Friday Facts #224 - Bots versus belts

[Nel]

I personally dislike bots, as I think of them as "cheating" because of their property to basically teleport items. At the same time, trying to build something like a Mall without them is incredibly tedious and results in a massive belt spaghetti construction. This is because belts can be contaminated, and as such, you want to keep them clean, i.e. only one item type on one belt. This in turn requires one belt lane per item type. In contrast, a train can transport different item types using the same infrastructure.

Basically, what I am getting at, is that some middle-of-the-road transport type is needed, that works on a rail-like infrastructure (i.e. different item types can be transported on the same item tile without contamination), which is significantly smaller than the current rails. I could imagine a kind of monorail-crane system, that is used in real factories. Its rails would occupy only one tile and a 90 degree curve at most 3 tiles. Logistically, it could either work like trains currently do and use dedicated transport entities, or the stations would act like provider / requester chests. I.e. a stop would request items from the network it is connected to, and a provider would then send a transporter entity on its way to the recipient.

[zebediah49]

I personally dislike bots, as I think of them as "cheating" because of their property to basically teleport items. At the same time, trying to build something like a Mall without them is incredibly tedious and results in a massive belt spaghetti construction. This is because belts can be contaminated, and as such, you want to keep them clean, i.e. only one item type on one belt. This in turn requires one belt lane per item type. In contrast, a train can transport different item types using the same infrastructure.

Basically, what I am getting at, is that some middle-of-the-road transport type is needed, that works on a rail-like infrastructure (i.e. different item types can be transported on the same item tile without contamination), which is significantly smaller than the current rails. I could imagine a kind of monorail-crane system, that is used in real factories. Its rails would occupy only one tile and a 90 degree curve at most 3 tiles. Logistically, it could either work like trains currently do and use dedicated transport entities, or the stations would act like provider / requester chests. I.e. a stop would request items from the network it is connected to, and a provider would then send a transporter entity on its way to the recipient.

It should be noted that you can do this using belts -- you just need a closed loop, and then either a circuit network contraption, or just to re-gather all of your ingredients back up via filter inserters. Dump from chests into belts, merge belts together with splitters; go around the loop; filter back into chests. It has astonishingly low throughput though.

A miniaturized version of the train system -- preferably with logistics capability -- would be a really cool alternative. Maybe it would basically move single-stack chests around. So you can request a stack of gears, and the system goes and delivers a stack of gears (assuming you were providing them somewhere). I'm thinking something like what MagneMotion does, where individual products (Factorio scale would require it to be stacks) are transported on independently controlled maglev carts.

[provet]

I don't think bots are the problem, but better belts as described, the complexity for bots could go up, yes. But I rather have them controlled by combinator than nerfing their logistic capability for increased complexity.

Also: Why not just introduce a start-up setting that disables them instead of forcing the issue either way! The same with nukes, artillery, trains, belts etc to increase difficulty/variety

[TiMatic]

Disabling things that are directly build by the player is not realy usesfull, because everybody can decide during the game if he uses it or not. So a disable option is only usefull for things controlled by the AI, for example enemies or pollution.

[ske]

introduce a start-up setting that disables them instead of forcing the issue either way! The same with nukes, artillery, trains, belts etc to increase difficulty/variety

Yes, I think that'd be the way to go. Let those who want keep it for now and change the default settings later. The game is in the stage where it could be considered feature-complete (for version 1) and the balancing can get serious. Usually, this would be the stage where only beta-testers have seen the game and you can easily make such drastic changes to the game mechanic. Now they're in bit of a pickle.

[BrickVoidFCT]

I find myself trying to get to bots all of the time in a Factorio LP but I run into a lot of issues trying to get there, so instead of bots vs belts I would like for the Factorio Devs to address early-game issues in a better manner. Such issues include, but are not limited to, the following:

1. Pollution makes biters eat my base problem: Usually I get to the stage of around Oil Processing, and with default pollution diffusion settings I find that constantly, biters will just come along and start eating my base. I made one production section for red and green potions too large in one save and had to abandon it because there were swarms of biters filling a large portion of the screen from an unknown unmapped section these biters were swarming from. It's one thing to have a biter swarm but when they just start chewing your base up they don't seem to stop or draw back for regrouping, nor do they scale down their attack once started, they just continue relentlessly eating anything in their path. I thought these aliens were supposed to have a purpose in life, but I did not think it was to feast on entire sections of metal and artificial bases!

Seriously, redesign early game attacks so that they can scale back somewhat to that which the player can deal with. I had a few gun turrets and some ammo but it was useless as I had just started setting up defenses and laying out walls and the aliens just waltzed right in and forced me to abandon the save. All that work for nothing. *sighs* I had to set the pollution dissipation rate to 500 in Factorio .15 from Steam in order to make the pollution problem go away, temporarily.

2. Mixed resource deposits: This is something that could easily benefit from tweaking yet you Factorio Devs are more concerned about robots before I've even had a chance to get anywhere near a satisfactory level of production. I usually find myself more concerned with trying to split off mixed deposits and the associated logistics once filter inserters are available to me. How about an improvement like a "filter splitter", for instance? I would envisage that a basic version could filter 1-3 types, an intermediate version could additionally filter to a specific lane, and an advanced version could filter to a logistics chest integrated into itself. That's what I keep telling myself I need in Factorio, and that's what I'd like you to focus on.

3. The resource balancing problem itself. A lot of the time I find myself trying to manage my iron to supply everything with all of what I need on yellow belts. I placed a restriction on myself recently, automate it or don't build it, unless I actually need to manually craft it. This means that early game past the very early game stages is mostly about finding and managing enough iron to get everything supplied. I seriously would like tools and visualization apparatus that better allow me to see exactly where my resources are and where they're going to. It is just too hard to manage it all not knowing which part of your base is going to spend it's entire life dragging your iron supply to nothingness. I would prefer to know exactly how much production a supply of iron could maintain, and there are currently no early game tools or features that help me figure that out. Please, make some!

4. Defenses eat all my iron. A short note to the Factorio devs saying, why is it just making Gun Turrets and Ammo for them eats every piece of iron I can produce? Is this seriously something I need to be doing, devoting an entire resource mining and smelting section to a single defensive production facility? I thought Factorio had options in this regard, guess I thought wrong. Very irritating to find that when I set up for defenses production, that I have to devote a heap of iron plates to it!

Logistics bots would be nice if I could actually get to them, I find myself wondering, have the Factorio Devs given enough thought to what players must do to get to them? So it's late-game, big deal, I can't even get to mid-game oil production without heavily altering the default settings, dammit, brain, work harder. Oh wait, maybe I'll just make the game easier on me. Meh, i would prefer a standardized game experience where there's some leeway given to the player.

That's all for this post, hope you all enjoyed a certified noob's experiences of early game Factorio .15! This game intrigues me a lot otherwise I'd stop playing it!

Cheers ...

BrickVoid

[Dave64738]

For the base game I think bots are probably overkill. But when you start getting into the complexity of something like Bob's mods, they are really essential. I can't imagine how my last Bob's build would have worked without bots. Dev team: if you think bots should be removed then I challenge you to climb the Bob's tech stack without them. Of course, they could always be modded back in if you leave the hooks in place.

[Green_Baron]

And i hoped for more and complex recipes, instead the plan seems to be streamlining and simplification.

I better read the features before updating from now on. It might come out that 0.16 is the peak in terms of game experience ... or ?

[MtNak]

I really prefer Universe 1. With the nerf i would only build more robots.

It lets me think at a whole different scale without thinking about 1k hours into a save with a whole lot of tedious work that i have already done hundreds of times.

[HerpicusMcDerpington]

I once tried to setup a smelter array with bots. It turned out to be less effective as if i just had build it with belts, no matter how many bots and ports i threw it it.
As ive seen in others peoples bases, bots, for a high end efficient base, actally take alot of thought too.

I use them for personal logistics and builder trains loading chests

If you dont like the fact that they can help in the midgame, when you "urgently" need that one item type where only a 500 belt long line could bring it, then just stop selling the game.

Its not the bots themself, i could work around that.
Its the fact that you want to force a playstyle on people that lets me want to play the game less right now, not when your crazy ideas are implemented.

On the other hand....
Isnt it a bit too early for an Aprils fools joke?

[Sebster]

I really like the idea of being able to disable bots at game start. I prefer to not even see them as an option if I have decided not to use them. And why stop there, as some ppl have mentioned. Options to disable Nuclear Power or anything else just increases replayability imo.

[SHiRKiT]

I say a couple of things about bots vs belts:

- Factorio is just too complex for Logistics Bots not exist. It becomes tiresome later on the game needing tons and tons of long belts for higher throughput.
- Mods like Bob's mods adds TONS of items and recipes. Belts becomes INSANELY hard to build a big base on that scenario.
- Belt bases are more performance heavy. Most players may not note it, but a lot of people don't have the brightest of the computers, and to build a somewhat big, we needed bots on 0.15 and before.

[factorio_player487]

I think taking away gameplay options away from players is almost never a good game design idea. The early to middle game is about mastering belts in order to unlock science packs. After a point, the player should naturally become more powerful and bots coupled with trains and belts helps in building a good megabase. Players who don't want to play with logistics bots could have an option to turn it off but don't nerf something useful to the rest of us.

[BuddEBoy]

Here's an idea for nerfing bots: roboport range. It would act the same as research speed. Let's say when the player finshes the research that gives them the robot (robotics i think) it will branch off of that. A roboport without the research would have like 1/4 of the og range. When it is fully rearched, it'll have the og range or better! This is probably the best way to nerf bots, without actually nerfing bots!
Sincerely,
Me (a.k.a. Buddeboy)

[Muppet9010]

interesting to see 3 boxing/palleting mods all our recently, wonder if this approach has been considered for core game as a buff for belts and trains?
These ones seem to block or limit bots ability to move these so bots can just do the final item delivery. by preventing pallet stacking and limiting internal contents to 1 stack they don't change train max unit capacity it seems either.

https://mods.factorio.com/mods/paradigmsort/boxes
https://mods.factorio.com/mods/twanvl/pallets
https://mods.factorio.com/mods/mckilljo ... lPalleting

[ske]

interesting to see 3 boxing/palleting mods

This looks quite interesting. Does it feel like vanilla factorio?

Is the stack size of 1 a technical limitation or just due to the mods? Increasing it to maybe 5 or 10 could increase the train capacity substantially...

[CmdrKeen]

My suggestions for the changes with bot/belt mechanics.

Merging logistic and construction bots into just Flying Robots.

Thats a great change.

Removing Active provider, Buffer and Requester chests and having just Passive provider and Storage chests.
Flying Robots will do everything construction bots do now and also supply the player.

Limit what the flying robots can do in a logistic capacity: IE: moving intermediate products only, or some type of weight metric.

Simplifying recipe complexity a bit so the belt spaghetti does not get ridiculous.

Don't simplify recipes. The belt spaghetti is largely from the lack of options for splitting materials and beacons. Remove beacons and compensate with more assembler tiers or increase the effect radius to allow more space for belts.


If the bot changes take place as intended, a lot of people will be flummoxed by the lack of ability to finely control how materials on belts are split. We have large designs to simulate the behavior of 2-lane splitters on a larger scale but they take a large number of 2-lane splitters to implement, and take up n^2 space on average. There needs to be more splitter sizes and multiple splitter behaviors.

[Zavian]

I find myself trying to get to bots all of the time in a Factorio LP but I run into a lot of issues trying to get there, so instead of bots vs belts I would like for the Factorio Devs to address early-game issues in a better manner. Such issues include, but are not limited to, the following:

1. Pollution makes biters eat my base problem: Usually I get to the stage of around Oil Processing, and with default pollution diffusion settings I find that constantly, biters will just come along and start eating my base. I made one production section for red and green potions too large in one save and had to abandon it because there were swarms of biters filling a large portion of the screen from an unknown unmapped section these biters were swarming from. It's one thing to have a biter swarm but when they just start chewing your base up they don't seem to stop or draw back for regrouping, nor do they scale down their attack once started, they just continue relentlessly eating anything in their path. I thought these aliens were supposed to have a purpose in life, but I did not think it was to feast on entire sections of metal and artificial bases!

Seriously, redesign early game attacks so that they can scale back somewhat to that which the player can deal with. I had a few gun turrets and some ammo but it was useless as I had just started setting up defenses and laying out walls and the aliens just waltzed right in and forced me to abandon the save. All that work for nothing. *sighs* I had to set the pollution dissipation rate to 500 in Factorio .15 from Steam in order to make the pollution problem go away, temporarily.

Sounds like we are almost playing a different game. Possibly we are just playing with different startup settings (which can change the challenge lvl significantly. I normally play on railworld settings, and with a large starting area. There are also some player who feel biters aren't dangerous enough, even on deathworld settings).

I can normally deal with the first few attacks with just my starting pistol and ammo. After that I'll carry a submachine gun, and a couple of turrets. That is normally enough to kill any biters that attack until after I have blue science running. They may chew up a few buildings before I can get there, but buildings are easy enough to replace. Note that it is almost always cheaper to rebuild a few building than to try to defend your entire base at this stage of the game. I'll place a few turrets (probably 6-12 turrets total, often in clusters of 2, and with only around 25-50 ammo each) in areas I expect biters to attack, but I'm not concerned if they get past them and hit something else. If I'm getting constant attacks from a particular area I might take some turrets and clear that base.

2. Mixed resource deposits: This is something that could easily benefit from tweaking yet you Factorio Devs are more concerned about robots before I've even had a chance to get anywhere near a satisfactory level of production. I usually find myself more concerned with trying to split off mixed deposits and the associated logistics once filter inserters are available to me. How about an improvement like a "filter splitter", for instance? I would envisage that a basic version could filter 1-3 types, an intermediate version could additionally filter to a specific lane, and an advanced version could filter to a logistics chest integrated into itself. That's what I keep telling myself I need in Factorio, and that's what I'd like you to focus on.

You only get mixed resource patches where 2 resource patches overlap. Most of the time you can just carefully place your miners and belts, and get most of the patch mined without needing to deal with mixed belts. If you want to mine the rest, you can use filter inserters.

3. The resource balancing problem itself. A lot of the time I find myself trying to manage my iron to supply everything with all of what I need on yellow belts. I placed a restriction on myself recently, automate it or don't build it, unless I actually need to manually craft it. This means that early game past the very early game stages is mostly about finding and managing enough iron to get everything supplied. I seriously would like tools and visualization apparatus that better allow me to see exactly where my resources are and where they're going to. It is just too hard to manage it all not knowing which part of your base is going to spend its entire life dragging your iron supply to nothingness. I would prefer to know exactly how much production a supply of iron could maintain, and there are currently no early game tools or features that help me figure that out. Please, make some!

Mostly you can just eyeball it. If iron is reaching the ends of the belts, the it's doing ok. If not, then you need more iron. If you want to know how much you need at each stage of the game, then a good approximation is half a yellow belt of iron is enough for 5 red and 6 green science assemblers. Add another 1.6 belts for 12 blue and another 1.4 belts of iron for 5 military (both of these ignore iron required for steel production, which is around another 0.5 belts of iron). So call it 3.5 belts iron and just over 1 belt copper until you are ready to build production science and/or high-tech. (Once you start production and hi-tech you'll need more of everything). If you want planning tools there are mods available (eg helmod), and web-based planners (I use https://kirkmcdonald.github.io/calc.html ). The maths is also simple enough to do yourself, so the game itself doesn't need to contain a planning tool.

[Sander_Bouwhuis]

I've only once ever used logistics robots (in my second playthrough when I discovered they existed at all). Never used them again. Belts are really visceral and are the reason I'm playing Factorio.

Construction bots, though, should be available OUT OF THE GATE! Having to put down building blocks one-by-one is soooooooo tedious after a few playthroughs. It keeps me from starting a new game. (this, and the uselessness of it all; I don't like to just build stuff to just get research bottles).

[meganothing]

This isn't 20 rockets-per-second surgery, here. It's right there in the big bold text at the top of the page. Sure, he walks it back a little further down the page, but if he's "thinking about it" today, that means he wants to do it tomorrow.

This is called a straw man argument. He specifically says "Don't worry, logistics bots won't be removed from the game, mostly because it's a feature that has been developed and polished quite a lot. Also many players love the feature, and we've all become used to it over the years.". Sure, if the whole forum had said "we love scenario 2" that would have been something else, but he clearly says that he doesn't expect that himself.

Also I've seen all the other nerfs Factorio's devs have packed into update after update in the past, and I've seen those updates do nothing but make the game more painful to play.

Also I've seen too many other promising indie games on steam nerf themselves into painfully unpleasant grinds just because the dev thought people "weren't playing it right."

Followed by an association fallacy aka guilt by association.

Sure, we can do the "always assume the worst"-game and panic at every thought experiment, but then the developers won't talk freely with us anymore but assume the worst as well. That would be the end of open dialog.