is getting science to unlock new tools. The first 10 launches from KSC will probably all yield some science.
Once we unlock more science equipment, we need to do this again. Remember to unlock tools first (the parts we need to make funds), everything leading up to placing satellites in orbit will help, science equipment comes second here.
Do small hops, nothing dangerous. After about 15+ of these we will have most KSC science grabbed for this level and some backup funds as insurance.
Go for orbit when ready, but that won't mean you are done with the KSC yet...there is still lots to be done down there
"Science Rollers" can easily be "driven" by using the trim controls (actually designed for aircraft).
Turn off SAS.
Hold down `left Alt`
and press `q`.
You should see the Roll control indicator move (bottom left of screen). But unlike when we control it with the keyboard (making the indicators max out on each key press), this trim will gently feather it to roll very slowly
You can release Alt now. This works for all controls (roll, yaw, pitch) and even translation controls (i, j, k, l, keys for RCS stuff, but i bound my rover wheel controls to them as well). Later you can make rover wheel "lead-footed" so that they keep driving forward without you there
If things get out of hand, just turn on SAS (it nullifies these trim values). You can also reset trim with `left Alt + X`
Always check to spot this one. A Terrier test contract is gold at this point, we can keep the contract (uncompleted) for 401 Kerbin days if we want, without any penalty. We don't have a Terrier part yet, but they will be extremely useful on the orbital craft we just built moments ago.
Keep this golden contract until we are ready to unlock our own Terrier part (meaning we also have to be careful about locking up a contract slot with this), but we can always complete it at-launchpad if we ever get into trouble at least.
With a Terrier onboard, it becomes much easier reaching Mun and Minmus, where we find a lot of extra science in orbit. This part gets a bit repetitive, because we might not have all the science equipment unlocked yet, meaning we have to go back later again. But since it is all orbital stuff, we will definitely be going through this region again anyway where we can scoop up left over science.
Once we are ready in terms of technology (farm all the easy science near KSC to help unlock unmanned rocket parts) we can start using probes for everything. They are cheaper, it hurts less when they are lost, and we can leave them in orbit to provide a benefit. Forget about the Communitron 16 for these, the HG-5 High Gain antenna is so much more powerful, and it provides relay.
The idea with these will be to:
build a relay communications network for future launches
scan Kerbin/Mun/Minmus for anomolies and mark them as navigation points
provide in orbit probes that can transmit science for contracts if we ever hit a dead end for funds
At first, some probes won't have solar panels. That is fine, they will relay comm signals as long as they had power the last time you checked them (if you ignore them for 100 years, they will still be functional relay sats, but if you go and control the craft personally, it will shut off). These craft will mostly help launches of newer and better craft.
Every launch should be accompanied by enough contracts to cover the launch cost. If we can't pull this off, we need to do silly launch pad contracts to make up the funds that we will need. The satellite contracts can payout 7 000 - 13 000 which cover the launch cost unless the craft has many science experiments onboard. A 2nd contract is pure profit at this point, allowing us to leave a relay/kerbnet probe in orbit somewhere.
We will need little bits of science from everywhere possible. This one is dangerous, and the payoff isn't even great, but i wanted some EVA flying science. I strapped Jeb into this...thing, let it launch and then parachute Jeb down to find some more science, somewhere.
The S2 roverwheel part is also pretty important, for clearing up the rest of the KSC science. We can get this via a contract for free, but later on it will also be useful for roving Minmus. It is worth researching, but other things are more critical.
Remember the hidden science. Some of these sound ridiculous, but they exist. "Landed at Kerbin's water"...gives a new meaning to ships finding land. There are many versions of these going on in game, like "Splashed down at Kerbin's Deserts" is obviously a craft floating in an oasis, but in the end they all give unique science that might not have been collected before.
Before landing anywhere other than Kerbin, remember to go get the contract for those. It is usually worth a lot of funds, and sometimes 1.
This part is all about the delicate balance of funds, science, and monoliths. Every satellite needs funding to launch, so we will have to find some contract mission that fits together nicely (2 satellite contracts around Minmus is a great start). Then we can determine what kind of science we gain while being there (maybe a landing is possible, or maybe it is better to leave a relay in orbit)
Don't be surprised after doing 10+ of these missions just for Minmus, we need to be thorough. Before solar panels, keep an eye on the EC supply (leave the craft before that empties out). Once the we solar panels, we can do a full KerbNet scan of where we are orbiting and mark those anomolies.
Anomalies on Minmus are easy: It has 1 black monolith (gives us a tiny bit of funds for discovering it), and 1 green monolith (unlocks a new tech for free)
It is worthwhile detecting all anomalies now, even if we don't visit/discover them (that might unlock some lowly T2/T3/T4 techs). Once we unlock the last T4 tech with science points, we can discover these anomalies for a guaranteed T5+ tech.
After the very long science collection missions, once we have all the critical parts unlocked, we can unlock the lowest tier nodes in order to hunt monoliths on Kerbin. Don't try that without planes, Kerbin has far too many anomalies to do a rocket launch for each.
We will find many false positives. These ground stations show up as anomalies, and they are actually easy to spot as long as you keep an eye on the mountain peaks.
Sometimes you get lucky finding an easter egg outside of the savegame
A well designed plane can easily get past the poles, even to Badlands.
Finally made it to Woomerang. This won't unlock the launch site (since difficulty settings disable it), but the same discovery of Cove or Island Airfield will allow you to launch from them.
After a lot of struggling on Kerbin, i decided to go get the Minmus monolith instead. They don't need Kerbals close by, only a controlled craft in physics range (some 2.3 km). Maybe this works with debris as well if the camera is tracking them.
At some point, we need larger launch vehicles. With docking ports unlocked, we can add boosters afterwards, on top of the full 18t limit. To build these forklifts, it help using a wing segment as the base structure.
...more to come, look at the Screenshots in the meanwhile