Disconnect roboports from logistic network

[ssilk]

There could be an option on the roboport screen to create/select logistic networks. The difference would be that the robots inside such network don't search for items on the other network.

That is exactly what is meant by "color" above.

However, the only situation where I thought this would be useful, was when I was trying to turrent-creep to biters bases close to my base. So it's not really something I would use. Besides I found it's easier to just shotgun their nests from the car and then kite them to my base for the turrets to kill the biters left.

Well, the usage is in my eyes only, when your base factory gets bigger than about 500 tiles in diameter. That happens not in all games, but especially, when you play with low resources.

Let me explain the logical steps, which brings you to this idea:
Assumed the base factory is in the center of your logistic network. Then expanding means, that a bot must drive in the worst case from the border to the center, pick up something and back to the border to place that.
This worst time should be low, I would say the worst time is about one minute.

So calculation is:
Form the border to the center and back = 500 tiles
Speed is 3 tiles/sec * speed bonus. Let's say at this time you have a speed bonus of 3-4, so the speed is between 6 and 7.5 (rounded).
So time is in the best case about 66 seconds and in the worst about 83 seconds.

Edit: I looked up the wrong bot, logistic basic speed is 3, but construction bot is 3.6 tiles/sec! Anyway the calculation with about 1 minute in max shouldn't change that much.

Much longer makes no sense. You cannot built things fluidly, if it takes longer.

So, logically, if you go beyond that size you need to change the strategy. One is to have logistic networks every 800-1000 tiles.

The logical consequence is: how could the item then be interchanged between the networks? There are the three options belts, train or logistic network. But this last is currently extremely complicated and not useable without the usage of the two others. So the next logical idea is to overlap the networks, but how to distinct them now?

Well, and so the idea of colored logistic networks (or areas) comes up as logical consequence. They have already colors, we just need to add some more.

[SHiRKiT]

There could be an option on the roboport screen to create/select logistic networks. The difference would be that the robots inside such network don't search for items on the other network.

That is exactly what is meant by "color" above.

However, the only situation where I thought this would be useful, was when I was trying to turrent-creep to biters bases close to my base. So it's not really something I would use. Besides I found it's easier to just shotgun their nests from the car and then kite them to my base for the turrets to kill the biters left.

Well, the usage is in my eyes only, when your base factory gets bigger than about 500 tiles in diameter. That happens not in all games, but especially, when you play with low resources.

Let me explain the logical steps, which brings you to this idea:
Assumed the base factory is in the center of your logistic network. Then expanding means, that a bot must drive in the worst case from the border to the center, pick up something and back to the border to place that.
This worst time should be low, I would say the worst time is about one minute.

So calculation is:
Form the border to the center and back = 500 tiles
Speed is 3 tiles/sec * speed bonus. Let's say at this time you have a speed bonus of 3-4, so the speed is between 6 and 7.5 (rounded).
So time is in the best case about 66 seconds and in the worst about 83 seconds.

Edit: I looked up the wrong bot, logistic basic speed is 3, but construction bot is 3.6 tiles/sec! Anyway the calculation with about 1 minute in max shouldn't change that much.

Much longer makes no sense. You cannot built things fluidly, if it takes longer.

So, logically, if you go beyond that size you need to change the strategy. One is to have logistic networks every 800-1000 tiles.

The logical consequence is: how could the item then be interchanged between the networks? There are the three options belts, train or logistic network. But this last is currently extremely complicated and not useable without the usage of the two others. So the next logical idea is to overlap the networks, but how to distinct them now?

Well, and so the idea of colored logistic networks (or areas) comes up as logical consequence. They have already colors, we just need to add some more.

Couldn't be clearer. Perfect explanation. Doing train setups in my own base seems overkill to the extreme.

[Khyron]

So, essentially when your logistics network has any dimension around 500 tiles you start to experience a problem with drones not delivering things in a timely manner. And as it scales out further it becomes completely unworkable at some point. I'm sure we all admit this can be resolved by segmenting and hybridising the network (clusters of roboports linked by belts or rail), but you guys would rather have a more advanced drone network.

At this stage I'm against this idea simply because such a feature request is well beyond what challenges the game currently presents. The game is eminently solvable with a much, much smaller and simpler base, regardless of resource settings (building resource outposts is unrelated). Obviously you guys are looking for ways to get more out of the game by role-playing and setting mega-project goals and I can respect that. I like to needlessly optimise my layout for energy consumption and speed. I also acknowledge the value in dreaming big.

This feature request just seems too far ahead of the game development AND you already have some capacity to deal with the problem. Rail for long range; belts for medium range; drones for short range.

[ssilk]

The game is eminently solvable with a much, much smaller and simpler base, regardless of resource settings (building resource outposts is unrelated).

Repeat: This is true for normal settings, but with low resources it is not.

This feature request just seems too far ahead of the game development AND you already have some capacity to deal with the problem. Rail for long range; belts for medium range; drones for short range.

Play with different settings.

[syneris]

The game is eminently solvable with a much, much smaller and simpler base, regardless of resource settings (building resource outposts is unrelated).

Repeat: This is true for normal settings, but with low resources it is not.

Or when your goal isn't to 'beat' the game, but to make cool factories and designs. This is why there is a lot of discussion on end-game content even though the game is in alpha. Endless play is fairly popular among the community here.

[Khyron]

Repeat: This is true for normal settings, but with low resources it is not.

Play with different settings.

My first couple of tries with the game were default res. By third game I finished it with default res. Since then I've only played low resource games. First one was all low res, low frequency and small deposits but i regened the map until i got a good chunk of iron and copper in the starting area. Coal and Oil were huge problems and I couldn't get access to a 3rd oil well or any coal until i killed a nest with 6 spawners. I had to use trees cut manually for fuel for everything until solar, but even then i couldn't make advanced chips for electric furnaces because I had no coal... It was a great game but in the end it was progressing too slowly due to oil. Since then I've kept oil field size on medium. I've played two other low resources games since then and the current game that I've just started has a descent iron supply but hardly any copper in sight, so I'll have to delay switching to solar until much later which should be really interesting.

I tend to build isolated bases connected by rail (due to low frequency settings). The rail passes through hostile territory. My guess is that we play low resource games quite differently. Do you wall off sections of the map? I don't. My resource bases tend to have one roboport and my main base has 2 or 3 covering the factory and all the solar fields and rail junctions.

So, when I say resource settings don't make a difference, that's where I'm coming from. It's completely fine if your play style is different and that's a testament to the flexibility of the game (a good thing). However a low resource game does not unequivocally mean you must have a logistics network spanning 500+ tiles.

[Khyron]

Or when your goal isn't to 'beat' the game, but to make cool factories and designs. This is why there is a lot of discussion on end-game content even though the game is in alpha. Endless play is fairly popular among the community here.

MMmm, I agree. It's just that I'd rather the devs do more work on end game content since that as an immediate impact on affects all existing and potential customers, whereas I think this request only applies to *some* of the existing customers. The needs of the many, etc. I really don't mean this in a disparaging way, this is a perfectly valid way to play the game. What I'm trying to say is that too much focus on satisfying the wishes of these extreme players is (imho) not a sensible thing to be doing right now.

And that's doubly true when there are some methods to cope with these really massive bases. You *can* use rail or belts to connect clusters of roboports and space the roboports out so that they form discrete logistics networks.

I see this as a parallel for those people who keep requesting smart splitters etc. There are already tools in the game to engineer the desired outcome, albeit by using multiple splitters. Similarly, you guys want new widgets to *better* implement these large scale bases.

I would be more on side if it was completely impossible to do this stuff. But even still, it would be working on content that only affects a subset of customers, so you have to weigh up that priority in terms of development. That's not my job, of course, but it is naturally how I regard feature requests in any of the forums for games that I visit. I hope this explains my perspective a bit.

[ssilk]

Even if I think, that we play low resource games quite similar (no, i don't built much walls) I think it is not you, who should decide, if I play with logistics or with train. Cause it depends on many more factors.

How splitted are the resources? It makes sense to harvest many small fields not by belts but with logistic bots.
How much energy do you have? Is there a big lake, you need to go around?

And more than a dozen more factors.

I think the best argument is with the upcoming multiplayer. Players could define areas, which they are responsible for. And then they can say: well we exchange our products here. It won't be possible, if everything is one big logistic area.

[SHiRKiT]

... I'm sure we all admit this can be resolved by segmenting and hybridising the network (clusters of roboports linked by belts or rail), but you guys would rather have a more advanced drone network. ...

It's just a suggestion, it's not a must. Right now we deal with this by building hybrid networks, by changing the factory design, by using belts to transport the resources from one production chain to the other. It's an idea to improve the logistic network, so it's more roboust.

If not implemented, we'll still deal with this by using alternate methods, which includes setting up trains in the middle of my base (which for me is an absurd), setting up belts that bleed off the production from one place to another (which isn't too bad of an idea in the end, since it'd be the same effect, but requires a lot of infra-structure to be built per product made) or modding the game and creating a new logistic robot that flies up to five/ten times faster.

But would be cool to have colored network it would. May never be implemented due to game mechanics or the amount of work required to do the task. I'm happy either way.

[hitzu]

Remember the new year friday facts? Devs asked us for ideas to make floors usable and functional. Colored logistics floors, that's exactly what we need

[ssilk]

But you don't see the floor everywhere you need to.