Contest: Longest train based nuclear reactor

Power Plants, Energy Storage and Reliable Energy Supply. All about efficient energy production. Turning parts of your factory off. Reliable and self-repairing energy.
quyxkh
Smart Inserter
Smart Inserter
Posts: 1027
Joined: Sun May 08, 2016 9:01 am
Contact:

Re: Contest: Longest train based nuclear reactor

Post by quyxkh »

mrvn wrote: ↑
Sun Oct 31, 2021 10:12 pm
quyxkh wrote: ↑
Sun Oct 31, 2021 9:16 pm
If the direct-to-wagon steam unload requirement is relaxed ...
But it isn't for the very reason of that design. There is no challenge there as you can tile that to any train length you want. It can fill any length of train.
Okay. The doubleheaded layout is fatally flawed, you can't get trains in and out fast enough with the unfueled-nuc-plant heat pipes you were using in hot-plant rows and heat-pipe columns. Instead, turn the columns sideways, an exchanger-pump pair tiles nicely with the nuc-plant connectors, quick sketch:
pic
Then you could doubleside the steam loading and supply the water separately, unload from the train at an earlier stop, and get yourself, what, 2s load time + roro intervals? which neatly cuts your locomotive requirement in half as well, so even quicker cycling.

mmmPI
Smart Inserter
Smart Inserter
Posts: 2674
Joined: Mon Jun 20, 2016 6:10 pm
Contact:

Re: Contest: Longest train based nuclear reactor

Post by mmmPI »

quyxkh wrote: ↑
Sun Oct 31, 2021 10:05 pm
That's what I was referring to. It meets my definition of unloading, I don't understand what you mean.
it's a bit absurd and the source of my argument with mrvn, the method from the rules only impose 3 pieces for the LOADING station.

the UNLOADING station hasn't been described, you calculated in your previous post how much time would a train need to clear the area it occupy when loading because you assume the train was to move but it is not clear in the rules of the contest if this amount of time is measured, and if it is the method of measurement is not known.

You did a sensible design assuming the steam was to be transported with the ability to tile it to choose the size of the train, i think it's great x) This allow the fun mechanism of triggering refuel by counting trains, that would be like measuring steam tanks if you had used steam tanks.

And the method you use to evaluate the amount of steam to be carried away is also making sense, steam/s/wagon for one ( i may be biais) as it allow to compare different size of trains which is more sensible than trying to have the longest since the rules do not explictly circumvent to possibilities of making one of any desired lengh.

Measuring the time it takes for the train to clear itself makes sense, but if the method of measurement for steam/s/wagon or just the vague "steam/s" was known there would be way to reduce that time. For example, by having a train that act like a conveyer belt, if your train is 4500 reactor long, you may need to only move 5% of it to unload one portion, then again 5% to unload a different area of the train at the same unload, and so on, and there would be a train right behind the first one so there would be no downtime. The giga long train could then take all its time to turn around and replace. This would also mean that each time the train stops, you would need to load like 5% of the wagon, then next stop 5% next stop 5%, so that when a portion is facing the unload it has 100% load ready to be unloaded all at once. ( abusing the internal buffer of exchanger with very short break between loading for moving so as to reduce their real down-time to 0 ) (instead of using tanks to hold the production that was made during train swap).

Those kind of weird things would allow maximizing "steam/s" as it is written in the rules and depending on the not-yet-disclose-not-even-in-his-own-head method of measurement but you did maximize "steam/s/wagon" in real conditions which is smarter x).

What you said about the rule is nonetheless true i think, it is required to make something useful and not just a proof that you don't need storage tank. The LOADING method need to be relaxed and open to interpretations and creativity !

quyxkh wrote: ↑
Sun Oct 31, 2021 10:05 pm
A little quick belt math will show how that layout can't feed more, but the real reason is, 4500 reactors in a row is already 22.5km and that's probably more than enough for anything I'll ever build. I'll leave going bigger for anyone who wants to build a map that can find an interesting use for more than 1.4TW. Got any examples?
I had not realised you mathed the throughput of the blue belts full of fuel. On my sheets the requirement for infinite array for fuel is dealt with having the train bringing it with a cargo wagon inserted inbetween the fluidwagon.

For your design you could always get rid of a column of exchanger to have room for a vertical blue belt refilling the horizontal one every 4500 reactor, or maybe every 9000 since you could have 2 vertical blue belt, turning horizontal, but one to the left and one to the right, every 45 km, sure little overheating but small cost to reach infinity x) !

With 1.4TW you could do pixel art with beacon visible from map view at the smallest zoom with a decent definition i think but i would never had thought of that if you didn't asked x)
Last edited by mmmPI on Sun Oct 31, 2021 11:13 pm, edited 1 time in total.

mmmPI
Smart Inserter
Smart Inserter
Posts: 2674
Joined: Mon Jun 20, 2016 6:10 pm
Contact:

Re: Contest: Longest train based nuclear reactor

Post by mmmPI »

mrvn wrote: ↑
Sun Oct 31, 2021 10:08 pm
mmmPI wrote: ↑
Sun Oct 31, 2021 8:58 pm
haha that' so stupid, if i make a 90000000 wagon long, then there's no other design to compare with, i just have to put 1 or 2 steam here and there and it's the longest train. Or you gonna try saying you already explained it's forbidden ?
Already did. Say I've already explained that is.
no you didn't you just said my wagon with 82% filled count as 100% filled and be judged as a 4 wagon train.

but then if they are filled with 0.1%, then it's different because you think it's me trying to break a rule that you didn't explain, nor write explicitly, not even realized was needed before i told you my train were 82% filled after you said there were counted as 4 wagon, which is why you keep avoiding answering properly.

This rule missing is what would make the difference between my design with 82% filled train as i use it for more precise refuel or with 100% filled train as i could have posted for higher steam/s/wagon.

You are a lier saying you already explained you keep avoiding explaining how you would measure steam/s

mmmPI
Smart Inserter
Smart Inserter
Posts: 2674
Joined: Mon Jun 20, 2016 6:10 pm
Contact:

Re: Contest: Longest train based nuclear reactor

Post by mmmPI »

quyxkh wrote: ↑
Sun Oct 31, 2021 10:56 pm
Then you could doubleside the steam loading and supply the water separately, unload from the train at an earlier stop, and get yourself, what, 2s load time + roro intervals? which neatly cuts your locomotive requirement in half as well, so even quicker cycling.
oh yeah mrvn was told that already in the first page, it did not made him realized that using this could lead to any size of reactor if you want, and then making a contest for the longest train doesn't make sense. Only steam/s would count, but for that there need to be a valid way to measure, since measuring steam at loading doesn't take in account train moving time it should be measure at a certain distance, which i told mrvn, but according to him there is no need for a distance, it doesn't matter since every distance would yied the same steam/s as he "explained".

mrvn
Smart Inserter
Smart Inserter
Posts: 5682
Joined: Mon Sep 05, 2016 9:10 am
Contact:

Re: Contest: Longest train based nuclear reactor

Post by mrvn »

mmmPI wrote: ↑
Sun Oct 31, 2021 11:00 pm
quyxkh wrote: ↑
Sun Oct 31, 2021 10:05 pm
That's what I was referring to. It meets my definition of unloading, I don't understand what you mean.
it's a bit absurd and the source of my argument with mrvn, the method from the rules only impose 3 pieces for the LOADING station.

the UNLOADING station hasn't been described, you calculated in your previous post how much time would a train need to clear the area it occupy when loading because you assume the train was to move but it is not clear in the rules of the contest if this amount of time is measured, and if it is the method of measurement is not known.
I you need a special unloading station then you can always provide one. It's left unspecified since it doesn't matter. Just assume the unloading isn't the bottleneck. Just think it's a magic mod that takes the trains, counts them and the steam and deconstructs them in one tick when the train reserves the block the magic stuff is in.
mmmPI wrote: ↑
Sun Oct 31, 2021 11:00 pm
You did a sensible design assuming the steam was to be transported with the ability to tile it to choose the size of the train, i think it's great x) This allow the fun mechanism of triggering refuel by counting trains, that would be like measuring steam tanks if you had used steam tanks.

And the method you use to evaluate the amount of steam to be carried away is also making sense, steam/s/wagon for one ( i may be biais) as it allow to compare different size of trains which is more sensible than trying to have the longest since the rules do not explictly circumvent to possibilities of making one of any desired lengh.

Measuring the time it takes for the train to clear itself makes sense, but if the method of measurement for steam/s/wagon or just the vague "steam/s" was known there would be way to reduce that time. For example, by having a train that act like a conveyer belt, if your train is 4500 reactor long, you may need to only move 5% of it to unload one portion, then again 5% to unload a different area of the train at the same unload, and so on, and there would be a train right behind the first one so there would be no downtime. The giga long train could then take all its time to turn around and replace. This would also mean that each time the train stops, you would need to load like 5% of the wagon, then next stop 5% next stop 5%, so that when a portion is facing the unload it has 100% load ready to be unloaded all at once. ( abusing the internal buffer of exchanger with very short break between loading for moving so as to reduce their real down-time to 0 ) (instead of using tanks to hold the production that was made during train swap).
Next you will make the train infinite long and always move N wagons instead of 5%.

If the train only moves 5% then only 5% get loaded or unloaded at any time. So I would count that at 5% of it's size for comparison with other reactors. It's the same as cutting the train into separate trains every 5% but you have less hassle with avoiding collisions with a conveyer belt train. Only way to get a sensible count.

Think about the throughput of a conveyer belt. The length of a conveyer belt does not determine the throughput, only the speed of the belt matters. Would you ever say a 10000 tile transport belt moves 80000 ore? No, it would be 15/30/45 ore/s no matter how long it is.
mmmPI wrote: ↑
Sun Oct 31, 2021 11:00 pm
Those kind of weird things would allow maximizing "steam/s" as it is written in the rules and depending on the not-yet-disclose-not-even-in-his-own-head method of measurement but you did maximize "steam/s/wagon" in real conditions which is smarter x).

What you said about the rule is nonetheless true i think, it is required to make something useful and not just a proof that you don't need storage tank. The LOADING method need to be relaxed and open to interpretations and creativity !
It's scary you think to know what is in my head combined how wrong you always are about what I said or mean. I think I've been clear enough about the measurement method. It's counting wagons and as deciding factor steam/s for comparable designs. Rule breaking or other special cases that make the train count just plain wrong gets your train size adjusted to something sensible.

If the rules get any more relaxed then there is only one winner: Unlimited long train. The standard tileable reactor with the turbines repalced by loading stations with a train as long as you want to make it does that. There is no contest there.

mmmPI
Smart Inserter
Smart Inserter
Posts: 2674
Joined: Mon Jun 20, 2016 6:10 pm
Contact:

Re: Contest: Longest train based nuclear reactor

Post by mmmPI »

mrvn wrote: ↑
Mon Nov 01, 2021 3:20 am
I you need a special unloading station then you can always provide one. It's left unspecified since it doesn't matter. Just assume the unloading isn't the bottleneck. Just think it's a magic mod that takes the trains, counts them and the steam and deconstructs them in one tick when the train reserves the block the magic stuff is in.
mrvn wrote: ↑
Mon Nov 01, 2021 3:20 am
If the train only moves 5% then only 5% get loaded or unloaded at any time.
That is false :

ABCDE are wagons, A gets loaded 10 second by 3 heat exchanger, represent 60 steam, then train move and A is now where B was, gets loaded 10 second, 120 steam total, then moves again to 180 steam the move again to 240 steam, when moves again to 300 steam, now comes the unload: all at once since the unload is not the bottleneck, why would it be as slow as 3 heat exchanger 60 /second ? so the train moves 5% of its lengh but the 5 % of its lengh that is unloaded is unloaded fully.

This means if you think about how much steam the train contains and where it is located, that you load every part a little every time, but your unload only the part that are the most filled and all at once.


When the train unload its last part, it unload 100% of its content not 5% easy to understand no ?. And when the train unload its first part, it's actually more than 5% of the total steam of the train even if it represent 5% of its lengh.

mrvn wrote: ↑
Mon Nov 01, 2021 3:20 am
Think about the throughput of a conveyer belt. The length of a conveyer belt does not determine the throughput, only the speed of the belt matters. Would you ever say a 10000 tile transport belt moves 80000 ore? No, it would be 15/30/45 ore/s no matter how long it is.
You're correct, that's probably why DaveMcW calculated the amount of wagon that 1 rail track could carry. But then you said you had 2 rails tracks hahahaha ! like it's going to fundamentally change the math instead of just x2.

this holds true as a comparaison because the belt filled up has the same density everywhere, not the train wagon in my example if there is something else than a magic imaginary non-bottleneck unload.

mrvn wrote: ↑
Mon Nov 01, 2021 3:20 am
It's scary you think to know what is in my head combined how wrong you always are about what I said or mean. I think I've been clear enough about the measurement method. It's counting wagons and as deciding factor steam/s for comparable designs. Rule breaking or other special cases that make the train count just plain wrong gets your train size adjusted to something sensible.


mrvn wrote: ↑
Sat Oct 30, 2021 3:49 pm
mmmPI wrote: ↑
Sat Oct 30, 2021 2:48 pm
i wish you find a way to be more explicit.
like 1 pump per wagon => ok not considered elongated artificially not cheating,


It depends but I probably would consider it elongating the train. Because you can fit at least 2 heat exchangers per wagon. Only reason not to use 2 is probably to make the train longer. Instead of making a longer train so you can move more steam.


This is how you answer when asked, and then after tons of useless argument you're like " i think i've been clear enough" no, you've just spammed a lot to avoid explaining clearly things

mrvn wrote: ↑
Mon Nov 01, 2021 3:20 am
If the rules get any more relaxed then there is only one winner: Unlimited long train. The standard tileable reactor with the turbines repalced by loading stations with a train as long as you want to make it does that. There is no contest there.
steam/s is not a method of measurement, it's what is being measured, for which you need a method, that is something practical like a protocol example:

-production graph average in 10 minutes 1 hours after the begining of the test.
-Number of steam passing beyond a certain point on the map coming from beyond another point in X amount of time counting how much wagons and how many in each wagon.
-how much electricity produced in an area accessible only for trains loaded with steam on average in the first 2 hours of opening the test.

not just a hypothetical steam/S magic bottleneck that anyone and his cousin have to decide to interpret how it functions and hope you will interpret it the same.

The rules as of now allow unlimited long train formally, you keep saying "no it's cheating" when you realize a design not explictly excluded by any rules is infinite. There is no contest already since only worse design that yours are accepted others are considered cheating .

mrvn
Smart Inserter
Smart Inserter
Posts: 5682
Joined: Mon Sep 05, 2016 9:10 am
Contact:

Re: Contest: Longest train based nuclear reactor

Post by mrvn »

mmmPI wrote: ↑
Mon Nov 01, 2021 3:59 am
mrvn wrote: ↑
Mon Nov 01, 2021 3:20 am
I you need a special unloading station then you can always provide one. It's left unspecified since it doesn't matter. Just assume the unloading isn't the bottleneck. Just think it's a magic mod that takes the trains, counts them and the steam and deconstructs them in one tick when the train reserves the block the magic stuff is in.
mrvn wrote: ↑
Mon Nov 01, 2021 3:20 am
If the train only moves 5% then only 5% get loaded or unloaded at any time.
That is false :

ABCDE are wagons, A gets loaded 10 second by 3 heat exchanger, represent 60 steam, then train move and A is now where B was, gets loaded 10 second, 120 steam total, then moves again to 180 steam the move again to 240 steam, when moves again to 300 steam, now comes the unload: all at once since the unload is not the bottleneck, why would it be as slow as 3 heat exchanger 60 /second ? so the train moves 5% of its lengh but the 5 % of its lengh that is unloaded is unloaded fully.

This means if you think about how much steam the train contains and where it is located, that you load every part a little every time, but your unload only the part that are the most filled and all at once.


When the train unload its last part, it unload 100% of its content not 5% easy to understand no ?. And when the train unload its first part, it's actually more than 5% of the total steam of the train even if it represent 5% of its lengh.
Well, that is not what a conveyer train looks like normaly. That would be more LABCDELABCDELABCDELABCDELABCDELABCDE where the train goes around in a loop and the last E nearly touches the first Locomotive. And you load/unload one set of ABCDE any time the train stops. That will be as slow as a simple LABCDE train except the rail congestion is eliminated.

And that is AGAIN the problem with your made up hypothetical designs. What you describe and what other people understand from your description are 2 completely different things. Maybe you describe it wrong, maybe the other person doesn't understand. It really doesn't matter. Each one thinks of a different design and you can't argue about it then.

What you describe now is just a fancy and inefficient way to load the train. Since you have heat exchangers for all 5 fluid wagons you can just drive in fully and load all 5 at the same time. There is 0 advantage of stopping along the way and you are just slowing down loading. If you want to do that then that is your problem. Someone who modifies your design so the train simply drives to the last loading stop will have you beat. The train moves faster without having to slow down all the time and the pumps don't have to connect 5 times before activating and all 15 pumps will work to fill the train, not just 50% or so on average. Saves a lot of extra delay.
mmmPI wrote: ↑
Mon Nov 01, 2021 3:59 am
mrvn wrote: ↑
Mon Nov 01, 2021 3:20 am
Think about the throughput of a conveyer belt. The length of a conveyer belt does not determine the throughput, only the speed of the belt matters. Would you ever say a 10000 tile transport belt moves 80000 ore? No, it would be 15/30/45 ore/s no matter how long it is.
You're correct, that's probably why DaveMcW calculated the amount of wagon that 1 rail track could carry. But then you said you had 2 rails tracks hahahaha ! like it's going to fundamentally change the math instead of just x2.

this holds true as a comparaison because the belt filled up has the same density everywhere, not the train wagon in my example if there is something else than a magic imaginary non-bottleneck unload.
And this has just become obsolete since your design isn't a conveyer belt design.
mmmPI wrote: ↑
Mon Nov 01, 2021 3:59 am
mrvn wrote: ↑
Mon Nov 01, 2021 3:20 am
It's scary you think to know what is in my head combined how wrong you always are about what I said or mean. I think I've been clear enough about the measurement method. It's counting wagons and as deciding factor steam/s for comparable designs. Rule breaking or other special cases that make the train count just plain wrong gets your train size adjusted to something sensible.


mrvn wrote: ↑
Sat Oct 30, 2021 3:49 pm
mmmPI wrote: ↑
Sat Oct 30, 2021 2:48 pm
i wish you find a way to be more explicit.
like 1 pump per wagon => ok not considered elongated artificially not cheating,


It depends but I probably would consider it elongating the train. Because you can fit at least 2 heat exchangers per wagon. Only reason not to use 2 is probably to make the train longer. Instead of making a longer train so you can move more steam.


This is how you answer when asked, and then after tons of useless argument you're like " i think i've been clear enough" no, you've just spammed a lot to avoid explaining clearly things

mrvn wrote: ↑
Mon Nov 01, 2021 3:20 am
If the rules get any more relaxed then there is only one winner: Unlimited long train. The standard tileable reactor with the turbines repalced by loading stations with a train as long as you want to make it does that. There is no contest there.
steam/s is not a method of measurement, it's what is being measured, for which you need a method, that is something practical like a protocol example:

-production graph average in 10 minutes 1 hours after the begining of the test.
-Number of steam passing beyond a certain point on the map coming from beyond another point in X amount of time counting how much wagons and how many in each wagon.
-how much electricity produced in an area accessible only for trains loaded with steam on average in the first 2 hours of opening the test.

not just a hypothetical steam/S magic bottleneck that anyone and his cousin have to decide to interpret how it functions and hope you will interpret it the same.

The rules as of now allow unlimited long train formally, you keep saying "no it's cheating" when you realize a design not explictly excluded by any rules is infinite. There is no contest already since only worse design that yours are accepted others are considered cheating .
mrvn wrote: ↑
Sat Oct 30, 2021 1:51 pm
Just for you if the reactor has N nuclear reactor entities I will measure (N % 100)% of the distance between reactor and the hypothetical consumer before said consumer.
And to be more clear I count fluid wagons as they pass with steam content and time between trains and fluid wagon count of the train. This then gives steam/s transported or steam/s/wagon if you like. Nothing hypothetical about that.

Apparently you still think it makes a difference if I measure trains leaving the reactor, trains on the way, trains arriving at unload, steam unloaded, electricity created from steam. Maybe even weather I use 100 turbines running 100% load or 200 turbines running 50% load when the average steam produces is good for 100 turbines. There is no loss of energy in factorio other than reactors overheating. No matter where you measure the result will always be equivalent. If one reactor is better counting the steam leaving the reactor it will also be better at electricity produced. It's all the same.

And obviously I'm not going to take any measurement at the beginning of the test. That goes without saying. Reactors need to heat up to a working temperature and all that stuff. Which was already one of the arguments why an infinite train wouldn't count. It fills the train during warmup and stops. If I don't measure the first train or the first hour or any other sensible way to define the warmup phase the infinite train will always produce 0.

mmmPI
Smart Inserter
Smart Inserter
Posts: 2674
Joined: Mon Jun 20, 2016 6:10 pm
Contact:

Re: Contest: Longest train based nuclear reactor

Post by mmmPI »

mrvn wrote: ↑
Mon Nov 01, 2021 3:05 pm
And that is AGAIN the problem with your made up hypothetical designs. What you describe and what other people understand from your description are 2 completely different things. Maybe you describe it wrong, maybe the other person doesn't understand. It really doesn't matter. Each one thinks of a different design and you can't argue about it then.

This is why i'm asking for the method, otherwise i'd spend hours making a design you'd just discard because of misunderstanding in the rules ending in you disqualifying it due to what be like arbitrary reason/interpretation of the rule to me.

mrvn wrote: ↑
Mon Nov 01, 2021 3:05 pm
What you describe now is just a fancy and inefficient way to load the train. Since you have heat exchangers for all 5 fluid wagons you can just drive in fully and load all 5 at the same time. There is 0 advantage of stopping along the way and you are just slowing down loading.


And this has just become obsolete since your design isn't a conveyer belt design.
the advantage of stopping along the way allows for not having to wait for 15 or 50 or 100 loco when it's time but instead have a train LCCCLCCCLCCCLCCCLCCC; This means the time between unload in averaged.

No it's not obsolete, you can think of it as a belt maximum troughput, and a belt actual throughput. The belt maximum throughput is the same all along the way, not the actual throughput this depend on where you load. Depending on where you look, on the distance , on how many producers on the left side/right side.

You can saturate a belt with 4 stack inserter working 100% of time, and then have a long belt maxed out.
Or you can saturate a belt with 120 yellow inserter working some % of time. This will take much more space, the belt maxed out will be so only for a few tile if you take 70 tiles measure, 60 inserter on each side, only the 10 last belt have to carry the whole throughput.

More distance allow to earn more time to load given the max throughput, and the max capacity of a single unit. (do you move 1 by 1 wagon to average ? do you move 5 by 5 ? do you move 100000 by 1000000 ?) the more distance you take, the more time you have to accelerate, if there is a fix distance, then there is an optimal.

mrvn wrote: ↑
Mon Nov 01, 2021 3:05 pm
And to be more clear I count fluid wagons as they pass with steam content and time between trains and fluid wagon count of the train. This then gives steam/s transported or steam/s/wagon if you like. Nothing hypothetical about that.

This is getting better, if you count the number of steam wagon, and the steam content, and a notion of time then it creates a metric that is possible to apply the same way to different design to compare them.

I shall say that my arguments would then go onto the way time between train is calculated, not because i want to be annoying but because i think you have the notion of distance in this time. Those value are linked together in the acceleration formula, which include time and distance. ( wagon count is non-interpretative i'm ok with it, sum of steam in the different wagons, i'm ok with that).

If i want to apply the same method of calculating time between train between different design it needs to be defined because it would differ if a train needs to fully clear himself before another take its place in this case acceleration can be high, but will be capped by the distance if you set it like 1km. OR if a train doesn't need to fully clear himself but only N% or only the locos/part, in other word i think time between train is in theory a value that must be optimized for a particular distance, FOR CERTAIN DESIGN NOT ALL I AGREE.

mrvn wrote: ↑
Mon Nov 01, 2021 3:05 pm
Apparently you still think it makes a difference if I measure trains leaving the reactor, trains on the way, trains arriving at unload, steam unloaded, electricity created from steam. Maybe even weather I use 100 turbines running 100% load or 200 turbines running 50% load when the average steam produces is good for 100 turbines. There is no loss of energy in factorio other than reactors overheating. No matter where you measure the result will always be equivalent. If one reactor is better counting the steam leaving the reactor it will also be better at electricity produced. It's all the same.
Try it., then explain your method, then i will apply the method, realise i'm wrong and shut up, what do you have to lose explaining how i should proceed to evalute design before spending tons of time making one ?

mrvn wrote: ↑
Mon Nov 01, 2021 3:05 pm
And obviously I'm not going to take any measurement at the beginning of the test. That goes without saying. Reactors need to heat up to a working temperature and all that stuff. Which was already one of the arguments why an infinite train wouldn't count. It fills the train during warmup and stops. If I don't measure the first train or the first hour or any other sensible way to define the warmup phase the infinite train will always produce 0.
Again this is unprecise and does not allow me to make a design choice, like if you set a distance = 1KM and a time = 1hours of waiting then 1 hour of counting it would be what i'm arguing for, then i can optimize my design based on those value.

look :
curvetemp.png
curvetemp.png (18.71 KiB) Viewed 4112 times
This is my steam production overtime, you don't measure the first hour? why not the first 61 minutes ? i know it's arbitrary, but the curve i showed you is from 1 design, if you say the amount of time you let things heat up precisely, with a number, like i'm arguing for the distance, then i can alter the design because i know where to crop the curve and plan my train for, otherwise i can just say, this design require 129 hours of initialization. And you know i would :D

This is why you need (imo) to be more precise in the method. Doing so is enough to circumvent many problems i'm causing you.


[Edit: ideally instead of /wagon it could be /rolling stock required for continuous delivery, this way you count 3xLCC train as the same as 1 LLLCCCCCC it's almost the same as per/wagon but if you say distance is 1km, then the math become 32 LCC train vs 9 LLLCCCCCCCCC or something or maybe 28 vs 11, the number of train is not hypothetical or interpretative, it makes it so that there is a theoric optimum for each compositions, depending on the design, you'd try to reduce that number of total amount of rolling stock, by limiting train-down time.]

mrvn
Smart Inserter
Smart Inserter
Posts: 5682
Joined: Mon Sep 05, 2016 9:10 am
Contact:

Re: Contest: Longest train based nuclear reactor

Post by mrvn »

mmmPI wrote: ↑
Mon Nov 01, 2021 3:41 pm
mrvn wrote: ↑
Mon Nov 01, 2021 3:05 pm
And that is AGAIN the problem with your made up hypothetical designs. What you describe and what other people understand from your description are 2 completely different things. Maybe you describe it wrong, maybe the other person doesn't understand. It really doesn't matter. Each one thinks of a different design and you can't argue about it then.

This is why i'm asking for the method, otherwise i'd spend hours making a design you'd just discard because of misunderstanding in the rules ending in you disqualifying it due to what be like arbitrary reason/interpretation of the rule to me.

mrvn wrote: ↑
Mon Nov 01, 2021 3:05 pm
What you describe now is just a fancy and inefficient way to load the train. Since you have heat exchangers for all 5 fluid wagons you can just drive in fully and load all 5 at the same time. There is 0 advantage of stopping along the way and you are just slowing down loading.


And this has just become obsolete since your design isn't a conveyer belt design.
the advantage of stopping along the way allows for not having to wait for 15 or 50 or 100 loco when it's time but instead have a train LCCCLCCCLCCCLCCCLCCC; This means the time between unload in averaged.

No it's not obsolete, you can think of it as a belt maximum troughput, and a belt actual throughput. The belt maximum throughput is the same all along the way, not the actual throughput this depend on where you load. Depending on where you look, on the distance , on how many producers on the left side/right side.

You can saturate a belt with 4 stack inserter working 100% of time, and then have a long belt maxed out.
Or you can saturate a belt with 120 yellow inserter working some % of time. This will take much more space, the belt maxed out will be so only for a few tile if you take 70 tiles measure, 60 inserter on each side, only the 10 last belt have to carry the whole throughput.

More distance allow to earn more time to load given the max throughput, and the max capacity of a single unit. (do you move 1 by 1 wagon to average ? do you move 5 by 5 ? do you move 100000 by 1000000 ?) the more distance you take, the more time you have to accelerate, if there is a fix distance, then there is an optimal.
And all of that is irrelevant. You said the train eventually arrives at the unload and gets unloaded all at once. So your LCCCLCCCLCCCLCCCLCCC arrives and 15 fluid wagons get unloaded in parallel. There is no averaging of anything. To take your belt design from above all you look at is the last belt where you would have a loader putting things in a chest. Everything before is a big black box labeled "Your reactor design". Only what comes out of the box is relevant for the measurement. If the belt is filled with 4 stack inserters or 120 yellow inserters makes 0 difference. I'm not going to measure in the middle of your reactor like you describe for the belt.

If you want a specific on this lets do this just for you: Put your reactor on the positive X axis. Unloading will haven on the negative X axis. How much space you leave between X=0 and your reactor is up to you. Take all the space you want.

mmmPI wrote: ↑
Mon Nov 01, 2021 3:41 pm
Again this is unprecise and does not allow me to make a design choice, like if you set a distance = 1KM and a time = 1hours of waiting then 1 hour of counting it would be what i'm arguing for, then i can optimize my design based on those value.

look :
curvetemp.png

This is my steam production overtime, you don't measure the first hour? why not the first 61 minutes ? i know it's arbitrary, but the curve i showed you is from 1 design, if you say the amount of time you let things heat up precisely, with a number, like i'm arguing for the distance, then i can alter the design because i know where to crop the curve and plan my train for, otherwise i can just say, this design require 129 hours of initialization. And you know i would :D

This is why you need (imo) to be more precise in the method. Doing so is enough to circumvent many problems i'm causing you.


[Edit: ideally instead of /wagon it could be /rolling stock required for continuous delivery, this way you count 3xLCC train as the same as 1 LLLCCCCCC it's almost the same as per/wagon but if you say distance is 1km, then the math become 32 LCC train vs 9 LLLCCCCCCCCC or something or maybe 28 vs 11, the number of train is not hypothetical or interpretative, it makes it so that there is a theoric optimum for each compositions, depending on the design, you'd try to reduce that number of total amount of rolling stock, by limiting train-down time.]
Just send a save game where the reactor is warmed up already and produces a stead and consistent amount of steam. Then I can wait an arbitrary time of my choosing and measure to my harts content. Nothing will change for you since you already waited long enough to get the reactor to full speed no matter how long I wait or don't wait.

And to already answer your next problem: You can specify a minimum time I use to count trains. I might count longer and average it but again that wouldn't change anything if you specify a long enough time to measure in the first time. So you can design you reactor to take 11 minutes to load a train and I can't just measure 10 minutes and say: 0 output.

mmmPI
Smart Inserter
Smart Inserter
Posts: 2674
Joined: Mon Jun 20, 2016 6:10 pm
Contact:

Re: Contest: Longest train based nuclear reactor

Post by mmmPI »

mrvn wrote: ↑
Mon Nov 01, 2021 4:29 pm

And all of that is irrelevant. You said the train eventually arrives at the unload and gets unloaded all at once. So your LCCCLCCCLCCCLCCCLCCC arrives and 15 fluid wagons get unloaded in parallel. There is no averaging of anything. To take your belt design from above all you look at is the last belt where you would have a loader putting things in a chest. Everything before is a big black box labeled "Your reactor design". Only what comes out of the box is relevant for the measurement. If the belt is filled with 4 stack inserters or 120 yellow inserters makes 0 difference. I'm not going to measure in the middle of your reactor like you describe for the belt.
you misunderstood me, only 1 portion gets unloaded at a time, but this portion is fully unloaded in 1 unit of time when it took 20 unit of time to load it 5% at a time.
mrvn wrote: ↑
Mon Nov 01, 2021 4:29 pm
If you want a specific on this lets do this just for you: Put your reactor on the positive X axis. Unloading will haven on the negative X axis. How much space you leave between X=0 and your reactor is up to you. Take all the space you want.
i understand that but this is a range, one part i understand is 0 distance, then the other part is up to max size of the map. If i leave 0 distance it's cheating ? just to make sure. You may want to bound the thing to avoid me saying that you need to place the blueprint 138 tile before the end of the map or that i used 0 spacing between reactor and unload.
mrvn wrote: ↑
Mon Nov 01, 2021 4:29 pm
Just send a save game where the reactor is warmed up already and produces a stead and consistent amount of steam. Then I can wait an arbitrary time of my choosing and measure to my harts content. Nothing will change for you since you already waited long enough to get the reactor to full speed no matter how long I wait or don't wait.

And to already answer your next problem: You can specify a minimum time I use to count trains. I might count longer and average it but again that wouldn't change anything if you specify a long enough time to measure in the first time. So you can design you reactor to take 11 minutes to load a train and I can't just measure 10 minutes and say: 0 output.
no but you could measure 12 and average or 20 and average if i say minium 11min, since it's arbitrary i cannot test myself. that doesn't quite work. also given the previous curve i could say you need to wait minimum 1500 hours before measuring because that's how long my 120201024328 wagon trains takes to load and you would think it's me trying to be annoying while i've been asking for specific things especially to prevent those things to happen.

Edit: i added bald on your quote, i already sent you a reactor, it didn't help.

mrvn
Smart Inserter
Smart Inserter
Posts: 5682
Joined: Mon Sep 05, 2016 9:10 am
Contact:

Re: Contest: Longest train based nuclear reactor

Post by mrvn »

mmmPI wrote: ↑
Mon Nov 01, 2021 4:49 pm
mrvn wrote: ↑
Mon Nov 01, 2021 4:29 pm

And all of that is irrelevant. You said the train eventually arrives at the unload and gets unloaded all at once. So your LCCCLCCCLCCCLCCCLCCC arrives and 15 fluid wagons get unloaded in parallel. There is no averaging of anything. To take your belt design from above all you look at is the last belt where you would have a loader putting things in a chest. Everything before is a big black box labeled "Your reactor design". Only what comes out of the box is relevant for the measurement. If the belt is filled with 4 stack inserters or 120 yellow inserters makes 0 difference. I'm not going to measure in the middle of your reactor like you describe for the belt.
you misunderstood me, only 1 portion gets unloaded at a time, but this portion is fully unloaded in 1 unit of time when it took 20 unit of time to load it 5% at a time.
Because you say one thing in one sentence, like it being a conveyer and another in the next, like "now comes the unload: all at once since the unload is not the bottleneck". Those things are contradictory. It would be so much easier to just look at an implementation that those contradictory descriptions. I'm going to stop commenting on anything hypothetical from you again. It just adds no value.

So pick one. Either the design has 5 loading station and each part of the train moves through all 5 spots to load as you describe and then gets unloaded all at once. Or it is a conveyer system, where A gets loaded, then the train moves on and B gets loaded, it moves on again and C gets loaded while at the same time A might get unloaded and so on. The later case I would count as 5 separate trains because that is the unit that gets loaded and unloaded every time the train stops.

Assuming you have no signals on the track or this won't work: You can actually split your LCCCLCCCLCCCLCCCLCCC train into LCCC components. You already have the train stops perfectly positioned along the way for the long train top stop right. The short trains will stop there too just as well. And if you make them all leave each stop in sync they will move without collisions and look like the long train. Only way to tell the difference is hovering over the train with the mouse to see they aren't connected. And since you can do that I consider the LCCCLCCCLCCCLCCCLCCC conveyer equivalent to 5 LCCC trains.
mmmPI wrote: ↑
Mon Nov 01, 2021 4:49 pm
mrvn wrote: ↑
Mon Nov 01, 2021 4:29 pm
If you want a specific on this lets do this just for you: Put your reactor on the positive X axis. Unloading will haven on the negative X axis. How much space you leave between X=0 and your reactor is up to you. Take all the space you want.
i understand that but this is a range, one part i understand is 0 distance, then the other part is up to max size of the map. If i leave 0 distance it's cheating ? just to make sure. You may want to bound the thing to avoid me saying that you need to place the blueprint 138 tile before the end of the map or that i used 0 spacing between reactor and unload.
mrvn wrote: ↑
Mon Nov 01, 2021 4:29 pm
Just send a save game where the reactor is warmed up already and produces a stead and consistent amount of steam. Then I can wait an arbitrary time of my choosing and measure to my harts content. Nothing will change for you since you already waited long enough to get the reactor to full speed no matter how long I wait or don't wait.

And to already answer your next problem: You can specify a minimum time I use to count trains. I might count longer and average it but again that wouldn't change anything if you specify a long enough time to measure in the first time. So you can design you reactor to take 11 minutes to load a train and I can't just measure 10 minutes and say: 0 output.
no but you could measure 12 and average or 20 and average if i say minium 11min, since it's arbitrary i cannot test myself. that doesn't quite work. also given the previous curve i could say you need to wait minimum 1500 hours before measuring because that's how long my 120201024328 wagon trains takes to load and you would think it's me trying to be annoying while i've been asking for specific things especially to prevent those things to happen.

Edit: i added bald on your quote, i already sent you a reactor, it didn't help.
It's not a range. It's a line in the sand at X=0. You build your reactor on one side, I measure on the other. If you want to build right up to X=0 then that is your choice. If I want to measure right at X=0 then that is my choice too. If you want a certain separation from the measurement, say 1km, then build where X > 1000. You determine the minimum distance from the measurement.

Same again with the time. If you say minimum 11 minutes and you have a train leaving the reactor every 5 seconds then you can figure out the maximum error no matter what time span I measure. You can even say measure a multiple of 10 minutes. Anything within reason. I truly don't care as it won't matter one single bit.

You asked, I gave and now you say I will not honor them anyway. So what is the point of asking for it if you don't trust me anyway? What is the point in explaining any of it if you then say I won't do as I said? Everything here is public. Anyone can do their own measurements and show that I measured wrong if it ever comes to that.

Yes, you send me a reactor, I told you the result. Surprisingly the result was: yes, it loads 4 fluids wagon. None of your imaginary fears about me corrupting results manifested because in any sane design they are that: imaginary. As long as you design your reactor to produce the most steam for the longest train you will get the result you expect.

mmmPI
Smart Inserter
Smart Inserter
Posts: 2674
Joined: Mon Jun 20, 2016 6:10 pm
Contact:

Re: Contest: Longest train based nuclear reactor

Post by mmmPI »

mrvn wrote: ↑
Mon Nov 01, 2021 7:59 pm
Yes, you send me a reactor, I told you the result. Surprisingly the result was: yes, it loads 4 fluids wagon. None of your imaginary fears about me corrupting results manifested because in any sane design they are that: imaginary. As long as you design your reactor to produce the most steam for the longest train you will get the result you expect.
yes it's surprising afer all you bargain about not elongating artificially a train in the contest of the longest train, then you count a partly filled train like a fully filled one probably by mistake because you didnt realised is was the case as you said yourself and then you refuse to explain why it was counted 4 wagon, and not less for this case and how it would be for other case.

also think about the bald sentence when you have some time, "has long as your potato is the biggest and the most orange you win". that is stupid. like " as long as your house is the biggest and the cheapest i will buy it". so the metric is price per square meter ?, no it's size, then if the size is the same we look at the price, so that 200 size house is not as good as this 201 size house that is 3 time more exepensive. No the 200 size house is better because the 201 size house is overpriced.
mrvn wrote: ↑
Mon Nov 01, 2021 7:59 pm
You asked, I gave and now you say I will not honor them anyway. So what is the point of asking for it if you don't trust me anyway? What is the point in explaining any of it if you then say I won't do as I said? Everything here is public. Anyone can do their own measurements and show that I measured wrong if it ever comes to that.
I'm not saying wou will not honor them in the future, it already happened, the design i send you was not properly measured according to you own words. It wasn't measured at all

you declare me the winner already of your stupid contest i will now stop losing my time with you, good luck to other participants.

mrvn
Smart Inserter
Smart Inserter
Posts: 5682
Joined: Mon Sep 05, 2016 9:10 am
Contact:

Re: Contest: Longest train based nuclear reactor

Post by mrvn »

mmmPI wrote: ↑
Tue Nov 02, 2021 3:00 am
mrvn wrote: ↑
Mon Nov 01, 2021 7:59 pm
Yes, you send me a reactor, I told you the result. Surprisingly the result was: yes, it loads 4 fluids wagon. None of your imaginary fears about me corrupting results manifested because in any sane design they are that: imaginary. As long as you design your reactor to produce the most steam for the longest train you will get the result you expect.
yes it's surprising afer all you bargain about not elongating artificially a train in the contest of the longest train, then you count a partly filled train like a fully filled one probably by mistake because you didnt realised is was the case as you said yourself and then you refuse to explain why it was counted 4 wagon, and not less for this case and how it would be for other case.
It must have surprised you. Thought you would trap me into counting badly so you could argue some more? Too bad even there you misunderstood the issue completely. It never was about how full you make the train. Not filling it just makes you loose against others that do fill it. Artificial elongation is and has always been about the heat exchangers.

The reason it's 4 fluid wagons it's simple: because you have the maximum number of heat exchangers possible in the design filling the train all at the same time. You didn't play any games to make the train longer than was waranted for the reactor and heat conduction.
mmmPI wrote: ↑
Tue Nov 02, 2021 3:00 am
also think about the bald sentence when you have some time, "has long as your potato is the biggest and the most orange you win". that is stupid. like " as long as your house is the biggest and the cheapest i will buy it". so the metric is price per square meter ?, no it's size, then if the size is the same we look at the price, so that 200 size house is not as good as this 201 size house that is 3 time more exepensive. No the 200 size house is better because the 201 size house is overpriced.
mrvn wrote: ↑
Mon Nov 01, 2021 7:59 pm
You asked, I gave and now you say I will not honor them anyway. So what is the point of asking for it if you don't trust me anyway? What is the point in explaining any of it if you then say I won't do as I said? Everything here is public. Anyone can do their own measurements and show that I measured wrong if it ever comes to that.
That sentence makes perfect sense in an biggest orange potato growing contest. If the contest is about best tasting then it would be different. Same with biggest house. What good is a 200 size house if your fenemy already has a size 200 house and you want to claim to have the bigger house? It always depends on what the goal is.
mmmPI wrote: ↑
Tue Nov 02, 2021 3:00 am
I'm not saying wou will not honor them in the future, it already happened, the design i send you was not properly measured according to you own words. It wasn't measured at all

you declare me the winner already of your stupid contest i will now stop losing my time with you, good luck to other participants.
Thank god, hope it lasts.

mmmPI
Smart Inserter
Smart Inserter
Posts: 2674
Joined: Mon Jun 20, 2016 6:10 pm
Contact:

Re: Contest: Longest train based nuclear reactor

Post by mmmPI »

mathbasic.png
mathbasic.png (11.08 KiB) Viewed 4055 times

mrvn
Smart Inserter
Smart Inserter
Posts: 5682
Joined: Mon Sep 05, 2016 9:10 am
Contact:

Re: Contest: Longest train based nuclear reactor

Post by mrvn »

Wow, it lasted a woopping 32 minutes.

And you bring out more imaginary made up numbers that are completely meaningless and contrary to the rules.

The rules clearly result in 1rst being first. Sorry that that is still confusing but you labeled them the same. But one of them is clearly first by the rules for the longest train based nuclear reactor. There isn't even a question about it that would require looking at the steam/s to decide between designs with the same train length.

PS: What does illimited mean? That is not a word. No, scratch that, you said you were done with it. Be done with it.

mmmPI
Smart Inserter
Smart Inserter
Posts: 2674
Joined: Mon Jun 20, 2016 6:10 pm
Contact:

Re: Contest: Longest train based nuclear reactor

Post by mmmPI »

mrvn wrote: ↑
Tue Nov 02, 2021 8:04 pm
The rules clearly result in 1rst being first.

mrvn
Smart Inserter
Smart Inserter
Posts: 5682
Joined: Mon Sep 05, 2016 9:10 am
Contact:

Re: Contest: Longest train based nuclear reactor

Post by mrvn »

mmmPI wrote: ↑
Tue Nov 02, 2021 3:00 am
i will now stop losing my time with you
Trying to save time by not adding any words to your post now?

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

Re: Contest: Longest train based nuclear reactor

Post by ptx0 »

maybe not a contest but just show big builds of nuclear trains? i would contribute but i don't feel like competing in anything, this thread kind of shows why, it kinda gets out of hand.

mmmPI
Smart Inserter
Smart Inserter
Posts: 2674
Joined: Mon Jun 20, 2016 6:10 pm
Contact:

Re: Contest: Longest train based nuclear reactor

Post by mmmPI »

Feel free to post any design you want, there is no competition, there was only 1 design submitted and accepted, but since then the rules have changed.

mrvn will evaluate if your train is artificially elongated or not, based on the " I-know-it-when-I-see-it country."

And as long as you use more than 4 wagon you win the contest.

You are not allowed to use infinite train, but you are allowed to use tileable design.

But even if you don't respect the rules you can still post ! it will help other contestant understand when a train is considered artificially elongated because mrvn has trouble defining it with word, he need people to spend hours to make design before he find a way to express why they are disqualified when he sees them.

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

Re: Contest: Longest train based nuclear reactor

Post by ptx0 »

mmmPI wrote: ↑
Wed Nov 03, 2021 1:10 pm
Feel free to post any design you want, there is no competition, there was only 1 design submitted and accepted, but since then the rules have changed.

mrvn will evaluate if your train is artificially elongated or not, based on the " I-know-it-when-I-see-it country."

And as long as you use more than 4 wagon you win the contest.

You are not allowed to use infinite train, but you are allowed to use tileable design.

But even if you don't respect the rules you can still post ! it will help other contestant understand when a train is considered artificially elongated because mrvn has trouble defining it with word, he need people to spend hours to make design before he find a way to express why they are disqualified when he sees them.
I hope that you understand that you've effectively ruined this whole thread.

Locked

Return to β€œEnergy Production”