How to handle the "over time" problem (And anyone know if squad is looking at this?)

[rodion_herrera]

Launches will lead to more launches, and more money.

But one would only think that there's "more money" if every item on the equipment list all have the same price. I was thinking perhaps Squad would have different prices for different kinds of say, the SAME class of fuel tank, but it gets more expensive if it's more robust and/or has more oomph (in the case of engines). Thus, if you keep going for the more pricey but reliable stuff, then I feel you'd get slowed down in your efforts considering you spend more to get the better item, and thus find yourself constantly bankrupt and just awaiting return of more science to get more money. This in turn, I believe, would bog things down to a point that all the imagined "over time" issues wouldn't be that much of an issue, considering that the pace at which things move are slow. So what if the guy keeps buying the cheap stuff? Progress will still be slow because he'd have greater chances at failure (cheapo stuff breaks more).

[Themohawkninja]

Before I throw my two cents out there, by "over time", you are referring to the fact that after (2^32)-1 seconds, the game crashes because the variable that holds the game time has run out, right?

[Spica]

Oh no. My favorite save is reaching that limit, at 64 years, the limit being 68 years, 35 days, 4 hours.

[Kegereneku]

I'm not sure that money alone will solve the issue, especially if successful missions result in financial gain. Launches will lead to more launches, and more money.

[...QUOTE...]

Agreed about the monthly budget. On it's own it's not a very elegant solution. I'm starting to think that there will have to be a very fine balance between the speed at which science is gathered, and money is accumulated.

[.....................]

Players will have more concurrent missions, which do necessitate some time-warping, but I feel that this is more acceptable than time warping with no active missions just to reach some future date (be it a monthly budget replenishment or a transfer window).

I must insist, a fixed-budget replenished each month could be a very elegant solution.

What we want to avoid is to grind for money or to have so much money it make budget meaningless.

The fixed-budget ideas take care of maintaining a minimal budget without keeping mission from increasing it.

The budget can be increased from time to time to match the increase in difficulty and parts cost.

More importantly it allow you to keep trying without being penalized for failure (as even reverting to VAB can be felt as a penalty).

Furthermore, it give a simple way for the player to trade "time" for any launch he deem necessary (be it a refueling station or a tug-ship), something that may not be possible if your budget only depended of mission or science return.

Lastly, a Reputation system used in conjunction with missions may be more than enough to avoid potential abuse.

It's simple, consistent and maybe balanced. So if this isn't at least little "elegant" I don't know what could be.

Before I throw my two cents out there, by "over time", you are referring to the fact that after (2^32)-1 seconds, the game crashes because the variable that holds the game time has run out, right?

I didn't knew about that but sort of.

The discussion have mostly been about using time as a direct gameplay mechanism. A little like production time in RTS-game but adapted so Timwarping actually have a visible & balanced impact on the game.

Knowing the game as it is now have a time-limit is indeed very important.

Edited December 19, 2013 by Kegereneku

[iMAniaC]

That's a really excellent post, I was nodding my head most of the way through.

Your comparison with minecraft is interesting, indeed I have a different topic about objective/reward game types. The bit that I find interesting is this: do you feel like there is a dfference between the castle you have finished building in miecraft, vs the base that you have, say, landed on Duna? It seems to me that once you've landed the base on Duna, it's basically just "art" that you've added to the lifeless KSP universe. But in minecraft you have actually affected the universe itself (and indeed, that castle might serve a purpose - I've never played minecraft but as I understand it has things like zombies, "electrical" systems you can create to make your castle "active", and so on).

Back to the time mechanic: Does this just come down to a distinction between career mode and sandbox mode? In sandbox we can do what we like whenever we like, that's basically the whole idea. But career mode seems, by definition, to be a mode that would require some kind of tangible connection between one mission and the next (in the sense that they are all working towards whatever the career mode objectives are), which also suggests, at least to me, that all the missions operate within the same "universe", and therefore the same time... at least eventually. Right now that's not the case, the "science points" don't care how many missions you flew, how long they took to fly, whether they were "profitable" (because that's not a thing in the game yet), and so on. But I expect that as career mode gets tied together a little more, the "common time" mechanic would make more sense.

Sorry about the late answer, I was, uh, busy (you can guess with what ). Anyway, to answer your direct question, I have to admit that I haven't reached the point where I have bases on other planets yet (I started playing not long ago), so I wouldn't be able to say for sure. However, based on what I have achieved, I think that there won't be much of a difference between my Minecraft castle and KSP Duna base (when I finally build one). In Minecraft, I usually get a certain satisfaction simply from having achieved my objective and I expect it would be the same with KSP bases. I'll probably move some Kerbal around the finished base, look at it from various angles and feel good about myself and then move on to bigger and better things. If this is not how things play out, though, I'll let you know

The last few days I have been playing with the idea of bypassing the idea of funds per time period. After all, it may be a very human thing to think along the lines of fiscal years or quarterly results (sometimes to the point of absurdity, but that's another discussion altogether). Instead of Kerbals siphoning off x money every year for something, they might have a mentality more like "Going to Mun? That's easily worth this much. Here, take my money!" That way, time is no longer an issue. You could do all the missions simultaneously or you could do them serially and Kerbalkind wouldn't care, because they focus on the achievement and ignore the time it took to get there.

Initially, I thought about a set budget for every contract (and so the challenge would be to make a rocket within budget), but I soon realized that only having those would kill the opportunity to do missions the game developers had not thought of. Also, if it were all about the science, doing missions that do not return science would be wasteful in Career, so a space station around Kerbin is pretty much pointless if you've soaked up all the science from Kerbin's orbit already. But what if there were free missions in addition to contracts, based on a general game mechanic that calculated worth based on payload sizes and types of various situations? So for instance, getting a payload of x tonnes into orbit is worth x money to the Kerbals. The money gained is then scaled with certain factors; distance from Kerbin or Kerbin's orbit, whether the payload lands on another celestial body, whether there are Kerbals aboard or if it's only a probe, etc. In a sense, it would in some ways mirror the mechanic for determining science, but while science is used for unlocking the tech tree, getting payloads into space has an intrinsic worth. Note, however, that these kind of missions are based on returns and not on budget: It's the money gained from your previous mission that is invested into the next, rather than getting a specific budget before the mission is conducted.

Such a system would also allow for an infinite number of missions that do not return science: For example, putting a satellite into orbit is worth a certain amount to Kerbalkind. Thus, if you manage to do it within the worth of the mission, you will get the money back, as it were, enabling you to build a whole network of satellites if you wish without going bankrupt.

Speaking of going bankrupt, that is a very real possibility if a mission fails, since there may not be any return on the investment, as it were. So in the beginning, to make it more newbie-friendly, there could be a specific contract: "Getting to orbit" that would refresh no matter how many times you fail, giving you a specific budget again and again until you reach orbit. Likewise, if you go bankrupt after a long and successful Career (and you've already done all the Contracts), there could be a contract "Back to space" that would give you a certain amount to restart your space program.

In summary, this is what I'm suggesting:

Resources:

-Science: Used for unlocking the tech tree.

-Savings: Money hoarded up by the space program, used to build rockets for Free Missions

-Budget: Money available for a specific Contract, only.

-Reputation: ??? Maybe something to with astronauts.

Mission types:

-Contracts: Missions with specific parameters. Money paid up front as a budget.

-Free Mission: Rockets built with savings. Successful launches return money to the savings based on certain factors.

In the end, this post doesn't touch much on the topic of "Over time"/"meanwhile" mechanics, but it's meant as an elaborate way to avoid it altogether, in order to not force the players to do missions they don't feel like. Moreover, with contracts with up front money, the player doesn't have to wait for the return of his investment into a Free Mission and it will still be possible to do several missions at the same time and thus get closer to a continuous space program, if that's what the player prefers.

I'm sure there are lots of things that have slipped my mind, so feel free to point them out. And congratulations if you made it all the way through

[vexx32]

Before I throw my two cents out there, by "over time", you are referring to the fact that after (2^32)-1 seconds, the game crashes because the variable that holds the game time has run out, right?

If you've a game with the universal time value being anywhere near that, you'll see a myriad of problems well before then.

But that's not what the topic is about, anyway.

I don't think that the timewarp v.s. time-based systems should become an issue. All you really need to do is make sure that the things you need to do over time are sufficiently non-tedious that you don't need time warp for them, and then simply put them in a place where it would be nonsensical to use them for time warp. For example, nobody uses time warp when collecting EVA samples. That's just silly. (Apologies if you do, and if you do, does it work at all?)

(Naturally the actual implementation will be far removed, that's just a quick example I dredged up at midnight).

[Themohawkninja]

If that is the issue being discussed here, why not just make each unit of time (seconds, minutes, hours, days, and years) its own 32 bit variable instead of having one variable for seconds and subdividing it to get all other units?

I doubt anyone will run into problems after 65,000+ years.

[AngelLestat]

I must insist, a fixed-budget replenished each month could be a very elegant solution.

This will remove the freedom of what mission you choose and it push you to be multitasking.

Pros: If you make all time dependent then all start to had sense. You can not abuse using time warp with the different mechanics. A great benefic with mining resources and science that it will take time.

Cons: If you wait for best launch windows, then you wait 1 extra year at the planet and maybe you fly over there, The whole mission takes you 3 years lets said. Then you find that you have a big budget or the fact that your reputation is ruin becouse you dint do nothing in those 3 years. So this force you to be multitasking, launch a mission, then plain the next mission meanwhile the first mission is not finish.

What if we dont want to be multitasking? I try some times make many mission at the same time and I dint enjoy it much.

I like to focus in one thing at the time. Searching the best way to improve that mission (this mean go back to the VAH many times if I need to and launch again), if in the middle of those attempts I also need to focus in many other missions doing some actions over time.

It will become tedious.

I really wish that there was a mechanic to not force you into multitasking. Or maybe there is...

so a space station around Kerbin is pretty much pointless if you've soaked up all the science from Kerbin's orbit already.

Well this can be fix it with a bit of reality.

The space station it does not make science by its own. It needs astronauts.

And each astronauts goes to the space station to test a particular theory or experiment.

In that case you will find that you dont get any benefic leaving kerbanauts up there once they already finish their experiments.

So you need to exchange with new kerbanauts if you wanna keep doing science.

This give a new use to the "hire kerbanaut" panel. You will hire some sicientist kerbanauts just to make science. Each one trying to prove their own theorys.

For now like you said, I keep with the idea of money based in goals with some kind of loan budget mechanic (in case you want to make a little jump in your ambitions).

Edited December 19, 2013 by AngelLestat

[Themohawkninja]

Well this can be fix it with a bit of reality.

The space station it does not make science by its own. It needs astronauts.

And each astronauts goes to the space station to test a particular theory or experiment.

In that case you will find that you dont get any benefic leaving kerbanauts up there once they already finish their experiments.

So you need to exchange with new kerbanauts if you wanna keep doing science.

This give a new use to the "hire kerbanaut" panel. You will hire some sicientist kerbanauts just to make science. Each one trying to prove their own theorys.

For now like you said, I keep with the idea of money based in goals with some kind of loan budget mechanic (in case you want to make a little jump in your ambitions).

I would like to see this in the game, or at least a mod.

[Kegereneku]

Not going to derail the topic (again) AngelLestat, but I totally disagree with you. If you compare with the usual alternatives and don't use such a severe reputation-loss to make your points you should find out that it globally preserve freedom while allowing multitasking.

Anyway, since the "over time" problem is linked to science and possibly budget and the Devs are working more on Career-mode, it should get addressed soon.

If that is the issue being discussed here, why not just make each unit of time (seconds, minutes, hours, days, and years) its own 32 bit variable instead of having one variable for seconds and subdividing it to get all other units?

I doubt anyone will run into problems after 65,000+ years.

No that really not the only issue discussed here.

The problem is that for now time is ridiculously cheap. Even science happen instantaneously which is kind of immersion breaking. I would be interesting to know what the devs have in mind.

[nholzric]

I think you can solve the problem with how the three resources (Science, Money, and Reputation) are earned. Science, as we have seen, is earned by what you do on your trip and where you go. Money is earned by accepting contracts to go on various trips (take this Kerbal to the Mun and back, take this one to Duna orbit and back). And Reputation is earned by how quickly to accomplish it depending on the mission (if you take 500+ days to get to Duna and back your reputation returns are nill). This along with the alarm clock idea mentioned earlier allow for multiple missions at a time of vastly different durations while penalizing spamming the time-warp button.

(Speaking of ruining immersion, I just watched this Scott Manley video http://youtu.be/HIjqvLcsz8g in which he nearly completes the tech tree in two flights... but oh yeah, the second flight takes over 2000 days. I really enjoy Scott's videos, not trying to pick a fight, but in my mind that's a big part of KSPs "Time Problem" when you talk about anything other than Sandbox)

[allmhuran]

Two flights is fairly straightforward. Here's 5528 on mission 1.

[nholzric]

Two flights is fairly straightforward. Here's 5528 on mission 1.

Nice! So, that took over 7800 days, how long did it take you to play?

That ship undoubtedly took a lot of skill to build and fly, and the mission took a cool head to execute, I'd need several tries and some luck if I was ever going to pull it off. But isn't this also a good example of KSP's Time Problem? More power to you if you can pull of these epic missions in sandbox, but it seems to me that this type of thousand days trip circumvents enough features to be an exploit. In this case you're not exploiting the transmit button, you're exploiting timewarp and the fact that Kerbals don't need anything to survive.

[allmhuran]

I flew it over about 3 nights IIRC.

And yes, it is absolutely a good example of the time problem. In fact I used it as an example earlier in the thread:

In total, the mission took something like 20 in game years. But surely that doesn't make sense in terms of the realities of running a space program. I essentially exploited the fact that nothing in the game itself "cares" how long it takes you to do something. When I eventually recovered my craft and went to build my next mission the fact that the first one had taken so long had no material impact.

[allmhuran]

I'm sure there are lots of things that have slipped my mind, so feel free to point them out. And congratulations if you made it all the way through

I did!

There are some good thoughts in there. I think you could add to that the possibility of missions that may not directly return "science points" themselves, but may assist other missions. (And I don't necessarily mean scripted/assigned missions, just player created missions). For example, I believe squad is looking at communications relays for transmission - which, by the way, I think is an excellent idea - and so getting a few satellites up around Kerbin to act as relays doesn't give you any direct science points, but will assist with later transmissions you make from wherever you may be in the solar system. And this is exactly why I think it is an excellent idea - it gives the player a real objective that is integrated into the gameplay, not a "gimmicky" addon (like science points).

[AngelLestat]

Sorry, I had to resurrect this topic, becouse I am feeling all mighty.

Not going to derail the topic (again) AngelLestat, but I totally disagree with you. If you compare with the usual alternatives and don't use such a severe reputation-loss to make your points you should find out that it globally preserve freedom while allowing multitasking.

Anyway, since the "over time" problem is linked to science and possibly budget and the Devs are working more on Career-mode, it should get addressed soon.

.

Lol, what are you talking about Kege? Derail the topic? I guess my question was very inside of this topic.

But well, if you dont wanna talk, I will understand.. snif snif.

.

.

.

.

.

.

I am still here... lets said that "it can work", can you give more details how you will balance this? with examples.

[Alchemist]

The main question is about tracking unloaded parts and vessels. Running "onUpdate" every frame for everything in the space is not a good option, but some good design can provide reliable alternative to that. And there already are some such solutions. It started with SatelliteRelayNetwork/RemoteTech, but for something more advanced... guys in Russian subforom are working on a life support mod with full resource tracking on unloaded ships.

[thereaverofdarkness]

In many cases, there are ways to make the "over time" element really simple. Similar to the "on rails" mechanic of orbits during time warp, many functions that happen over time can be mapped out in a procedural schematic, and the results can be computed easily simply by plugging in relevant details such as the starting conditions and time, the ending time, and anything that profoundly affected it during that time. A mapping satellite need not give data of all parts of the ground that it mapped, but rather the game can simply note its orbit and when it started and stopped mapping, any orbital changes from passing spheres of influence and when, and a map of power levels during the time (which itself boils down to simple effects such as rate of depletion and times entering and exiting sunlight). and thereby know what was mapped by the satellite. Scripting it like this allows these functions to happen easily at high time warps.

[Kegereneku]

Sorry, I had to resurrect this topic, becouse I am feeling all mighty.

Lol, what are you talking about Kege? Derail the topic? I guess my question was very inside of this topic.

But well, if you dont wanna talk, I will understand.. snif snif.

I am still here... lets said that "it can work", can you give more details how you will balance this? with examples.

Don't mind, I don't want to launch a debate about "budget" here and I just remarked that you deliberately choose to use a very-severe "reputation loss" (for not doing anything) in your argument against my suggestion, and your argument depended heavily on it with no though given to alternatives (ex : not loosing reputation for doing nothing but warp).

None of us know if it can work or cannot work, but I have found no fundamental problem with it (like say a recursive loop giving infinite money).

But again, this isn't the place to make an actual suggestion about it.

The main question is about tracking unloaded parts and vessels. Running "onUpdate" every frame for everything in the space is not a good option, but some good design can provide reliable alternative to that. And there already are some such solutions. It started with SatelliteRelayNetwork/RemoteTech, but for something more advanced... guys in Russian subforom are working on a life support mod with full resource tracking on unloaded ships.

It's exactly this.