Reverse-facing locomotives contribute to total train power

Ideas that are too old (too many things have changed since) and ones which won't be implemented for certain reasons or if there are obviously better suggestions.

Moderator: ickputzdirwech

Qon
Smart Inserter
Smart Inserter
Posts: 2091
Joined: Thu Mar 17, 2016 6:27 am
Contact:

Re: Reverse-facing locomotives contribute to total train power

Post by Qon »

ickputzdirwech wrote:
Fri Oct 23, 2020 5:39 pm
What am I missing?
Single headed require way more space for those loops. Especially noticable for smaller stations where the loop requirement might not be easily satifiable and at the early stages where you don't have the space for those right where you want it.
ickputzdirwech wrote:
Fri Oct 23, 2020 5:39 pm
Double headed trains are way more complicated to build and have generally a lower throughput. Both will be true even if this suggestion would be implemented (especially if backwards facing locomotives only add ~60% of their power as was also suggested in this topic) and I would continue to use single (!) headed trains as I do right now.
Double headed are simpler. They get it at the cost of reducd throughput. But they have enough throughput anyways, you don't really stress the limits of trains until you reach megabase scale. At that point it might be worth switching over to single headed since you have the infrastructure and requirements etc to really afford the downsides of single headed trains (and they are reduced, relativly) and you have advanced enough to not really need the simple rail network since you are trying to push for performance.

I'm using single headed trains, but I think I might switch to shorter double headed ones for the early game in the future. The signaling isn't more complicated for either type...

mmmPI
Smart Inserter
Smart Inserter
Posts: 2675
Joined: Mon Jun 20, 2016 6:10 pm
Contact:

Re: Reverse-facing locomotives contribute to total train power

Post by mmmPI »

ickputzdirwech wrote:
Fri Oct 23, 2020 5:39 pm
Disadvantages of double headed:
  • Higher fuel consumption
  • Lower speed and acceleration per locomotive.
  • More complicated signaling
  • Lower throughput
What am I missing?
Qon wrote:
Sat Oct 24, 2020 1:24 pm
Double headed are simpler. They get it at the cost of reducd throughput
One disadvantage of the double headed trains , that could make them a bit more complicated for unexperienced player, is if you use more than one wagon, or if the different wagon don't share the same content. The train if going from A to B to A to B would end up "switching" its front loco every time, and thus if you have 2 wagon, they will alternate between being the 1rst wagon and then the 2nd. This can lead to mixing up material and/or make player choose a different system so they don't bother anticipating all the configuration if you have a 3 destination schedule it makes it worse. And if you want to add a destination it also requires planning.

On the other hand i don't think signaling is more complicated it's basically the same signaling except for the stations which are themselves not as complicated as a stacker or a junction. I would even argue that it could be a simpler signaling as it forces player to use signals in a certain way but in a simple configuration. So i think it's good to familiarize with signals to at least try double headed trains.

When talking about throughput i tend to not think about single or double headed trains. In the discussion i think it would be more precise to differenciate the station's throughput, the trains' throughput on a track, and the material's throughput on a track or overall. Station for double headed trains are a bit slower to allow a 2nd train after the 1rst one due to the impossibility for the 2nd train to be halfway in the station while the 1rst one is only halfway out. Which i think is what is called lower throughput in the previous discussion. It can be counteracted by having more stations, which wouldn't take necessarily more space than a Roro/loop station.

The overall throughput of train on a track however doesn't depend on single/multi directionnal train: If you have 1 loco for 10 wagon ( LCCCCCCCCCC) trains are slower vs 2 loco in each direction per wagon. (LLCLL). sorry for stating the obvious :D The overall throuput of material depends more on this particular aspect ( imo ) that the throughput of the stations because adding more stations is often way easier than changing existing train composition which needs the player to adapt existing station .
boskid wrote:
Thu Oct 22, 2020 1:39 pm
There are also other problems when locomotive in wrong direction would contribute: would that mean a train with all locomotives heading into one direction should be able to find a path in opposide direction and travel entirely in reverse? What if there would be 100 locomotives in one direction and a single one in opposite?
For the first part i don't think a train should try to pathfind using the "wrong" direction; if no locomotive is facing it, just don't even try. ( player's intent would be required explicitly by placing at least 1 loco each side as other player already said).
I was convinced by the second part:
Take a train with 100 loco facing one direction. add a loco on the opposite direction. ( to show that player's intent ). Now the train is able to reach max speed on each direction.
Even if acceleration is only 60 or 80%; the way the max speed is calculated is based on the fuel type, not the number or locomotive or wagons. (according to wiki's formula : max_speed = 1.2 * fuel_top_speed_multiplier).

Sounds a bit ridiculous/gimmicky/exploit to me. I imagine seeing one such train, at full speed, in the 'wrong' direction, and it doesn't sound right despite my wish to see the backward locomotive be something else than a dead weight :).

On a side note :
ssilk wrote:
Fri Oct 23, 2020 2:50 am
- this idea is a very good mod idea (already mentioned existing mod in this thread?)
- a way to turn around locomotives at train stops: ❤️
- a way to de-/couple wagons/locos/trains: ❤️
- shunting yards/switch yards/store items in decoupled wagons: ❤️ ❤️ ❤️ Even if I have no idea how that could work. :)
I have played with this mod when it was new : https://mods.factorio.com/mods/NiftyManiac/TrainCoupler
It was very difficult to use for the less experienced player i was. Lots of little details to plan for not much real gain.

However now that you mention it i realised there are new mods that were developped with the same idea, https://mods.factorio.com/mod/Signalized_Couplers this one is up to date with version 1. It doesn't include "turn around locomotives at train stop" (which i would appreciate seeing a picture of it IRL if it exists:) ). But i think you could do the shunting yards and store item in decoupled wagons. You could also decouple only the locomotives and add them back on the other direction :D

I'm not sure it would be helpful at solving one's problem, i think it's a way to give yourself a problem: how to make it work ? which is totally my taste for a mod :D

Edit: I forgot to mention that single and bidirectionnal train can coexist peacefully on the same network, i tend to mix those more often than not.

User avatar
KoblerMan
Fast Inserter
Fast Inserter
Posts: 184
Joined: Sat Mar 05, 2016 12:59 am
Contact:

Re: Reverse-facing locomotives contribute to total train power

Post by KoblerMan »

mmmPI wrote:
Sat Oct 24, 2020 5:20 pm
One disadvantage of the double headed trains , that could make them a bit more complicated for unexperienced player, is if you use more than one wagon, or if the different wagon don't share the same content. The train if going from A to B to A to B would end up "switching" its front loco every time, and thus if you have 2 wagon, they will alternate between being the 1rst wagon and then the 2nd. This can lead to mixing up material and/or make player choose a different system so they don't bother anticipating all the configuration if you have a 3 destination schedule it makes it worse. And if you want to add a destination it also requires planning.
This is a player design challenge, you can argue it adds complexity to the game that the player has to account for, but people like me say this is what makes the game fun. :D At that point it's up to the player how they want to solve this problem. The easiest thing to say would be "don't do this in the first place, har-har" but that doesn't really address the problem. Also, this is already something you have to consider if you use double-headed trains. Contribution of locomotive power makes absolutely no difference to this.

Potential workarounds for this:
  • Use filter inserters that will only pick up certain items, and couple this with more train conditions that make them leave if there is inactivity (so that they path around again and arrive facing the right way)
  • Filter the contents of each wagon, so that only X amount of each item type can fill the wagon, then copy those filter conditions to other wagons so that there is even resource distribution
  • Simply use single-headed trains instead
Personally I see all of these solutions as supreme jank that would reduce efficiency, and would probably just adhere to Occam's Razor for this one. Simplicity in this case means each train/station only contains one type of resource and is symmetrical, for instance a classic 1-4-1 setup.

Ultimately though I'm not here to tell you how to play the game, I'm just saying what works for me. And my original suggestion would give people more options and more design considerations to have fun with. :)
mmmPI wrote:
Sat Oct 24, 2020 5:20 pm
On the other hand i don't think signaling is more complicated it's basically the same signaling except for the stations which are themselves not as complicated as a stacker or a junction. I would even argue that it could be a simpler signaling as it forces player to use signals in a certain way but in a simple configuration. So i think it's good to familiarize with signals to at least try double headed trains.
I would argue that moving from a two-track, one-way main line to converge into a two-way single line, single-block station is more complex than just designating the station as its own one-way block section, like what you would do with a single-headed train. Also it's definitely not hard, I've done all this before as has anyone who's ever used double-headed trains. If new players have difficulty wrapping their head around these concepts then they have plenty of resources to nudge them along (see: built-in signaling tutorial). Veteran signalers will have no problem either way.
mmmPI wrote:
Sat Oct 24, 2020 5:20 pm
When talking about throughput i tend to not think about single or double headed trains. In the discussion i think it would be more precise to differenciate the station's throughput, the trains' throughput on a track, and the material's throughput on a track or overall. Station for double headed trains are a bit slower to allow a 2nd train after the 1rst one due to the impossibility for the 2nd train to be halfway in the station while the 1rst one is only halfway out. Which i think is what is called lower throughput in the previous discussion. It can be counteracted by having more stations, which wouldn't take necessarily more space than a Roro/loop station.
Again this is a player design consideration. One-way stations can be divided up into different block sections if you really want to do this just to optimize throughput. But at this point you could also just place signals everywhere in a straight line to make block sections one tile long each so that trains effectively never stop unless they're right on top of each other. I daresay, having that level of granularity is wholly unnecessary and the resource cost would make it not worth the effort. Ultimately though that's the player's choice. So, where would you draw the line? :|

Throughput in this sense has to do with acceleration. If rear-facing locomotives contributed to the total power, double-headed trains with N locomotives would get up to speed just as fast as single-headed trains with N locomotives. As it is right now, throughput and fuel efficiency is lower on double-headed trains because they accellerate more slowly. The long-distance implications are a lot more significant than how quickly you can get trains to accellerate over a very short distance, like going from a stacker into a station.
mmmPI wrote:
Sat Oct 24, 2020 5:20 pm
The overall throughput of train on a track however doesn't depend on single/multi directionnal train: If you have 1 loco for 10 wagon ( LCCCCCCCCCC) trains are slower vs 2 loco in each direction per wagon. (LLCLL). sorry for stating the obvious :D The overall throuput of material depends more on this particular aspect ( imo ) that the throughput of the stations because adding more stations is often way easier than changing existing train composition which needs the player to adapt existing station .
See? You've answered your own question! :D Having multiple stations is a design consideration, and the example you gave with locomotives to cargo wagons affects accelleration, but none of this affects top speed. So, higher thoughput would be situational depending on how far away your outposts are, what kind of stackers you have, how many stations, etc. In other words, it's player-designed. That means how competent something is depends on how players do it. Also, double-headed trains don't HAVE to only go to two-way track stations. One big advantage of double-headed trains is that they can conform to both conventions at the same time, if they're symmetrical - which again is player design.
mmmPI wrote:
Sat Oct 24, 2020 5:20 pm
boskid wrote:
Thu Oct 22, 2020 1:39 pm
There are also other problems when locomotive in wrong direction would contribute: would that mean a train with all locomotives heading into one direction should be able to find a path in opposide direction and travel entirely in reverse? What if there would be 100 locomotives in one direction and a single one in opposite?
For the first part i don't think a train should try to pathfind using the "wrong" direction; if no locomotive is facing it, just don't even try. ( player's intent would be required explicitly by placing at least 1 loco each side as other player already said).
I was convinced by the second part:
Take a train with 100 loco facing one direction. add a loco on the opposite direction. ( to show that player's intent ). Now the train is able to reach max speed on each direction.
Even if acceleration is only 60 or 80%; the way the max speed is calculated is based on the fuel type, not the number or locomotive or wagons. (according to wiki's formula : max_speed = 1.2 * fuel_top_speed_multiplier).

Sounds a bit ridiculous/gimmicky/exploit to me. I imagine seeing one such train, at full speed, in the 'wrong' direction, and it doesn't sound right despite my wish to see the backward locomotive be something else than a dead weight :).
This 100%. I had forgotten to mention the whole pathfinding thing, the whole point of double-headed trains is that you have one locomotive facing in each direction anyway. So this would work.

The more I think about it, the more boskid's argument seems to be a straw one (and I mean no disrespect with that), because none of this really matters when it comes to powering rear-facing locomotives. These are all challenges that already exist with the current state of double-headed trains. It should be possible to adjust the propulsion without changing the pathfinding behavior. :|

Also let's be honest here. In what realistic scenario is anyone ever going to see 100 locomotives in one direction and 1 in the other? If you're a player and you've done that, then I can conclude that you've failed at the game in spectacular fashion. :lol:
mmmPI wrote:
Sat Oct 24, 2020 5:20 pm
Edit: I forgot to mention that single and bidirectionnal train can coexist peacefully on the same network, i tend to mix those more often than not.
Yes, and this is one of the advantages of having both types of trains. :D
ImageImage
System Specs
OS: Windows 10 Professional 64 Bit
CPU: AMD Ryzen 5 3600X (@~3.8 gHz)
GPU: Nvidia RTX 2080
RAM: 32GB DDR4 (2400)
DRIVES: 2x 1TB NVMe SSD, 1x 6TB HDD

mmmPI
Smart Inserter
Smart Inserter
Posts: 2675
Joined: Mon Jun 20, 2016 6:10 pm
Contact:

Re: Reverse-facing locomotives contribute to total train power

Post by mmmPI »

KoblerMan wrote:
Sat Oct 24, 2020 9:41 pm
This is a player design challenge, you can argue it adds complexity to the game that the player has to account for, but people like me say this is what makes the game fun. :D At that point it's up to the player how they want to solve this problem. The easiest thing to say would be "don't do this in the first place, har-har" but that doesn't really address the problem. Also, this is already something you have to consider if you use double-headed trains. Contribution of locomotive power makes absolutely no difference to this.
It does indeed makes the game fun :). Contribution of locomotive makes no difference i agree. I just wanted to highlight that double headed trains induce more thinking than the straightforward single headed train in the design challenge. (imo)
KoblerMan wrote:
Sat Oct 24, 2020 9:41 pm
I would argue that moving from a two-track, one-way main line to converge into a two-way single line, single-block station is more complex than just designating the station as its own one-way block section, like what you would do with a single-headed train
I don't think it is more complex, in one case you have a single junction (entrance = exit), in the other one you have 2 junctions ( one entrance and one exit).

The more complex is still (imo) doing a junction for a two-track, one-way main line crossing the path of another one. Which would be required with any kind of trains ( i'm assuming past a certain scale), and imo is enough to call the difference in complexity of signaling for single headed train or multidirectionnal train null, or even in favor of the double-headed train since a station would be a learning step for junctions. It forces player to properly use chain signal at least in one case, which can help for later puzzle, such as signaling a junction.

mmmPI wrote:
Sat Oct 24, 2020 5:20 pm
When talking about throughput i tend to not think about single or double headed trains. In the discussion i think it would be more precise to differenciate the station's throughput, the trains' throughput on a track, and the material's throughput on a track or overall. Station for double headed trains are a bit slower to allow a 2nd train after the 1rst one due to the impossibility for the 2nd train to be halfway in the station while the 1rst one is only halfway out. Which i think is what is called lower throughput in the previous discussion. It can be counteracted by having more stations, which wouldn't take necessarily more space than a Roro/loop station.
KoblerMan wrote:
Sat Oct 24, 2020 9:41 pm
Again this is a player design consideration. One-way stations can be divided up into different block sections if you really want to do this just to optimize throughput. But at this point you could also just place signals everywhere in a straight line to make block sections one tile long each so that trains effectively never stop unless they're right on top of each other. I daresay, having that level of granularity is wholly unnecessary and the resource cost would make it not worth the effort. Ultimately though that's the player's choice. So, where would you draw the line? :|
It can be necessary at stations, while having it everywhere is wholly unnecessary. here is where i would draw the line :).
But you can't divide a station into many many tiny blocks to increase the station's throughput of train if it's a dead-end terminal station, like the common station for double headed trains. You can only do this for roro/loop station. It is not only player design consideration.

Even if you speed up double headed trains (or increase their acceleration ) it would still require them to exit the station entirely before another one takes their place, vs having 2 train following each other in quick succession for single headed train wich can be further exploited by spamming the station with regular signal.
KoblerMan wrote:
Sat Oct 24, 2020 9:41 pm
Throughput in this sense has to do with acceleration. If rear-facing locomotives contributed to the total power, double-headed trains with N locomotives would get up to speed just as fast as single-headed trains with N locomotives. As it is right now, throughput and fuel efficiency is lower on double-headed trains because they accellerate more slowly. The long-distance implications are a lot more significant than how quickly you can get trains to accellerate over a very short distance, like going from a stacker into a station.
I understand the balden portion as a contradiction. If what is more significant is long-distance implications, then there are no difference between single or double headed trains in terms of throughput on tracks. (not station's throughput !)

Double headed-trains do not always accelerate slower than single headed trains. (LCL accelerate faster than LCCCC). It is not single vs multi directionnal, it is train composition the main factor. (assuming same fuel).

If you were to compare the material throughput, i think it would be fair to consider 2 setups with the same footprint, and try to unload or load the same amount of material. In one case you would have a single loop station, in the other one you could have 2 dead-end station.

Meaning that throughput of trains on track is impacted the most by train composition = acceleration speed = not favouring any kind.
However , throughput of trains in a single station = granularity for loops = favouring single headed trains.

BUT in overall throughput of material, you could easily compare 1 single headed train station vs 2 double headed train station. If you do not consider station vs station, but area on the map and expected loading/unloading of material.

( this may sounds like i'm cheating to proove my point , using 2 stations vs one, but it is only a change in perspective, if you consider throughput of material on a particular area, is it better to have only 1 singleheaded station, or 2 doubleheaded station? i tend to see things this way when designing)

All this to nuance the asseration that double-headed trains have less throughput. I think it's true only with a set of initial assumptions that are not the one i choose when playing.
KoblerMan wrote:
Sat Oct 24, 2020 9:41 pm
Having multiple stations is a design consideration, and the example you gave with locomotives to cargo wagons affects accelleration, but none of this affects top speed. So, higher thoughput would be situational depending on how far away your outposts are, what kind of stackers you have, how many stations, etc. In other words, it's player-designed. That means how competent something is depends on how players do it
Yes, and in this case i hope i have explained well enough my reasonning that makes me say double-headed train don't have 'lower throughput'. It depends on what throughput you are considering, trains in network, trains in stations, material per area, material per station .... and more importantly how they are implemented.

Maybe my wording was not precise enough, in my native langage, throughput, flow rate ,bandwidth and others are often translated by the same word: "débit" which makes it difficult to grasp the subtle nuances and use the appropriate word in english :(

number or train crossing a rail in a time period, number of trains entering and leaving a station in a time period, number of material being loaded/unloaded in a time period and/or per area, : for all of those i used the word throughput. Only the bold one is favouring single-headed trains, but the station being bigger, if you include "per area" you don't necesarily conclude to the lower "throughput" depending on your designs, not depending on the intrinsic nature of double-headed train.
KoblerMan wrote:
Sat Oct 24, 2020 9:41 pm
mmmPI wrote:
Sat Oct 24, 2020 5:20 pm
Take a train with 100 loco facing one direction. add a loco on the opposite direction. ( to show that player's intent ). Now the train is able to reach max speed on each direction.
Also let's be honest here. In what realistic scenario is anyone ever going to see 100 locomotives in one direction and 1 in the other? If you're a player and you've done that, then I can conclude that you've failed at the game in spectacular fashion. :lol:
Well i was not precise here, the current behavior already allows one such trains to reach max speed on either direction. But one will be very slow :). So slow that it wouldn't even make sense to try.

The proposed change, would make it way faster to reach the max speed in the "wrong" direction so much that the acceleration could seem as fast in either way for human eyes.

Wouldn't that give incentive to players to do what you consider failing at the game in a spectacular fashion ?:D

More realistically, and taking into account the fact that in a 2-station-schedule trains flip every time their front and back. Here is how i would attempt to abuse the new rule :

Make your trains LLLLCCCCCCCCL (4-8-1) for example, and this way you have them leaving the main base with 4 loco pushing at full power and 1 at 60% acceleration,good for busy area. When they leave an outpost they would be much slower to reach max speed due to having only 1 loco pushing at full power and 4 at 60% but it'd be in desertic area so no problemo :).

To reformulate previous question in a less provocative way : Wouldn't that be an undesired side effect of the proposed change ?

(side question : what happens if the lone loco run out of fuel ? all 4 should stop contributing ?)

I also want to nuance the assertion saying that fuel consumption is favouring single headed train. i'll try to make it quick. The longer the distance without braking the easier it is to neglect the cost in fuel for carrying some locos as dead weight. ( you only increase travel time by a short margin in %). Instead one could see the rear-facing locos as carrying the fuel for the way back.

Or in other words LCCCL train have almost twice as much autonomy as LCCC train. It takes 2016.67 second to burn a single uranium fuel cell. with the LCCCL train you can go 2K second in one direction, AND 2K second in the other one after that. ( or 8K and 8K if you fill up locos with fuel). Sure in 2K second you will cover more distance with LCCC than with LCCCL due to the lower acceleration at first. but it tend to become negligible over long distance at full speed. and i think the tradeoff doesn't bring shame to bi-directionnal train. (they are my favourite already and i don't think they NEED to receive a buff even though i wouldn't mind :D).

LLLC trains, is less fuel efficient than LCCCCC trains once max speed is reached. because you burn 3x times as much fuel to maintain the same speed. Or am i missing something ?

One smaller buff that could be thought off would be to use backward locomotives as counting toward braking power instead of the proposed acceleration. it has less impact but would make more sense in my eyes :)

User avatar
ptx0
Smart Inserter
Smart Inserter
Posts: 1507
Joined: Wed Jan 01, 2020 7:16 pm
Contact:

Re: Reverse-facing locomotives contribute to total train power

Post by ptx0 »

after that comment ^ i think this should be moved back into Ideas and Suggestions and out of Not Implemented.

User avatar
NotRexButCaesar
Smart Inserter
Smart Inserter
Posts: 1120
Joined: Sun Feb 16, 2020 12:47 am
Contact:

Re: Reverse-facing locomotives contribute to total train power

Post by NotRexButCaesar »

ptx0 wrote:
Sun Oct 25, 2020 4:44 pm
after that comment ^ i think this should be moved back into Ideas and Suggestions and out of Not Implemented.
Movement of posts is dependent on developer input, not community opinion.
—Crevez, chiens, si vous n'étes pas contents!

User avatar
ptx0
Smart Inserter
Smart Inserter
Posts: 1507
Joined: Wed Jan 01, 2020 7:16 pm
Contact:

Re: Reverse-facing locomotives contribute to total train power

Post by ptx0 »

AmericanPatriot wrote:
Sun Oct 25, 2020 5:11 pm
ptx0 wrote:
Sun Oct 25, 2020 4:44 pm
after that comment ^ i think this should be moved back into Ideas and Suggestions and out of Not Implemented.
Movement of posts is dependent on developer input, not community opinion.
actually the moderators are a part of the community and not developers, and they move tickets around. thanks for your useless feedback.

User avatar
NotRexButCaesar
Smart Inserter
Smart Inserter
Posts: 1120
Joined: Sun Feb 16, 2020 12:47 am
Contact:

Re: Reverse-facing locomotives contribute to total train power

Post by NotRexButCaesar »

ptx0 wrote:
Sun Oct 25, 2020 5:12 pm
actually the moderators are a part of the community and not developers, and they move tickets around. thanks for your useless feedback.
But they don’t just move posts because they feel like it, they move posts when a developer gives input on something (in the ideas and suggestions subforum).

Your sarcasm is not useful or appreciated.
—Crevez, chiens, si vous n'étes pas contents!

Squelch
Filter Inserter
Filter Inserter
Posts: 346
Joined: Sat Apr 23, 2016 5:31 pm
Contact:

Re: Reverse-facing locomotives contribute to total train power

Post by Squelch »

I ran some experiments on my test map and found that a train comprising of mainly forward facing locos, and one rearward facing loco could be another option.

Travelling left to right

____LL>CCCCCCCCCCFFFFFFF<L-LL>_____

The single rearward facing loco allows the train - albeit slowly - to reverse into a turning head, and then back out to the next destination. The whole layout is slightly smaller than a loop, and the top speed/acceleration does not appear to be impeded by the dead weight too much. It still arrives with the "correct" orientation at stations, and can access branch lines as simply as any other double headed train. This seems to have some of the best of both worlds combined.

I'm not sure I'd use such a contraption, but it might be useful for those distant locations served by a single track.

User avatar
KoblerMan
Fast Inserter
Fast Inserter
Posts: 184
Joined: Sat Mar 05, 2016 12:59 am
Contact:

Re: Reverse-facing locomotives contribute to total train power

Post by KoblerMan »

mmmPI wrote:
Sun Oct 25, 2020 9:57 am
Maybe my wording was not precise enough, in my native langage, throughput, flow rate ,bandwidth and others are often translated by the same word: "débit" which makes it difficult to grasp the subtle nuances and use the appropriate word in english :(
Well lucky for you, I'm a bit of an english nerd. :geek:

I would describe the difference as this:
  • Throughput is how much of something moves through any particular medium over an amount of time, and is the most appropriate term here. Measuring throughput is usually what you do with most systems in Factorio, because ultimately it's the throughput that determines how fast you can consume the raw resources and produce completed products. The all-famous measurement of SPM (science per minute) is an example of this, and having high SPM is a pretty good benchmark test for how powerful your megabase is post-rocket launch.
  • Flow rate is how fast something moves through any particular medium over an amount of time, but does not quantify how much (you can have high flow rate but still end up with lower throughput, IF the amount of stuff going through is miniscule). To represent this, let's say you have an LCC configuration where the locomotives move very fast, and get offloaded quickly at each station and continue moving along, but they drop off less material in total. As soon as the ore hits the chests, it gets offloaded onto the belts which carry it away. The trains might have a high flow rate through the station, however the belts become empty again just as quickly as it took the train to stop. In contrast you might also have a different station with an LCCCCL convention, where the trains are slower to flow in but because they each carry twice as much more material but take the same amount of time to unload, the belts might remain full the whole time, and thus have a higher throughput. So a lower flow rate of trains in this sense might not be better than a higher throughput of materials in general.
  • Bandwidth is how much of something can move through any particular medium at the SAME time. Higher bandwidth for instance would mean your trains might have a two-way 4-track network instead of a two-way 2-track one. The bandwidth is how many trains can travel on that same network going from point A to point B. However, high bandwidth in this case does not always guarantee high throughput. Ultimately although 4-track networks have higher bandwidth, it is not worth it for many people because it's a case of diminishing returns - you get less than double the throughput as a 2-track network, even when doubling the bandwidth. Another instance of having higher bandwidth would be putting 8 belts of iron plates on your main bus instead of the standard 4. But again the throughput depends on how well you can satisfy the bandwidth.
When I use the term "throughput" it's more of a colloquial, abstract use - how much resource can you get from point A to point B by train? Technically all three of these things are considered in the player's design. But they can all be quantified. And to be honest, I should have done a better job of quantifying them for the purposes of this suggestion. I got in trouble for that last year on my suggestion thread where I asked for a rebalancing of nuclear power. :oops:
mmmPI wrote:
Sun Oct 25, 2020 9:57 am
KoblerMan wrote:
Sat Oct 24, 2020 9:41 pm
mmmPI wrote:
Sat Oct 24, 2020 5:20 pm
Take a train with 100 loco facing one direction. add a loco on the opposite direction. ( to show that player's intent ). Now the train is able to reach max speed on each direction.
Also let's be honest here. In what realistic scenario is anyone ever going to see 100 locomotives in one direction and 1 in the other? If you're a player and you've done that, then I can conclude that you've failed at the game in spectacular fashion. :lol:
Well i was not precise here, the current behavior already allows one such trains to reach max speed on either direction. But one will be very slow :). So slow that it wouldn't even make sense to try.

The proposed change, would make it way faster to reach the max speed in the "wrong" direction so much that the acceleration could seem as fast in either way for human eyes.

Wouldn't that give incentive to players to do what you consider failing at the game in a spectacular fashion ?:D

[...]

To reformulate previous question in a less provocative way : Wouldn't that be an undesired side effect of the proposed change ?
Generally speaking, I'm inclined to say no. Personally I have the opposite mindset on this, but I realize that this is all just a matter of opinion, what each player believes is realistic and less harmful to their suspension of disbelief.

My justification for saying no is because I would find it much harder to believe that any engines in a configuration are dead weight and not contributing power, which is the case now. I am of the opinion that those 100 locomotives in the back with just 1 in front would reflect this as well. (See also: push-pull train mechanics IRL)

Ultimately though, this suggestion brings up a potential change in balance with the proposed behavior adjustment, so it's less about suspension of disbelief and more about practicality and gameplay. :)
mmmPI wrote:
Sun Oct 25, 2020 9:57 am
(side question : what happens if the lone loco run out of fuel ? all 4 should stop contributing ?)
I don't think so, they should be able to continue chugging along regardless. Just like in a single-headed config with multiple locomotives, if one of them runs out of fuel the other will keep going.
AmericanPatriot wrote:
Sun Oct 25, 2020 5:15 pm
ptx0 wrote:
Sun Oct 25, 2020 5:12 pm
AmericanPatriot wrote:
Sun Oct 25, 2020 5:11 pm
ptx0 wrote:
Sun Oct 25, 2020 4:44 pm
after that comment ^ i think this should be moved back into Ideas and Suggestions and out of Not Implemented.
Movement of posts is dependent on developer input, not community opinion.
actually the moderators are a part of the community and not developers, and they move tickets around. thanks for your useless feedback.
But they don’t just move posts because they feel like it, they move posts when a developer gives input on something (in the ideas and suggestions subforum).

Your sarcasm is not useful or appreciated.
Can you guys relax please? :shock:

It's up to the moderators to interpret these posts and determine in good faith what should be done with them. In this case Koub interpreted boskid's criticism of this suggestion as a "no" to it being implemented and took action on that. To be fair I did as well, but ultimately it is up to the mods what to do with these posts. In rare cases devs will override mod decisions though (I've seen it happen before).

I don't always agree with everything the forum mods do either so don't think I'm being an apologist for that. But I'm also not here to start WW3 so I won't extrapolate on that point. :)
ImageImage
System Specs
OS: Windows 10 Professional 64 Bit
CPU: AMD Ryzen 5 3600X (@~3.8 gHz)
GPU: Nvidia RTX 2080
RAM: 32GB DDR4 (2400)
DRIVES: 2x 1TB NVMe SSD, 1x 6TB HDD

Koub
Global Moderator
Global Moderator
Posts: 7175
Joined: Fri May 30, 2014 8:54 am
Contact:

Re: Reverse-facing locomotives contribute to total train power

Post by Koub »

[Koub] I might not look like I do, but I actually read all the new posts several times a day, so I've been watching what's been written here. I like letting people moderate themselves, what they most of the time do, because I'm a lazy sloth convinced most people are sensible enough to behave by themselves. When time proves me wrong, a simple reminder is usually enough. This is such a reminder. Please behave.
I've moved this to Outdated because the boskid seemed pretty clear that it would be a "no". If any of the devs think otherwise, they all have the privileges to move it back to Ideas and suggestions, or a simple post here asking to move it back will be enough for me.

mmmPI wrote:
Sun Oct 25, 2020 9:57 am
Maybe my wording was not precise enough, in my native langage, throughput, flow rate ,bandwidth and others are often translated by the same word: "débit" which makes it difficult to grasp the subtle nuances and use the appropriate word in english :(
In case it might be of some help :
Throughput = débit (plutôt pour des object)
Flow rate = débit (plutôt pour un fluide)
Bandwidth = bande passante (the maximum throughput/flowrate the system can achieve). If you need more output than that, your are throughput limited.
Koub - Please consider English is not my native language.

mmmPI
Smart Inserter
Smart Inserter
Posts: 2675
Joined: Mon Jun 20, 2016 6:10 pm
Contact:

Re: Reverse-facing locomotives contribute to total train power

Post by mmmPI »

Koub wrote:
Mon Oct 26, 2020 7:03 am
In case it might be of some help :
Throughput = débit (plutôt pour des object)
Flow rate = débit (plutôt pour un fluide)
Bandwidth = bande passante (the maximum throughput/flowrate the system can achieve). If you need more output than that, your are throughput limited.
mmm to be honest, in this direction i can translate, since it's "débit" everytime :D . (even bandwidth is often translated into "débit" mainly in internet coloquial or advertising langage).

the unit in which the variable is expressed i think is the most significant to translate the other way around.

Throughput is number of " train per hour , per day , per week "
flow rate is " distance per time " ( speed of train or average )
bandwidth is a maximum number of " train at any instant T"

For car traffic analysis, they do something like a multiplication between "flow rate" (average speed of cars km/h) and "amount of car in 1KM of road (car/km)" for example. to calculate throughput of a portion of road.

if you multiply both , the "km" unit disappear and you are left with a variable in car/h. Which would be the throughput.

In the analogy the amount of car in 1km is the % of bandwidth used right ?

Using those more precise term it's easier for me to try and translate what i read in french about traffic management. Thanks for helping me improve my english :)

Post Reply

Return to “Outdated/Not implemented”