Add functionality to limit the number of incoming trains per station.

[jendenbm]

The absence of this feature is what made me quit Factorio last time. I always come back about once a year, play for a few months and then quit when I realize the game is missing some very needed fundamental mechanic.

The ability to limit how many trains are being sent to the station is absolutely crucial, in my opinion. It becomes so tedious late game to manage mining/smelting that you just cannot take it anymore. I don't know why would devs not have this feature introduced - you can clearly track how many trains are traveling to each station by clicking on individual trains, but you cannot control it. Why not give player an option to exclude a station based on how many trains are going there?

[ptx0]

The absence of this feature is what made me quit Factorio last time. I always come back about once a year, play for a few months and then quit when I realize the game is missing some very needed fundamental mechanic.

The ability to limit how many trains are being sent to the station is absolutely crucial, in my opinion. It becomes so tedious late game to manage mining/smelting that you just cannot take it anymore. I don't know why would devs not have this feature introduced - you can clearly track how many trains are traveling to each station by clicking on individual trains, but you cannot control it. Why not give player an option to exclude a station based on how many trains are going there?

they did, you can install a mod, like the others listed in this thread!

[jendenbm]

they did, you can install a mod, like the others listed in this thread!

You're missing the entire point. The idea is to NOT use mods. Many people (myself included) resent the idea of any kind of modding, period. It ruins the game experience.

The concept of disabling/enabling the station ALMOST works in pure vanilla, the only thing missing is the train tracker to prevent multiple trains from being sent to the same station at once.

Besides, there are NO mods currently available that fully mitigate the problem anyway due to limitations of the mods themselves. LTN does not allow you to use the same name for providers (forces you to name your outposts iron-1, iron-2... iron-50) and receivers (does not allow you to balance between smelters with the same name) which defeats the entire purpose of the mod. It's the same as using dedicated trains from each mining outpost going to dedicated smelter - ugly and cumbersome solution.

Again, adding a simple incoming train tracker (counter) would fix 100% of the problem in vanilla. We wouldn't even be discussing this if the feature was available.

[ptx0]

they did, you can install a mod, like the others listed in this thread!

You're missing the entire point. The idea is to NOT use mods. Many people (myself included) resent the idea of any kind of modding, period. It ruins the game experience.

[snip]

Again, adding a simple incoming train tracker (counter) would fix 100% of the problem in vanilla. We wouldn't even be discussing this if the feature was available.

i don't see why you despise mods so much, they 'ruin' the experience?

but you just said that NOT having this is ruining the experience:

The absence of this feature is what made me quit Factorio last time. I always come back about once a year, play for a few months and then quit when I realize the game is missing some very needed fundamental mechanic.

you can just add ONE mod that does what you want, without "ruining the experience".

imo, people should just stop asking to have moddable features included in base - it's a waste of wube development time, they can be doing many other more important things that would make modding better. because then YOU can make whatever experience you want.

remember, modding is a part of the developers' intention - if you feel something is missing, add it.

don't be lazy and expect wube to do it for you.

[ssilk]

Thank you ptx0, this is part of what I think about it all the time when this discussion began in General (viewtopic.php?f=5&t=80501 )

I’m not against this suggestion, especially the second part “Add some more signals for trains and stations so players can implement the limit themselves“ , it may offer some new possibilities to play with trains. What me triggered is mainly the rant against mods and that someone quits Factorio because of a missing feature.

I I looked back to the OP: For the first part “Add LTN-like functionality that limits the number of trains that can have a particular station (out of many with the same name) as a destination“ I’m sceptical.

I’m sceptical due to thousands of hours playing mainly with trains. I think a limitation will not solve the targeted problem completely and is questionable from gameplay value (it is eventually hard to hear that), but it looks for me like an idea, that tries to solve a problem, that only exists because of players behavior.

What I mean is: this suggestion looks like a solution to a problem, but it stretches only the time that makes the trains unuseable a bit. Why?

Think about this:

- problem arises the first time, when too many trains try to come to one station at a time.
- this is (for example) because player added too many outposts that targets too many trains with one station
- by implementing this suggestion the problem is eventually solved, but the trigger was the players behavior
- so logical conclusion: if player doesn’t change behavior, and he/she will add more trains for the same station, another problem will arise. Something similar or completely different.

I know that it is so, because I’ve tried to play this scenery several times: if you solve one problem (and yes, you can solve such problems with vanilla but it is complicated and many won’t like my solutions) and continue to add pressure to one station or the whole rail system another problem will come up. And you can never foresee, what the next problem will be. And this doesn’t end in one perfect solution, it ends in a total mess of solutions.

So in my opinion this problem can only be solved by the game by limiting players behavior.

LTN is a good example: it limits players behavior to avoid the described problems. What can be seen as disadvantage of that mod is the solution (one solution of many possible solutions) to this problem.

And to hit on top more arguments pro modding: viewtopic.php?f=6&t=30240 I want to make suggestions, but I got links to mods!!
There is a much, much higher chance, that someone is willing to implement a mod instead of implementation in vanilla, if you like it or not.

Again: I’m not against this, but it will not solve problems, it gives the player just more and easier possibilities to add more complex problems.

[P.E.T.A.R.]

don't be lazy and expect wube to do it for you.

Seriously?

Boskid already made a prototype in the thread that ssilk mentioned in the General section (image below).
I don't like to use any mods either and I would like to see something like this with signals implemented in vanilla.
Also I don't want to downplay the massive effort the mod makers invest in making additional features (seriously you guys are amazing!).

The truth is I don't like to mess with mods, and I'm sorry that this will trigger some but it's simply the way it is, I like to play a game made purely by it's original developers in its original form. Mods are not the answer to everything.

[ptx0]

Boskid already made a prototype
[...]
The truth is I don't like to mess with mods [...] Mods are not the answer to everything.

in this case, they are - boskid has said the limit trains feature is "postponed indefinitely". it probably won't happen.

[ssilk]

The truth is I don't like to mess with mods, and I'm sorry that this will trigger some but it's simply the way it is, I like to play a game made purely by it's original developers in its original form. Mods are not the answer to everything.

It triggers me a bit, but not too much, I really understand your opinion as I had some years ago the same idea.

I just want to show the situation as I see it:
- this feature has a very low chance to be implemented. What boskid wrote in short: feature has been implemented as a spike to estimate effort and it looks bad to make this complete in contrast to all the other things, that can be done. Which means in Factorio (as ptx0 already interpreted): never. Just an economical decision. It can change, but this will take much time.
- my argument (that I cannot prove but is based on very long experience) is, that this feature will not really work well, because it covers only the real problem, which is that a train station cannot handle unlimited numbers of trains. You just shift this limit a bit. It is not a complete solution.
- this next argument deepens the former: this feature is useless, if you make the train network a little bit bigger, because the trains will take some time, and - much more irritating - they will take different times, depending on travel distance. You need to take then into account that if you limit the number of trains it could still be so,that there are some times no trains at the station (because they are all on the way) and some times still too much (because they had different travel times and came in all at the same time).


It’s like in reality, a little bit delay can introduce a big chaos into the travel plans. You can control this chaos only up to some level. After that level you need to make bigger changes to keep control. This is typical for Factorio, the robots have for example the exact same problem. I would bet with you, this is part of the game.
So - logically - limiting the number of trains is just one screw of many, and it is not very effective - from my experience.

Much more effective is:
- building big. I mean really big. The scales changes, when you begin with trains. You need large areas, where the trains can wait, before and after the train station. Like in reality.
- limit the number of trains that targets one station. The simple way is to remove trains from the schedule to that train stop (make a new target for example). Also like in reality
- building parallel stations (if done correctly nearly doubles the throughput).
- build correctly: if the train cannot be emptied because the output chests are full for example. Or you need to count correctly, because of the large space needed.

There are more tricks. It depends on the situation and cannot be generalized. See finding those as part of the game, something which can be fixed with the things you have. I guarantee if you find a solution your brain releases a nice cocktail of good hormones.

But quitting Factorio, because it “misses” a feature that must be implemented in vanilla otherwise you cannot play ... you don’t see me nodding.

[scooter010]

Or you can use HTN. "Haphollas Train Network".

It works in Vanilla just with circuit networks. But has the sam eproblems than LTN. Maybe too complex.

[disentius]

they did, you can install a mod, like the others listed in this thread!

You're missing the entire point. The idea is to NOT use mods. Many people (myself included) resent the idea of any kind of modding, period. It ruins the game experience.
...

You want your specific playstyle, NxN train networks with several same-named stations, to be a vanilla experience.
At the same time you argue that you want to play the game as the Devs intend it to be ("vanilla")

To me that sounds a bit contradictory.

Either use Mods if you want to change a game mechanic, or play vanilla.
Personally i like the challenge of trying (and failing spectacularly) to design a train network with same names which works.

[jendenbm]

imo, people should just stop asking to have moddable features included in base

I repeat, this is not a moddable feature because LTN does not solve this particular problem. Besides, how's it even an argument to force people to use mods if there could be a simple optional feature in vanilla? Why even have anything in vanilla then ? By the same token - just export trains and rockets into mods and let people use them. It's no different from what you're saying.

What me triggered is mainly the rant against mods and that someone quits Factorio because of a missing feature.

That was the problem with factorio for many people since the beginning. You start playing, eventually you start implementing more efficient and more elegant solutions to a problem until you face the wall of the game limitation that cannot be avoided. Then you have nothing left to do but quit. Happens way too often.
The devs openly admit that problem. They stated many times that new changes/features are geared to address learning curve and annoying game mechanics (such as removing mining picks, re-working oil and many others) to improve the experience of less-hardcore players and make the game more enjoyable. But clearly there's still much more work left to be done.

but it looks for me like an idea, that tries to solve a problem, that only exists because of players behavior.
- problem arises the first time, when too many trains try to come to one station at a time.
- this is (for example) because player added too many outposts that targets too many trains with one station
- by implementing this suggestion the problem is eventually solved, but the trigger was the players behavior
- so logical conclusion: if player doesn’t change behavior, and he/she will add more trains for the same station, another problem will arise. Something similar or completely different.

I'm disappointed that you brush it off by labeling the problem as "players behavior". That is extremely arrogant and condescending. Or perhaps you're totally missing the point.
The goal is to treat all mining as a "cloud service" where you don't micromanage individual mining outposts and name them all the same name, so that the trains figure out which station to use. The trains make this decision, not the player. The player can then increase the number of trains as the number of mining outposts increases.
The problem why LTN does not work is the hardcoded limitation that forces you to use unique names for providing and receiving stations, which breaks the logic if you want your receivers load-balanced by using the same name. I.e. the receiving station is to be chosen based on which one is free (with regular signals right before the receiver split), and not at the time of schedule creation (like it's done with LTN) as you can easily end up with 1 schedule for receiver-1 and 10 schedules with receiver-2. The non-LTN approach works perfectly for receivers but does not work with providers - because of multiple trains choosing to travel to the same station until the first one reaches it.
If only there was in-game feature to limit the number of trains that provider station (or any station for that matter) could have at any given time - that would literally solve the whole dilemma. There would be NO other problems, not sure why you claim otherwise. The trains would attempt to travel to the first closest outpost, and if there's already a train going there - pick the second closest outpost and so on. In the long term the closest outpost will see more visits simply because of proximity - and that is totally fine, that exactly is the goal.

[ssilk]

You start playing, eventually you start implementing more efficient and more elegant solutions to a problem until you face the wall of the game limitation that cannot be avoided. Then you have nothing left to do but quit. Happens way too often.

I don’t know. If I have to do with new technology I try to understand their limits before I plan to use them. That works well in real life and in games like Factorio.

I'm disappointed that you brush it off by labeling the problem as "players behavior". That is extremely arrogant and condescending. Or perhaps you're totally missing the point.

I currently think more you miss it. I try to show it from another side.

You described that well as a “cloud service”. That is a nice picture. I add: Player doesn’t want to micromanage, just wants to play.

But how are cloud services working? That is quite different and how that cloud services route your request and what will happen etc. is very much dependent of what you really want to achieve. Facebook has surely a completely different setup as google. Even WhatsApp will work different than FB. In other words: it’s hard work to make a cloud service running smoothly.

Here it is similar. And what I mean with “players behavior” is just this: player wants to have a smooth cloud service, but the game cannot know how it would work in this special case. Because it is always a special case!

Why? Assume you have this feature and it is working. Or better: it pretends to work. Then you add a new station to this cloud and suddenly you run into the same problems. Why? Because you cannot automate complex or chaotic systems. (I reference to https://en.m.wikipedia.org/wiki/Cynefin_framework )

You can try to keep a system in complicated state, then such automation will work, no question. But players behavior is complex. He has the ability to bring the system from complicated to complex. It has nothing to do with your play style or that I want to forbid anything, but what I want to explain is, that the train network is super fragile (In this “change of system behavior” from simple to chaotic) and adding such kind of automatic (limit number of trains per stop) - without including, that humans are involved and can change things - is in this case a way to implement something, which will not work in every case. Bug reports will be written. More players that want to quit, because the game doesn’t do what they want to do. And so on.

Me as a software developer would really avoid to implement anything which should do X but also can have the effect of -X and Y. It’s a software-anti-pattern.

The problem why LTN does not work is the hardcoded limitation that forces you to use unique names

I had lately a discussion with Optera about this, and in the end I didn’t have an argument left. viewtopic.php?f=214&t=85023
(Or could say I’m waiting for inspiration of a reason, that would make this unavoidable )

If only there was in-game feature to limit the number of trains that provider station (or any station for that matter) could have at any given time - that would literally solve the whole dilemma.

No it wouldn’t. For example: how would you count the limit?
I would count it so, that a train that targets a station will add a counter. And if it leaves it will decrease that counter.
We assume now you have a limit of 5 trains. That will not avoid that your station is overwhelmed or under supplied . Why? Because the trains have different travel times. And because there are signals which will force trains to wait. So you cannot foresee,how long any travel will last.

You can estimate it. You can say: the travel to this station will take in 90 percent of travels between 1 Minute and 3 minutes. But the left 10% are outside of this range. They are chaotic or will add chaos to the system.
And now take into account, that you have more stations, more trains, that also need to wait on signals and more traffic will add more chaos. There is some point, where you add one more station and the whole system breaks. And then player writes bug reports or quits game.

There would be NO other problems, not sure why you claim otherwise.

I hope I explained it in detail. I underline, that I tested such things in several configuration and tried to keep the system out of chaos. But once you install limitation like this and add more and more components (tracks, signals, trains, more transports) to your train system it suddenly will behave chaotic.


Edit:

after reading my answer - what might be misunderstood-able - is that your suggestion will work in the case of a more or less static train system. Once you have installed the limits of the train stops properly and don’t touch the system anymore (no increasing production, no added trains, signals, stops...) it will work forever. Of course, why not... but I doubt that players will play like this. And that’s what I (also) mean with players behavior.

I doubt, that someone likes playing with this limitations, I would not say “play” to it anymore.

On the other hand I can imagine myself using this feature - very practical feature and I can think of several use-cases. But I know what I do, when I’m using it. I know what can happen, because I learned how the system works. In most cases, but it will always surprise me again.

You admitted above that you start implementing without thinking. I interpret that as you will not be able to use this . Sorry to say that, I don’t mean it so hard as it sounds in written language. Because that is also part of the fun of Factorio and I would say the majority of players sees that similar to you. Which is fine.

That is the reason, why I think this feature will not bring the wanted effect, it has a questionable gameplay value, because the feature deludes players to think of it as a solution to their problem. Which it isn’t as I tried to explain.

[ptx0]

The problem why LTN does not work is the hardcoded limitation that forces you to use unique names for providing and receiving stations, which breaks the logic if you want your receivers load-balanced by using the same name. I.e. the receiving station is to be chosen based on which one is free (with regular signals right before the receiver split), and not at the time of schedule creation (like it's done with LTN) as you can easily end up with 1 schedule for receiver-1 and 10 schedules with receiver-2

man that was an awful lot of unnecessary text you had in the post other than this actually important info.

LTN doesn't have this limitation anymore... as of a few weeks / a month ago. and in either case, you're not using LTN's features correctly if you're trying to "load balance" this way. LTN has train limit settings. go to the LTN forum, ask questions.

don't worry, i used to be like you, and ask for behaviour to be in vanilla.

mods are for this situation, a user who has played the game for a hundred hours and now knows what they would change to tailor the situation.

i don't thiink aanyone will have any sympathy for you not wanting to use mods - no one is forcing you to do anything, plenty of people seem to get by just fine with the current train mechanics. see the 20k SPM base posted lately? no mods.

[ssilk]

TL;DR: I changed my mind.


There is this parallel thread in viewtopic.php?f=5&t=80501

And Ambaire posted something which makes suddenly a lot of sense to me:

I don't really see the problem here.

Suppose a train stop is set to allow 1 2-4 train to path there, with enabled condition of ore > 8000. There are 4 trains waiting in a stacker near the ore unload. The train stop opens up, and one train paths there. The other three wait, since the limit of 1. The train loads up on ore and heads to unload, and the train stop still has 10k ore left, so a second train heads there once the first leaves. After that one leaves, there is 5k ore left, so the two trains heading to unload resume after in the stacker and no additional trains leave the stacker.

No deadlocks / jams / what have you.

Why is this not already a feature? I could have really used it in my latest vanilla playthrough...

This will work quite reliable - of course. Why? Because the distance between the two stations is so short. No chance that any type of chaos can influence this. The player created a really simple setup.

It opened me a bit my eyes, because it where be so simple stupid setup, that I never would like to play with. ahm, no, I just could not imagine one setup, that guarantees no problem, never. But this one does.

As I mentioned above, other setups cause problems with growing rail system. This will not be the case if setup is so, that the limits will make sense; I mean that it brings the train system back into a simple or complicated state. Which are existing. The players task is now to find such kind of systems, which work well balanced with the train limitations. Which has a good game-play value.

As already said, I changed my mind.

[Nosferatu]

I'm using this mod for 4 months now:
https://mods.factorio.com/mod/Concurren ... estriction

It allows you to simply limit your stations and I have no issues.
Of course you have to take care that at least on one end of the schedule stations have enough waiting room so that all trains can go somewhere but that's a problem you also have in vanilla.
If you build more trains than your stackers can take you get congestion.

[MiniHerc]

TL;DR: I changed my mind.


There is this parallel thread in viewtopic.php?f=5&t=80501

And Ambaire posted something which makes suddenly a lot of sense to me:

I don't really see the problem here.

Suppose a train stop is set to allow 1 2-4 train to path there, with enabled condition of ore > 8000. There are 4 trains waiting in a stacker near the ore unload. The train stop opens up, and one train paths there. The other three wait, since the limit of 1. The train loads up on ore and heads to unload, and the train stop still has 10k ore left, so a second train heads there once the first leaves. After that one leaves, there is 5k ore left, so the two trains heading to unload resume after in the stacker and no additional trains leave the stacker.

No deadlocks / jams / what have you.

Why is this not already a feature? I could have really used it in my latest vanilla playthrough...

This will work quite reliable - of course. Why? Because the distance between the two stations is so short. No chance that any type of chaos can influence this. The player created a really simple setup.

It opened me a bit my eyes, because it where be so simple stupid setup, that I never would like to play with. ahm, no, I just could not imagine one setup, that guarantees no problem, never. But this one does.

As I mentioned above, other setups cause problems with growing rail system. This will not be the case if setup is so, that the limits will make sense; I mean that it brings the train system back into a simple or complicated state. Which are existing. The players task is now to find such kind of systems, which work well balanced with the train limitations. Which has a good game-play value.

As already said, I changed my mind.

Here's a second example I thought of, which seems to work perfectly. Please tell me what if any problems with it you see.

Suppose a single copper ore mine has two stations each with dedicated ore feeds, and both stations have the same name. The stations are set to activate on copper ore > 8k, limit of 1 2-4 train apiece. There are 6 copper ore trains in an idle stacker. Both stations activate at the same time, and 2 trains head to the station, leaving 4 in the stacker. The trains deplete the ore of the stations and head to unload, then back to the stacker. As the stations get enough ore, they reactivate, each time requesting a single train.

To be honest, I don't really see a need to allow more than 1 train to path to a single station. If this suggestion were implemented, I would most likely always have incoming trains limited to 1 per.

I'm sure there are some more advanced / complex use cases I'm not aware of.

[Nosferatu]

Currently if a station is disabled the train will skip that station und proceed directly to the next one.

[Lubricus]

I like the idea, it would solve the problem to "spread out" trains to many train-stations with the same name. I don't see any way to completely solve that in vanilla in any reasonable way. Trainsystems like HTN (https://www.reddit.com/r/factorio/comme ... haphollas/) doesn't 100% solve this problem and will break down with an to big factories with many spread out stations. There is other variants and tricks that can be added to reduce the train hunting in packs problem further. Systems like LTN in vanilla (https://www.reddit.com/r/factorio/comme ... la_part_1/) that 100% solve it is unreasonable complex.

I think it should be an better way to solve it in vanilla, that you can mod the game should not be an excuse to not improve the game.
Having an easy box to just fill out the limit number sounds to easy and boring I would more like have a way to send a train to a specific train-stations with circuits, so something more like the mod Dispatcher (https://mods.factorio.com/mod/Dispatcher) in vanilla than LTN.

It's very easy to install and use mods in Factorio but the devs have been good to incorporate functions of good mods in the base game to improve it and i think they should continue to do so.

[ptx0]

[...] Trainsystems like HTN (https://www.reddit.com/r/factorio/comme ... haphollas/) [...]

man, that thing is just incredibly awful for UPS.

so if this is the only way to do train limiting in vanilla, well, i think it should be implemented in the game engine too.

[ssilk]

To be honest, I don't really see a need to allow more than 1 train to path to a single station. If this suggestion were implemented, I would most likely always have incoming trains limited to 1 per.

Well, I see that. It makes sense to have a limit of 2, if you really have extreme throughput, so that the next train is already waiting at the station entry, and not in the stacker.

And on the other side, that’s also the point, where the problems begin! When you set the limit to more than one things can get complex quite fast.

So I think to add the most game-value this combination would be useful:
- first research for train limitation is after signal-research. Then you can set the limits only to endless or 1. This gives the player a nice learning curve.
- the second research enables to set the limit higher than 1. This is also a warning for the player: this is a bit different now.