About the Tech tree

[Beowolf]

I'd like to even see "Staging" as an early item on the tech tree. When I was studying the history of space flight, I was surprised to learn how big a hassle it was getting a liquid-fueled second stage engine to fire reliably. Until they got the Titan working (see Youtube for lots of videos of exploding early Titans), all our launches had to be a single liquid booster stage, with clusters of tiny solid rockets for the upper stages.

The Atlas booster was a direct result of those problems. Atlas was an ingenious "1.5 stage" kludge that's definitely my kind of engineering! I suppose it'd count as the simplest possible version of asparagus staging.

Alternately, hypergolic fuels could be on the tech tree to handle early staging, with another advance later allowing for internal engine igniters, which enables both cryogenic upper stages, and the ability to stop/restart those engines in-flight.

Doubtless that's more realism than some will want. A simplified version would be to just add a single "starter" liquid engine that can only be ignited while on the pad. Then a tech advance would open up the more advanced engines that work as they do now.

[ddavis425]

Think harder then just that. Solar Panels weren't used till Vanguard 1. Manned flight is a lot more tech then just probes. ASAS needs computers. Aerospike are yet to be used in real space flight. Different strength engines take work look at the original Redstone rocket compared to the Saturn V engines. Docking ports didn't originate until Voskhod/Gemini. Landing legs well...Luna and Surveyor.Engine gimbal could also be a step.

But I mean that the parts we have now aren't much of a step up from each other. Look at the engines. There are different sized engines to fit with different sized tanks, and each one of them is specialized for something. Heavy lift, medium power, low power/lander, radial engines. They aren't a step up from each other, they just have different functions. I'd rather have a tech tree work by starting out with multiple inefficient engines each suited for a purpose and move up to be upgraded versions of each one. The Mainsail could be bottom-tech; it would overheat easily, not gimbal, etc., but then there would be more versions of the Mainsail, each better than the last.

[PwnedDuck]

An abstract science rating makes some sense, I guess, but how do data about the composition of Dres and the thickness of Eve's atmosphere come together to allow you to unlock larger landing legs/small radial engines...

[Frederf]

I know this departs at at least 80 degrees from the current concept of R&D, but I still have thoughts leaking all over the floor.

My vision of R&D is an evolution of parts based on player feedback. It relies heavily on being able to produce parts semi-procedurally with fine grain adjustments of attributes. At simplest example take three standard "seed" parts, engine, fuel tank, parachute. The player elects to direct the R&D department to develop new parts of these categories with directed changes. R&D "dice rolls" some new parts with more desirable stats in the preferred area with possibly some degradations in others. Given a handful of possibilities (1-10) the player accepts or rejects these potential parts into the stable of usable parts. "We need an engine with more thrust" or "this parachute needs to open higher but with less drag" or "this fuel tank needs to be shorter and stronger at the cost of capacity." Thus the player directs their own part development qualitatively and ends up with a selection of parts uniquely their own.

Gameplay restrictions could include R&D cost per role, cost to support a larger stable of actively available parts or hard limits on number.

A tech tree and R&D evolution would synergize well. One might give up an old model honed to perfection in favor of a new technology that's still raw. New techs could open up new attributes to adjust or extensions on the values allowable such that existing models could be brought up to contemporary tech.

[ngianoplus]

But I mean that the parts we have now aren't much of a step up from each other. Look at the engines. There are different sized engines to fit with different sized tanks, and each one of them is specialized for something. Heavy lift, medium power, low power/lander, radial engines. They aren't a step up from each other, they just have different functions. I'd rather have a tech tree work by starting out with multiple inefficient engines each suited for a purpose and move up to be upgraded versions of each one. The Mainsail could be bottom-tech; it would overheat easily, not gimbal, etc., but then there would be more versions of the Mainsail, each better than the last.

I would LOVE to see this in game. Now, I don't expect it now because if they take this route, there would be less of a need to develop more parts (specifically engines). Say the mainsail started worse than the current stats (no gimbal, more weight, less thrust and isp). Then you upgrade to make it have gimbal and more weight/thrust. Then you upgrade to more isp and thrust (at this point the mainsail is at its current stats), then you upgrade the thrust further and reduce the weight. Now you could do this perhaps 5 levels. 1= terrible, 2=better, 3=same as current states, 4=even better, 5=best. Now take it even further, you could do an upgrade path sort of deal. Once you get to the current stats (level 3), you can choose to upgrade the thrust, isp, reduce the weight, or further gimbal. You could have combinations of, say isp and weight. Now this may not be a good choice, it could be simplified to where upgrades 4 and 5 increase all stats a little bit. Now imagine that the texture changes with the level... it would be quite the amount of work to implement but it would make R&D so much better.

[animalcross]

When I think of Tech tree, I think of Civilization V.

While the tech tree in that game is much more strategic and also made for tailoring the development of your game entity to a specific path towards victory, it had it right in that it was relatively difficult to screw it up. Sure there were ways of making your tech gain great advantages in one way or another, no path through it was really a way of preventing progress. Analogy; regardless of which basket had eggs in it or how many, it was still possible to succeed. Such is not the case in less well-constructed trees of gameplay.

But one thing I remember is a very notable detail: the industrial era in the tech tree is a pinch point of only three technologies (to the left of the middle in this image) and how a given Civilization approached those three technologies (Biology, Steam Power and Dynamite) made all the difference in the world of how quickly and easily they would progress to further goals only a few turns down the line.

Making such career lynch-pin tech tree decisions would make it much more interesting if cleverly employed, as long as it did not become too involved.

I can imagine a few different technologies that could be on the tree:

Solid Fuels

Monopropellant Liquid

Bipropellant Liquid

Tripropellant Liquid

Hypergols

Open Cycle Combustion

Closed Cycle Combustion

Turbopump Assembly (Level 1, 2, 3?)

Nuclear Thermodynamics

Don't want to sound rude, but I am a civ nerd, and must inform you that that tree is civ V vanilla. The current tree for BNW is much changed, no more industrial pinch.

[Spyritdragon]

I love the idea of using in-flight science equipment for 'R&D points'

[AlternNocturn]

This means that that little "Career" radio button on the New Game dialog is going to be clickable, and that makes me tingle with excitement.

[DresCroffgrin]

The big thing with R&D isn't to impose a grind where parts are unlocked as a reward for something. What I want to get out of it is a mechanic through which the game presents itself gradually to players, to not welcome you with an overwhelming wall of content right from the start.

As with the content, the tree itself is being designed to expand the further you get. At, first, you get a few general technologies to research, which in turn open up new nodes which are more specialized, and so on until you're getting to some really specialized stuff towards the higher levels.

We're assigning parts to each of these nodes now with this idea in mind, and so far I think it's going rather well. Each part seems to find its place nicely in the tree we have here.

As for how nodes will be researched, here are my thoughts:

We can't impose any amount of time to unlock something. You could just sit on the pad and timewarp if that were the case, and that's not the sort of gameplay we'd like to see.

I also don't like the idea of spending money directly to research something, as that would reduce R&D to be just a shop catalog.

Instead, I think the most authentic mechanic would be for nodes to require a 'science' rating, which you earn through experimentation (flying) and gathering data (packing science gear aboard possibly). This system opens up a lot of very cool possibilies for later, which will give a lot of meaning to what you can do out there on your missions, even before we have an actual game economy or contracts.

To paraphrase a friend developer, it's not about realism, it's about authenticity. Realism means putting you into a simulator cockpit and making you wait around for months to reach another planet. Authenticity means making sure what the game lets you do feels right, even if it's presented with some degree of abstraction, like using a numerical value to represent a more complicated process.

Cheers

Every time I read a lot of discussion about upcoming features, I get worried that the game will implement either some sort of creativity-destroying realistic feature, or an immersion-breaking gimme. Then, I read the devs' responses and play the actual game, and am always amazed by what I see.

Thank you, devs.

[joppiesaus]

Remembers me of the Age of Empires II tech tree:

[Akhar]

Well, if part failures are implemented for new tech, I'd like to see them as something that requires thinking. Not like "Every 3 minutes there's a 10% chance that your engine will derp." But "If you run this engine at higher throttles in thinner atmospheres/vacuum could damage it."

[daffy]

I'm still trying to figure out what Delta-V means! Tech tree???????

[Sandworm]

I'm all for a tech tree so long as it moves KSP away from arcade and closer to simulation. Say mainsails aren't available immediately. That's fine. But if the way you get to them is by launching ridiculous asparagus-staged monsters that in no what resemble actual rockets (ie wider than tall) then the tech tree is not a good thing.

Harvester claimed that 160 parts was enough. It isn't. Take engines alone. Stock has only one engine in each weight class. So we can assume that the tech tree will be segmented according to thrust/weight class. That's not how technology works. The power of a motor is a function of the job, not the technology. A rocket motor today, and one 50+ years ago, will be in exactly the same thrust range but be radically different under the hood. Just look at the redesign work on the Saturn-5's main engines. The new models are almost exactly the same thrust as their predecessors. We need multiple models WITHIN each class before a tech tree makes sense. Maybe this means improving engine performance as one advances or maybe it means making them lighter (I'd add reliability, but that seems off the table for now).

Or take fuel tanks. We have only one in each size range with no dry weight separation (ie improved material science).

Edited August 8, 2013 by Sandworm

[Diche Bach]

Ah very good news!

In my experience there are two major stumbling blocks to tech trees. The first is that design choices are trivialized because you just use whatever you most recently unlocked because its better. The second is that the unlocks are trivialized because by the time you have aerospikes, you've already spent the last 10 hours of play escaping the atmosphere with whatever you did have available so the 'upgrade' doesn't actually help you.

Neither problem will kill the game for me, but I hope Squad manages to avoid the issues somehow.

Agree on this. The 'solution' for me in Civ was a mod that slowed the game down dramatically. Instead of only having 3 turns to make use of your nifty knew bronze spear points, you'd get 30 or 40.

[Necandi Brasil]

I'm still trying to figure out what Delta-V means! Tech tree???????

Delta-V

http://en.wikipedia.org/wiki/Delta-v_budget

Tech Tree

http://en.wikipedia.org/wiki/Technology_tree

[maro]

Very interesting alternate route for tech development and it would be cool to have test pilot missions where new engines are put through their paces and the results used to improve their reliability..

That's something else to wonder about and contemplate: reliability. Will career mode include realistic reliability stats for newly-unlocked parts?

Was also my intention, but was told that part failures will not be a "feature" of this game.

Therefore a new, much easyer idea to implement: What if SQUAD implements only the financial result of experience...?

Example: One can buy parts in numbers to reduce the price of a single piece AND one can buy a stack of parts to reduce the price of the whole stack. Therfore you can invent whole "stages" and as more they are used unchanged they cheaper they get....as a result of the experience on the assembly line.

Would work like this - if you take any parts out and put it together you pay 100% of its single peace price. As you can save the whole launcher you can save the stage or a probe, a lander, a satellite.... And for any stage-configuration that one uses again the price for the parts-set will drop.

Using a stage a second time will drop its price for 10%, a third time will drop by additional 5%, from 4-10th launch the drop each time is 2,5%, up to the 25th launch is 1,25% and there after ist 0,75% up to the 45th launch - then the price of the stage configuration has decreased allmost to 50% and will stay there.

Altering designs would also be possible this way.

Removing a part out of a stage will reduce the stage price only by 25% of the price of this part

Adding a part will increase the price of the stage by 10% + the parts price.

Example: After 10 launches one adds RCS thanks and thrusters to a stage - starting price was 10.000 at the first launch, by the 10th it was 7.161. Now the price increases again to 7.877 + the price of the RCS stuff.

This will result in a more real live behavior of the financial aspect of KSP

[Spyritdragon]

Delta-V

http://en.wikipedia.org/wiki/Delta-v_budget

Tech Tree

http://en.wikipedia.org/wiki/Technology_tree

Best use of that smiley i've seen yet.

But "If you run this engine at higher throttles in thinner atmospheres/vacuum could damage it."

If you run mainsails at higher throttles in Kerbin's atmosphere for longer durations it has a tendency to fail in a dramatic uncontrolled oxidation of itself.

Edited August 9, 2013 by Spyritdragon
"Tendence"? Seriously, me?

[Tommygun]

Was also my intention, but was told that part failures will not be a "feature" of this game.

Therefore a new, much easyer idea to implement: What if SQUAD implements only the financial result of experience...?

Example: One can buy parts in numbers to reduce the price of a single piece AND one can buy a stack of parts to reduce the price of the whole stack. Therfore you can invent whole "stages" and as more they are used unchanged they cheaper they get....as a result of the experience on the assembly line.

Would work like this - if you take any parts out and put it together you pay 100% of its single peace price. As you can save the whole launcher you can save the stage or a probe, a lander, a satellite.... And for any stage-configuration that one uses again the price for the parts-set will drop.

Using a stage a second time will drop its price for 10%, a third time will drop by additional 5%, from 4-10th launch the drop each time is 2,5%, up to the 25th launch is 1,25% and there after ist 0,75% up to the 45th launch - then the price of the stage configuration has decreased allmost to 50% and will stay there.

Altering designs would also be possible this way.

Removing a part out of a stage will reduce the stage price only by 25% of the price of this part

Adding a part will increase the price of the stage by 10% + the parts price.

Example: After 10 launches one adds RCS thanks and thrusters to a stage - starting price was 10.000 at the first launch, by the 10th it was 7.161. Now the price increases again to 7.877 + the price of the RCS stuff.

This will result in a more real live behavior of the financial aspect of KSP

That's an interesting idea.

You would also have to decide whether to stay with an older cheaper system or go with a newer more expensive model.

[lazarus1024]

Speaking of....I do like the idea that if you are able to return any parts to the KSC, you get some/all of the value back to be able to spend on future missions. It increases the incentive in career mode to actual get stuff back (assuming there isn't also a mechanism of needing to, to complete a mission).

I think it should be modified based on how far from the KSC you land. Just land somewhere on Kerbin, and maybe you only get 10% of the value back. Get within 500km of the KSC and maybe you get 25% of the value back. Land within 100km and maybe you get 50% of the value back. Land within 10km and you get 75% of the value back and land within 1km of the complex and you get 90% of the value back.

Just a thought on recovery and finances.

[Sandworm]

Nowhere in the history of space travel have 90% of the cost of building a vehicle been saved via recovery. There have only ever been TWO craft, shuttle and the shuttle's SRB, that have ever been "recycled" and they still cost millions and millions and months to turn around. (That virgin thing is less a spacecraft, more a rollercoaster for rich people). Worse yet, if you look at those programs as a whole they would have been far far cheaper had they never been designed for recycling, and most likely far safer.

Accurate savings? How about 10% money back... but 200% penalties when you drop SRBs anywhere over land.

[Necandi Brasil]

Nowhere in the history of space travel have 90% of the cost of building a vehicle been saved via recovery. There have only ever been TWO craft, shuttle and the shuttle's SRB, that have ever been "recycled" and they still cost millions and millions and months to turn around. (That virgin thing is less a spacecraft, more a rollercoaster for rich people). Worse yet, if you look at those programs as a whole they would have been far far cheaper had they never been designed for recycling, and most likely far safer.

Accurate savings? How about 10% money back... but 200% penalties when you drop SRBs anywhere over land.

If you design a space program that uses spent stages in orbit as fuel depots, If you have a 'Grasshopper' type vehicle (even if it's only possible to recover the first stage) I'm pretty sure costs will go down.

You see, the Shuttle was really intended for low cost space flight, the REAL problem with them was the lack of the 60 flights per year... Had the Challenger distaster not happened the U.S would reach the low cost goal...

Source: MIT course 16.885J avaible on youtube

[LameLefty]

Here's another real-world analog to a possible Tech Tree ... deep throttling for engines and variable ISP. For instance, most real rocket engines cannot be run at 5% or 10% throttle like you can in KSP. Actually most engines are not variably throttleable at all and if they can be throttled, they have specific set points (often only 1 set point). For instance, the newest American engine is the J-2X which is wrapping up testing currently. This hydrogen-oxygen ("hydrolox") engine has only 2 throttle points - 100% and 80%. The RS-25 (a/k/a "Space Shuttle Main Engine") is variably throttleable in 1% increments, but only between about 60% and 109% of rated thrust. So a real way to add a LOT of challenge for dedicated players would be to include "Engine Throttling" and "Deep Throttling" to the Tech Tree.

Similarly, when engines are run at lower throttles, they produce lower pressure and temperature and thus tend to lose efficiency. So the ISP tends to be several seconds (and potentially much more - several TENS of seconds) of ISP when throttled down deeply. So variable ISP would be an important thing to add to the game.

[LameLefty]

If you design a space program that uses spent stages in orbit as fuel depots, If you have a 'Grasshopper' type vehicle (even if it's only possible to recover the first stage) I'm pretty sure costs will go down.

You see, the Shuttle was really intended for low cost space flight, the REAL problem with them was the lack of the 60 flights per year... Had the Challenger distaster not happened the U.S would reach the low cost goal...

Source: MIT course 16.885J avaible on youtube

The Space Shuttle would NEVER have reached 60 flights per year. There were never more than 3 launch pads (two at KSC, one at Vandenberg Air Force Base). 20 flights per year per pad would have cost HUNDREDS OF BILLIONS of dollars. The best "realistic" flight rate actually planned for, with all three pads up and running, was about 24 flights per. And that was a lot, compared to the reality. A big reason Challenger happened in the first place was the "Go Fever" that accompanied a rushed flight rate and pad/ground ops pacing, which led to decisions being made based on data points being ignored or downplayed. Challenger STS-51L was hardly the first mission to have SRB field joint damage; it was just the one that actually led to catastrophe. Similarly, Columbia was lost because managers again felt that the understood and could deal with the system behaving in ways it wasn't designed to (e.g., shedding foam debris at very high velocity, causing impact damage to thermal protection systems). STS-107 was not the first one to have TPS damage, just the first one to have CATASTROPHIC TPS damage.

Source: Rogers Commission Report; CAIB Report.

[rkman]

Had the Challenger distaster not happened the U.S would reach the low cost goal...

Source: MIT course 16.885J avaible on youtube

How could the 25th shuttle flight affect the cost of the (all over budget) flights that went before?

[Necandi Brasil]

Well all I'm saying is that the shuttle WAS planned initially for 60 flights per year ...

After the Challenger disaster the Air force decided to go with expendable launchs... Lowering even more the use of the Vehicle.

" early on in the justification, in the selling of the Space Shuttle, and this primarily happened at the Washington level, Washington headquarters level, people created these economic models and flight rate models, mission models, they called them at the time, that had very, very high flight rates. The highest was sixty flights a year originally specified, and that gradually ramped down to today we're probably capable of ten or twelve if we had the activity to support it.

But they did the economics based on high flight rates so that all of the fixed costs would be a low part of the recurring cost per flight, and it was assumed that it would be easy to amortize fixed base costs like the populations at Johnson Center and at the Kennedy Center, and that the only extra costs would be the tanks and rockets. As a matter of fact, the first number I remember for what was supposed to be the cost per flight was about 10 million dollars per flight, 10 million dollars per flight. Of course, that was in '72 dollars, which, if you escalate it today, would be quite a bit more than that. "

src: http://www.jsc.nasa.gov/history/oral_histories/LunneyGS/LunneyGS_12-9-99.htm

And btw, I'm in favor of being able to recover parts for later use in KSP (After all this is a game) and designing spacecraft that can be recovered is both challeging and fun ...

Edited August 10, 2013 by Necandi Brasil
adding more information :)