Friday Facts #414 - Spoils of Agriculture

[Knuth]

Hi,

Watt per Second is the rate at which the Factory power consumption grows. XD

the factory grows exponentially, so it'd be watt per second per second

So it's parabolic...

Exponential would be joule (or watt) ^ sec
or even joule (or watt) ^ -sec

[DiegoPro77]

You could create the mod that just removes the spoil time, if you don't like it.

It's obvious that the spoil time is a balance choice. If you introduce the way to halt it, it makes the whole feature pointless. You just replace the chests with freezers and that's it.

Or just put this feature in a proper place in progression. Maybe in really really late stages of the game. :3
We believe in progress after all, don't we?

[picklock]

Interesting news again in this FFF. It's unbelievable what ideas you developers come up with. I can't wait to see how the whole thing plays.

[Rammix]

Biter farms maybe?
1) Biological products ("meat", methane, etc).
2) Electricity generation by biters running in a wheel.
3) Feed biters so that they grow and breed (to replace those dying of age).

[Lord Bumbleton]

Also, is the train logic going to prioritize trains with 'spoilable' cargo? If not by default, could we perhaps get a way to do it via circuits? (reading train cargo as it passes through a signal?)

This may be the interrupt schedule thingy that the devs said would be useful for one of the planets. I hope it is at least, because one of my trains spoiling on the ride is my worst nightmare.

[vipm23]

Everyone talks about freezing, refrigerator cars, etc...

...no one's said anything about the original solution to this problem:canning.

[Amarula]

...no one's said anything about the original solution to this problem:canning.

Pickles!

[pleegwat]

Everyone talks about freezing, refrigerator cars, etc...

...no one's said anything about the original solution to this problem:canning.

Canning (or rather the boiling afterwards which makes it actually preservable) changes the chemical makeup of the biological material. That's not (usually) a problem for food, along the science path it probably would be.

[vipm23]

Everyone talks about freezing, refrigerator cars, etc...

...no one's said anything about the original solution to this problem:canning.

Canning (or rather the boiling afterwards which makes it actually preservable) changes the chemical makeup of the biological material. That's not (usually) a problem for food, along the science path it probably would be.

Oh, probably. For industrial purposes, though, it should be mostly fine, particularly if we can synthesize any necessary enzymes in storable form.

On the gripping hand, you can put Gleba fruit into barrels you've just emptied the oil out of.

[rootgear]

I think spoilables should be substantially delayable but only after manufacture, not through refrigeration but rather through preservatives added during the production line to things that are deemed worthy/neccesary of such which itself is also gathered/processed from some other plant. This preserves the mechanic affecting the layout design of factories/transport networks by neccesitating things to be used on site and only being attainable at the moment of production but reduces the annoyance of working around the objects freshness while shipping them to other planets, or inversely being incentivized to make gleba your new base to avoid this. Moreover it's meaningfulness is not reduced so long as the final result is still weighted according to the freshness of it's constituents. To facilitate this the preservative would primarily be inserted at the biochamber.

Also forgot to add... it looks strange that you have to destroy the tree alltogether in order to gather the fruits. If you go that route you should not use trees but something else, like mushrooms or whatever.

Suggestion: gather the fruit only, leave the tree intact. Instead of "free product growth", you have to fertilize it with some byproduct to speed up fruit production to usable levels (as natural speed would be way too slow... or slow only if there are too many trees in the same area)
Also interesting: When trees have to grow first before reaching full "fruit production capacity", you can't just exponentially ramp up production in an instant. It takes time to grow...organically. It would fit the theme of the planet

Also seconding this; odd that plant life would forego being perennial even on an alien planet, and moreso that any static instrument which harvests it's surrounding would work on something as low-density/fast-growing as an annual which due to such has heavy incentives for high amounts of space usage and constant rotation, due to these points if plants are all effectively annual it would make more sense to have huge fields that are tended to by logistic or new agro-bots.. not that I'm seriously suggesting such.

Anyhow, aesthetics/lore aside the above suggested focus on the importance of fertilizer as the main driving force of production as opposed to seed emphasizes the focus of time on gleba and adds a level of modularity and complexity to the supply chain. If spoils produce compost, and everything spoils, then anything can serve as fertilizer. So lots of options to consider when deciding how your main orchards are "refueled" from one another which overlaps with the managing of spoils; Since, being agricultural, resource pools are presumably made by the user alongside the factories, this allows much greater thought to be had towards the layout of each crop relative to another, which I think is cool. Alternatively the slash and burn method also proposed in that post would have a substantial increase of enemy-aggravation and risk of unwanted spread (presumably the planet is more heavily forested) if improperly managed but would reduce the immediate need for a supply chain and aforementioned layout considerations for fertilizing and so has tradeoff, perhaps incentivizing use of such early on should you posess sufficient munitions until you eventually establish your infrastructure.

[KuuLightwing]

I skimmed through the thread, and it seems like in 20 pages nobody mentioned this point.
I think the mechanic itself could be interesting (aside from the ugly white bars all over the place, same issue as with quality visual representation), but what I am concerned is that I don't think the idea of limited lifetime works well with existing factorio mechanics, unless some new mechanics are introduced.

Factorio logistics essentially is based on backpressure and buffers. Your simplest transport belt setup is producer -> belt -> consumer. If consumer stops consuming, the belt starting to back up, and when it's full and the output buffer of the producer is full, the process stops. Any belt network is just a more complicated version of that. Train stations rely on buffers to ensure the steady flow of items, and trains themselves are essentially buffers too. Same goes for the bots which operate via moving products between chests. So, I'm not sure how does one handle the challenge of "no buffers" and such when the whole logistics of the game is designed around buffers and backpressure. If the consumption is not constant, which is a case for honestly most things, including even science (if you haven't built the next stage, and machines are idling), then most of the products will spoil where they are, and will need to be purged.

There's some "on demand production" solutions that could be made with circuit networks, I suppose, but those are rather complicated, and IIRC rather slow - I think I only seen them as proof-of-concept. I would prefer to not have to resort to complex circuit networks for basic production chains like that, so something simpler is required. So at the moment I quite frankly don't even see how would I deal with this mechanics given the tools that I have available.

[robot256]

I skimmed through the thread, and it seems like in 20 pages nobody mentioned this point.
I think the mechanic itself could be interesting (aside from the ugly white bars all over the place, same issue as with quality visual representation), but what I am concerned is that I don't think the idea of limited lifetime works well with existing factorio mechanics, unless some new mechanics are introduced.



There's some "on demand production" solutions that could be made with circuit networks, I suppose, but those are rather complicated, and IIRC rather slow - I think I only seen them as proof-of-concept. I would prefer to not have to resort to complex circuit networks for basic production chains like that, so something simpler is required. So at the moment I quite frankly don't even see how would I deal with this mechanics given the tools that I have available.

A very valid concern. Luckily, when 2.0 releases you won't have to do it with just the tools you have now. I know Wube has said that combinators should be optional for the base game, but I don't know if the same is true for Space Age.

I imagine Wube has solved these problems in many different ways during their playtests, and the new 2.0 combinators almost certainly have something to do with it. They appear intended to simplify many of the operations that were nearly impossible with the old combinators.

[Stringweasel]

what I am concerned is that I don't think the idea of limited lifetime works well with existing factorio mechanics

Be careful not to confuse "existing factorio mechanics" with "current playstyle". And speaking with experience the spoilage mechanic can be solved with existing tools. You just need to do a bit of a mindshift: think what problem was the backpressure-meta solving, and how does spoilage change the problem?

Ultimately though, new mechanics requires new playstyles, and new mechanics is the idea of an expansion pack. And if you don't like the mechanics you can always play vanilla or one of the many epic mods

[KuuLightwing]

Be careful not to confuse "existing factorio mechanics" with "current playstyle". And speaking with experience the spoilage mechanic can be solved with existing tools. You just need to do a bit of a mindshift: think what problem was the backpressure-meta solving, and how does spoilage change the problem?

Ultimately though, new mechanics requires new playstyles, and new mechanics is the idea of an expansion pack. And if you don't like the mechanics you can always play vanilla or one of the many epic mods

I am talking about mechanics, I'm not confusing them with anything. The fact that machines work based on the state of the output buffer is not a "playstyle", it's a mechanic and arguably a core one. The fact that items on belts are designed to stop when they reached the end of the belt is also a mechanic, and so on. Other two options to halt a machine is to either starve it of resources which probably would require circuit conditions, and (IIRC) direct circuit conditions.

Any solution that doesn't use backpressure would rely on circuit network. Which might be fine, but needs to be seen, and definitely not all players are adept at using them. Also introduces the issue with the latency, because without saturated belts, you will have to wait until items placed on a belt reach their destination, which would drastically reduce overall throughput.

Other option is to never stop the machines at all, and just void anything that didn't get used by the machines. Potential issue is that due to the inserter cycles, some items would be skipped, so you'd have to overproduce them - another example how solutions like this aren't working nice with the mechanics of the game.

I don't doubt there's still ways to deal with it, but what I am saying is that existing logistics mechanics don't play into that idea, as they weren't made for cargo with limited lifetime. Even optimizations for belts are made with an assumption that belt is fully compressed and each lane is filled with the same item.

Regarding the last remark. I don't know what's the purpose of it, as it's just the "don't like it don' play it" clause. I may be interested in some parts of the expansion, but not all of them.

[hovestar]

Be careful not to confuse "existing factorio mechanics" with "current playstyle". And speaking with experience the spoilage mechanic can be solved with existing tools. You just need to do a bit of a mindshift: think what problem was the backpressure-meta solving, and how does spoilage change the problem?

Ultimately though, new mechanics requires new playstyles, and new mechanics is the idea of an expansion pack. And if you don't like the mechanics you can always play vanilla or one of the many epic mods

I am talking about mechanics, I'm not confusing them with anything. The fact that machines work based on the state of the output buffer is not a "playstyle", it's a mechanic and arguably a core one. The fact that items on belts are designed to stop when they reached the end of the belt is also a mechanic, and so on. Other two options to halt a machine is to either starve it of resources which probably would require circuit conditions, and (IIRC) direct circuit conditions.

I think the point is that this is a totally new mechanic meant to change the meta. We've always filled our buffers, but now spoilage gives us a fundamentally different problem to solve. The fact that we view buffering as fundamental is the exciting part of this change. On Gelba we'll need to design our factories different, it's not the same game there.
If you've played the space exploration mod, or even watched a let's play, you quickly realize that a common complaint is that each planet is the same, just with extra recipes. Spoilage makes us use the existing tools we have in a totally new way.

[Stringweasel]

I am talking about mechanics, I'm not confusing them with anything. The fact that machines work based on the state of the output buffer is not a "playstyle", it's a mechanic and arguably a core one. The fact that items on belts are designed to stop when they reached the end of the belt is also a mechanic, and so on. Other two options to halt a machine is to either starve it of resources which probably would require circuit conditions, and (IIRC) direct circuit conditions.

Sure, what is or isn't intended game-mechanics subjective to us as players. However, these things you mention though sounds to me like the constaints put in place by the game, and you need to solve a way around it. And it is typically done using back-pressure - meaning letting your belts back up. I do see this as a play-style though. But again, I do agree with you that typically doing it any other way will require a little bit of circuits.

I think you understand the problem of handling spoilage. Typically you could use circuit networks to minimize losses, or just let the machines keep running forever and deal with the losses. Note: you will probably always have losses. However, I don't really understand your concern against the latter approach of letting the machines just run. Over-producing is usually recommended in the current Factorio meta? Pretty sure the recipes are balanced to take losses into account, so it's not so much having to build a bigger factory, but instead can be seen as a type of by-product that the recipe produces.

Minimizing losses do require some circuitry though, but it's not more complicated than the current challenge of balancing your heavy-oil/light-oil/petrol. And also not more complicated than what you would need in other parts of Space Age, like the space platforms. And they've also made using circuits a lot easier. For example by revamping the Decider Combinator and allowing you to read the internals of your machines. I understand circuits can be scary to new players though, and many players struggle to fully understand it.

Even optimizations for belts are made with an assumption that belt is fully compressed and each lane is filled with the same item.

As far as I know this is a common misconception about belt performance, but I can't remember where to find proof of this

Regarding the last remark. I don't know what's the purpose of it, as it's just the "don't like it don' play it" clause. I may be interested in some parts of the expansion, but not all of them.

Yeah, this remark came across more ill-intended than I meant to. My point was more to try and keep an open mind for new features. We want new things for expansion packs, but it's subjective how much different they should be. And before judging the feature as bad, maybe wait until you've tested it out first.

[KuuLightwing]

Sure, what is or isn't intended game-mechanics subjective to us as players. However, these things you mention though sounds to me like the constaints put in place by the game, and you need to solve a way around it. And it is typically done using back-pressure - meaning letting your belts back up. I do see this as a play-style though. But again, I do agree with you that typically doing it any other way will require a little bit of circuits.

I think you understand the problem of handling spoilage. Typically you could use circuit networks to minimize losses, or just let the machines keep running forever and deal with the losses. Note: you will probably always have losses. However, I don't really understand your concern against the latter approach of letting the machines just run. Over-producing is usually recommended in the current Factorio meta? Pretty sure the recipes are balanced to take losses into account, so it's not so much having to build a bigger factory, but instead can be seen as a type of by-product that the recipe produces.

It's less the problem with overproducing, but more of a problem with inserters not playing well with a system like this. In a regular production chain, if you get your ratios roughly correct (producing slightly more than the machines down the line would consume), then you can usually be sure than all machines would be fed and happy cause belt will end up saturated and inserters would grab whatever is needed. But if you have your belt constantly flowing, as you void the excess, due to the inserter cycles it's possible that machines won't be saturated because it just simply missed some of the items on the belt, despite you technically producing enough. I guess it maybe not that big of a deal, but it's something that would definitely annoy me, and also something that seems to be hard to predict and plan for. Similar issue currently could happen with Kovarex process, but it could be resolved by splitting off some U235 via priority splitter thus "capturing" it and not letting it pass. This can't be done with spoilage because it would make items spoil on the waiting bay.

Minimizing losses do require some circuitry though, but it's not more complicated than the current challenge of balancing your heavy-oil/light-oil/petrol. And also not more complicated than what you would need in other parts of Space Age, like the space platforms. And they've also made using circuits a lot easier. For example by revamping the Decider Combinator and allowing you to read the internals of your machines. I understand circuits can be scary to new players though, and many players struggle to fully understand it.

I have some experience with circuitry, but need to be in a mood to work with those, so I often resort to mechanical solutions where I can. I don't count things like cracking, cause that's just a couple of conditions that don't even need combinators. I was envisioning a more "on demand" system, but I think that's going to be just way too complicated and slow.

As far as I know this is a common misconception about belt performance, but I can't remember where to find proof of this

Is it? I remember the old belt optimization FFF and how they described it by essentially storing items on belts as clusters and gaps between the clusters. So different items would each be stored as a separate entity, thus increasing the memory and computation times required. That's why big sushi belts are so horrible for the performance.

Yeah, this remark came across more ill-intended than I meant to. My point was more to try and keep an open mind for new features. We want new things for expansion packs, but it's subjective how much different they should be. And before judging the feature as bad, maybe wait until you've tested it out first.

I am not judging the feature as bad, but merely pondering about how it works with the rest of the game. I am skeptical, yes, but I recognize that it could be interesting to work with depending on how it actually plays.