Not sure what the dual track setup is - here are some screenies of the map (sry for the late reply, had to solve some issues starting factorio).
How common is it to deconstruct EVERYTHING and start over late-mid game?
-
- Inserter
- Posts: 20
- Joined: Tue Jun 28, 2022 7:21 am
- Contact:
Re: How common is it to deconstruct EVERYTHING and start over late-mid game?
-
- Smart Inserter
- Posts: 2765
- Joined: Tue Apr 25, 2017 2:01 pm
- Contact:
Re: How common is it to deconstruct EVERYTHING and start over late-mid game?
2 one-way tracks running in opposite directions placed in parallel. Like your 5Cross.jpg image.
(Edit: Yeah, I made mistakes in this. Been on break and don't have BPs with 1 track spacing, so had to do it all from scratch without testing. Sue me.
![Laughing :lol:](./images/smilies/icon_lol.gif)
That said, and I know others have mentioned how to do signalling, but that 5Cross intersection is not properly signaled.
![Smile :)](./images/smilies/icon_e_smile.gif)
(Image is obviously wrong now.) For example, I'd signal your 5Cross like this (signaled for 1-1 trains to keep the image smaller; just push the signals for the exit blocks further out on your actual setup to fit your sized trains. You'd also want to adjust the rail signals for inbound to your preference):
You can also check out this thread for others' intersections (with BPs) that you can use or at least check out to see how they did it: viewtopic.php?f=194&t=100614
Last edited by FuryoftheStars on Tue Sep 06, 2022 4:18 pm, edited 2 times in total.
My Mods: Classic Factorio Basic Oil Processing | Sulfur Production from Oils | Wood to Oil Processing | Infinite Resources - Normal Yield | Tree Saplings (Redux) | Alien Biomes Tweaked | Restrictions on Artificial Tiles | New Gear Girl & HR Graphics
-
- Inserter
- Posts: 20
- Joined: Tue Jun 28, 2022 7:21 am
- Contact:
Re: How common is it to deconstruct EVERYTHING and start over late-mid game?
As far as I see it, your chain signals are just keeping the train further out of the intersection than necessary. As I said, 80 trains, not a single crash in 700 hours with just the 4 signals.
-
- Smart Inserter
- Posts: 2765
- Joined: Tue Apr 25, 2017 2:01 pm
- Contact:
Re: How common is it to deconstruct EVERYTHING and start over late-mid game?
That's the idea. If there's nothing blocking its path, then it just goes straight through non-stop. But if there's a train stalled on the other side, they'll stop another train from entering the intersection and thus blocking another train on a different track from being allowed through.realcyberleon wrote: ↑Tue Sep 06, 2022 3:07 pmAs far as I see it, your chain signals are just keeping the train further out of the intersection
Edit: Here's an example (and as I re-examined it, I did make a couple of mistakes. But that's to be expected designing something on the fly without testing):
(Edit2: Pfft, yup, and found more mistakes. See next post.
![Rolling Eyes :roll:](./images/smilies/icon_rolleyes.gif)
Fixed version without the trains on it for better clarity (Edit2: Pfft, not really
![Laughing :lol:](./images/smilies/icon_lol.gif)
Last edited by FuryoftheStars on Tue Sep 06, 2022 4:21 pm, edited 3 times in total.
My Mods: Classic Factorio Basic Oil Processing | Sulfur Production from Oils | Wood to Oil Processing | Infinite Resources - Normal Yield | Tree Saplings (Redux) | Alien Biomes Tweaked | Restrictions on Artificial Tiles | New Gear Girl & HR Graphics
-
- Smart Inserter
- Posts: 2765
- Joined: Tue Apr 25, 2017 2:01 pm
- Contact:
Re: How common is it to deconstruct EVERYTHING and start over late-mid game?
Cripes, I think I see what you were talking about, now. Yeah, sorry, the exit signals don't need to be chain. Been on break for a while, and I don't have any BPs with a 1 track spacing like what you are using.
Edit: You can also place additional signals before that fork for the left to right rail line if you want to allow waiting trains to move in sooner. I just had it spaced like that to allow the 1-1 train this was signaled for a stopping point prior to the first possible point for a train to be backed up/stalled at.
And forgive me, I never made these comments thinking that your trains would crash in the other. I gave these recommendations because I can see trains stopping in the crossing and blocking other traffic from carrying on while the stopped train is waiting for clear space and then moves out of the way itself.
Edit: You can also place additional signals before that fork for the left to right rail line if you want to allow waiting trains to move in sooner. I just had it spaced like that to allow the 1-1 train this was signaled for a stopping point prior to the first possible point for a train to be backed up/stalled at.
And forgive me, I never made these comments thinking that your trains would crash in the other. I gave these recommendations because I can see trains stopping in the crossing and blocking other traffic from carrying on while the stopped train is waiting for clear space and then moves out of the way itself.
My Mods: Classic Factorio Basic Oil Processing | Sulfur Production from Oils | Wood to Oil Processing | Infinite Resources - Normal Yield | Tree Saplings (Redux) | Alien Biomes Tweaked | Restrictions on Artificial Tiles | New Gear Girl & HR Graphics
-
- Fast Inserter
- Posts: 208
- Joined: Tue Apr 24, 2018 5:42 am
- Contact:
Re: How common is it to deconstruct EVERYTHING and start over late-mid game?
FuryoftheStars wrote: ↑Tue Sep 06, 2022 4:14 pmCripes, I think I see what you were talking about, now. Yeah, sorry, the exit signals don't need to be chain. Been on break for a while, and I don't have any BPs with a 1 track spacing like what you are using.
Still has errors man.
Hint -
you only need two chain signals to make that work fine.
Hint 2 -
signalling before a split or after a merge doesn't actually do anything in the vast majority of cases.
Solution
-
- Smart Inserter
- Posts: 2765
- Joined: Tue Apr 25, 2017 2:01 pm
- Contact:
Re: How common is it to deconstruct EVERYTHING and start over late-mid game?
Well, that's just optimizing at that point.ColonelSandersLite wrote: ↑Tue Sep 06, 2022 4:38 pmFuryoftheStars wrote: ↑Tue Sep 06, 2022 4:14 pmCripes, I think I see what you were talking about, now. Yeah, sorry, the exit signals don't need to be chain. Been on break for a while, and I don't have any BPs with a 1 track spacing like what you are using.
Still has errors man.
Hint -you only need two chain signals to make that work fine.
Hint 2 -signalling before a split or after a merge doesn't actually do anything in the vast majority of cases.
Solution
![Smile :)](./images/smilies/icon_e_smile.gif)
My Mods: Classic Factorio Basic Oil Processing | Sulfur Production from Oils | Wood to Oil Processing | Infinite Resources - Normal Yield | Tree Saplings (Redux) | Alien Biomes Tweaked | Restrictions on Artificial Tiles | New Gear Girl & HR Graphics
Re: How common is it to deconstruct EVERYTHING and start over late-mid game?
If you never place 2 trains into a single segment then they can never crash. You have to seriously mess up placing trains to make them crash once you have signals. The problem with signals is avoiding deadlocks.realcyberleon wrote: ↑Tue Sep 06, 2022 3:07 pmAs far as I see it, your chain signals are just keeping the train further out of the intersection than necessary. As I said, 80 trains, not a single crash in 700 hours with just the 4 signals.
Re: How common is it to deconstruct EVERYTHING and start over late-mid game?
Still wrong or at least suboptimal.FuryoftheStars wrote: ↑Tue Sep 06, 2022 4:14 pmCripes, I think I see what you were talking about, now. Yeah, sorry, the exit signals don't need to be chain. Been on break for a while, and I don't have any BPs with a 1 track spacing like what you are using.
Generally when a track splits you want a chain signal before the split if both ways can lead to the same destination. Because then the train will wait before the split and picks whatever way opens up first (white arrow). If it truly is a fork in the road going to separate destinations then don't put a chain signal before the fork. This can be important for train stops, where you fork, one side has the stop and then merges back. Put a chain signal there because otherwise a train can decide to go through the stop but the stop then is blocked by a unloading train.
You do not need a chain signal before a merge (purple arrow). You can turn both of them into a normal signal and for the same effect but with a larger exit block. Sometimes those few extra tiles means the exit block will actually be big enough for a train.
Note: If you have a merge (without a crossing like above) you can put a chain signal on on side to give the other preference. This is useful for train station exits so trains moving fast on the main track are less likely to be stopped by a train pulling out of the station. Or when you have congestion the train will stay in the station till the main line clear up some.
-
- Smart Inserter
- Posts: 2765
- Joined: Tue Apr 25, 2017 2:01 pm
- Contact:
Re: How common is it to deconstruct EVERYTHING and start over late-mid game?
Yeah, ColonelSandersLite already beat you to that. Like I said there, we're just nitpicking and optimizing at this point, but the intersection itself works.
My Mods: Classic Factorio Basic Oil Processing | Sulfur Production from Oils | Wood to Oil Processing | Infinite Resources - Normal Yield | Tree Saplings (Redux) | Alien Biomes Tweaked | Restrictions on Artificial Tiles | New Gear Girl & HR Graphics
Re: How common is it to deconstruct EVERYTHING and start over late-mid game?
Thank you, you won’t believe how long I’ve studied this. Will post soon with a diagram of what I don’t understand.mrvn wrote: ↑Sun Sep 04, 2022 1:02 pmWhen you get "no path" that means you either a) place the signals on the wrong side of the track, or b) connected 3 (or any odd number) loops so one runs clockwise, one anticlockwise and then the third would need to run both ways. Think of each loop as a gear wheel. The gear wheels must be able to turn.eBagger wrote: ↑Sun Sep 04, 2022 8:12 am"No need to ever place signals on both sides of a track" - Trust me I think I get what you're saying, and that's what I tried, but unfortunately if I don't place signals on both sides, I get no path, which I assume is because I only have 1 main rail both coming and going. Doesn't make sense to me either but I'd place everything how it did make sense to me, nothing worked, and then when I started placing signals on both sides it started working.mrvn wrote: ↑Sat Sep 03, 2022 11:10 amYou can't have any normal signal on any two-way track. That will always deadlock eventually. You can only use chain signals there. When you have a fork you must have the chain signals before the fork so a train does not enter the fork unless it can leave it. And if the track is one-way you can place a normal signal after the fork if there is enough space AFTER the signal for the train.
Also some of your loops seem to be designed as one way. All the train stations assume the train drives clockwise. But then you put signals on both sides of the track. That makes no sense.
My suggestion would be to first remove all the signals except those at train stations. And then make all the loops go one way (clockwise meeting anti-clockwise) and signal them on one side only. You can make the spurs going to ore fields two-way but place the signals so that trains going in opposite directions don't block the other train from exiting the two-way section. One chain signal and one normal signal at each end where the track has already split. No need to ever place signals on both sides of a track.
loop-fail.png
With the loops all going clockwise alternating with anti-clockwise you should have no problem getting paths. And then you can add stations and spokes. Note that I did add chain signals on both sides of the two-way track at the fork. They are not required but at forks is the one place where adding (chain) signals on both sides of the track does make sense and improve things.
loops-with-spokes.png
On the other hand if you want to go with parallel tracks I do favor the roundabout for it's simplicity. The only rule you absolutely have to follow is that there must be space for a full train after every normal signal. If roundabouts are too close together replace the signals with a chain signal. And if you do make loops it's bets to make them big, otherwise there is a risk of grid-lock (as in every grid).
roundabout.png
Re: How common is it to deconstruct EVERYTHING and start over late-mid game?
There is another way one can explain signaling:
You can place chain signals just everywhere. And I mean everywhere you can. Then there are three simple rules:
Note: You can enable "show-train-breaking-distance" in the debug settings (F4) if you want to see how much that is. Depends on the breaking research and speed (depends on fuel) of the train. So it changes over time.
You can place chain signals just everywhere. And I mean everywhere you can. Then there are three simple rules:
- If you have crossing then you need signals before and after the crossing as close as possible . But that's already covered with "place chain signals everywhere you can".
- If you have space for at least a train after a chain signal with no crossings and no signals then turn the chain signal into a normal signal. For best effects you want the space after a crossing as small as possible (exactly one train length). After that larger spaces are OK and save resources.
- If you have a chain signal in the middle of 2 signals (any combination, only 2 shown) with no crossings then remove the middle one. Do that as often as you can while checking if rule 2 applies again.
Note: You can enable "show-train-breaking-distance" in the debug settings (F4) if you want to see how much that is. Depends on the breaking research and speed (depends on fuel) of the train. So it changes over time.
Re: How common is it to deconstruct EVERYTHING and start over late-mid game?
Getting back to the original question.How often do players get overwhelmed later in the game, and then use robotics to completely dismantle their base and start over more organized? Like lay out a bunch of storage chests, select the entire factory, and dismantle that thing?
I don't deconstruct my whole base.
Let's say I need to rework my green circuit production.
I leave the current green circuit production in place, including trains.
I build an entirely new green circuit production in a different area.
I create new trains to supply the new production area.
Once it is working, (including delivering green circuits from the new area to where they need to be), I stop the trains to the old production area.
I wait until the old area stops production due to being out of raw materials.
Then I deconstruct the old production area.
I've gone through this sequence many times, most recently with oil and gas production.
Of course sometimes there are problems, but I do enjoy the re-work of production areas.
As to trains, I really recommend going through the tutorial on chain signals again.
It doesn't get complicated, but that's the basics you need to know.
Learn how to see where the rail segments are located by holding a signal in hand.
You may have noticed in the replies that even some long-time players don't fully grasp the signals.
I do recommend the dual-track layout -- it's easier to avoid deadlocks than a single track layout (I've tried both).
Roundabouts are great -- just use chain signals within them, use standard signals at the exits from the roundabout.
I have never tried double-ended trains, I generally create some sort of big loop and put the station as a branch/reconnect on the loop.
-
- Filter Inserter
- Posts: 365
- Joined: Mon Jul 03, 2017 9:14 am
- Contact:
Re: How common is it to deconstruct EVERYTHING and start over late-mid game?
I usually leave my initial setup for my mall.
Science gets a new place where I start from scratch.
But to be honest it took me several attempts to be able to start with a fairly good initial setup that can be used later on.
But its still hard to build a good inital base.
Every piece of belt is valueable in the beginning. Later on it does not matter.
At a certain point I only use blue underground belts and splitters, just out of convenience
Science gets a new place where I start from scratch.
But to be honest it took me several attempts to be able to start with a fairly good initial setup that can be used later on.
But its still hard to build a good inital base.
Every piece of belt is valueable in the beginning. Later on it does not matter.
At a certain point I only use blue underground belts and splitters, just out of convenience
Re: How common is it to deconstruct EVERYTHING and start over late-mid game?
I want to point out that usually tutorials or let's play videos are made by people with hundreds or thousands of hours into the game. E.g. Nilaus, one of the popular video tutorial makers, says that he has 7000 hours of playtime. Beyond just raw playtime, you may also not see all the planning that goes into a neat and organized factory. One person I watch, Zisteau, recently spent about half a year planning and making one blueprint.
Basically, you may be comparing yourself to people who have maybe 10 times or even 100 times your playtime or time investment into their setups. It's only normal that you can't make the same setups as effortlessly as they can.
I'd say to not expect so much from yourself. You (hopefully) don't expect to be as fast at the game as a world record speedrunner. You (hopefully) don't expect your first grade writing to be bestseller novel quality. So don't expect your first factory to be as organized and neat as the expert tutorialmakers factory!
So no, don't deconstruct everything. Instead, lower your expectations.
(But yes, this seems to be somewhat common. It's why I made the newgameplus mod, to offer players a way to restart that preserves some progress. I hoped it would help break people out of the restart, build basics, be unhappy, restart cycle that some get trapped in. But mod authors rarely get feedback, so I don't know if it ever helped anyone in that situation.)
I'm an admin over at https://wiki.factorio.com. Feel free to contact me if there's anything wrong (or right) with it.
-
- Fast Inserter
- Posts: 208
- Joined: Tue Apr 24, 2018 5:42 am
- Contact:
Re: How common is it to deconstruct EVERYTHING and start over late-mid game?
K, here's my own personal take on this subject.bll wrote: ↑Mon Sep 12, 2022 3:25 amGetting back to the original question.How often do players get overwhelmed later in the game, and then use robotics to completely dismantle their base and start over more organized? Like lay out a bunch of storage chests, select the entire factory, and dismantle that thing?
I tend to play deathworlds and there is a much greater time pressure and more restricted build space on a DW than on normal settings. I try to, and generally do a *fair* job, of keeping things organized and such but I always end up with a few nasty kludges. I end up finding out that an area is too small for what I'm trying to do or find out that I need to stuff a train with a load of goods somewhere or that I need to tack some production on in a weird area or that I need to cram some belts through a place I didn't plan on putting them or that I need to get something up and running yesterday so I gotta kludge it in or *something* and I just gotta deal with it because the behemoths are just around the corner.
So here's the key - you can't really spend the time and resources screwing around with doing something like that until things are stabilized and by the time your situation is really and truly stable, you're on the verge of launching a rocket anyways.
So what are your goals? Are you *just* trying to launch a rocket? My advice would be to just go for it. Take the lessons you learned and do better next time.
Are you planning on sticking around after the rocket launch to reach a different personal goal? Maybe the classic "I want to produce and consume x science per minute"? In that case, yeah, build anew once you're stable and you have a few quality of life things under your belt, like maybe a few artillery ranges upgrades and a few bot speed upgrades.
This process is usually exactly what leads me to deconstructing my old base.bll wrote: ↑Mon Sep 12, 2022 3:25 amI don't deconstruct my whole base.
Let's say I need to rework my green circuit production.
I leave the current green circuit production in place, including trains.
I build an entirely new green circuit production in a different area.
I create new trains to supply the new production area.
Once it is working, (including delivering green circuits from the new area to where they need to be), I stop the trains to the old production area.
I wait until the old area stops production due to being out of raw materials.
Then I deconstruct the old production area.
Once you're stable, start up a new headquarters to make your things for you. Once you have built that, start building offsite production centers and send resources back to your new HQ. Your old HQ is still there to provide you with things you need while you do these things but will slowly become less and less relevant. Eventually, it won't really be relevant anymore and you can just rip it out and free up the space to do something else there.
I have done the thing where you demolish and rebuild your base on the same site before and I always regret it. It takes forever and you're bound to run short on *something* and may well be forced to scab on some temporary production just to get things going again. For some reason, even though I was hesitant to do it, I decided to do this on my current map. I just really wanted my HQ to be by the crash site and starting lake. I definitely got reminded just how much I hate doing that
Re: How common is it to deconstruct EVERYTHING and start over late-mid game?
To summarize: It's more common to start over late-mid game and only then deconstruct EVERYTHING.
Re: How common is it to deconstruct EVERYTHING and start over late-mid game?
I'm a bit late, maybe more than a bit. I did that, once. Made a massive field of storage chests, as in 4 chunks worth, with roboports and power poles space within it. I started deconstructing everything, from the edge inward to preserve the power and keep the logistics network intact. Those thousands of chests filled up way too fast. All the incomplete products from belts and assemblers had to be put somewhere. I finally ended up shooting, with the SMG chests full of intermediates, and then used the resource sweeper mod to remove all the "products" from the factories.
Hours later an empty map awaited me grand designs.
About 1/3 of the way into my brand new factory I started running short on the basics, belts, poles, inserters, etc.
Plan B. Build the factories needed to build the things I needed to built the mall to build the things I needed to support my grand designs.
I did finally make all the factories for making the main intermediates. I even managed to get enough stuff build to launch a rocket.
Rocket launch was at the 1000 hour mark.
Now, and ever more, it's going to be plan c.
Plan C: build a new section to handle something. Then deconstruct the old version. Rinse and repeat for each section.
It was fun, and certainly a learning experience. I even learned enough to never, ever, not once, do that again!
Hours later an empty map awaited me grand designs.
About 1/3 of the way into my brand new factory I started running short on the basics, belts, poles, inserters, etc.
Plan B. Build the factories needed to build the things I needed to built the mall to build the things I needed to support my grand designs.
I did finally make all the factories for making the main intermediates. I even managed to get enough stuff build to launch a rocket.
Rocket launch was at the 1000 hour mark.
Now, and ever more, it's going to be plan c.
Plan C: build a new section to handle something. Then deconstruct the old version. Rinse and repeat for each section.
It was fun, and certainly a learning experience. I even learned enough to never, ever, not once, do that again!
Re: How common is it to deconstruct EVERYTHING and start over late-mid game?
Not sure this ever got answered. The usual reason that gets disabled is cause you respawn. The game does this as when you respawn you are back to base inventory size without armor bonuses. Also you probably had a lot of the items in your old armor suit that you are about to go pick up. So if the bots started delivering to your new body you could end up with 2x everything.
To answer the original question, it's rare to do a full deconstruct. Unless you are careful, you'll break your robots by either removing electricity poles or roboports too early, or you will make your job hard by removing your radars, so you need to do it in stages. You need heaps of chests for all the plates and circuits you will deconstruct. Usually you would at least keep your mall (the place where you build all your infrastructure) intact and start a new base somewhere else. It will be a lot easier to build a new base if your old base is still making belts, inserters, assemblers, poles, chests, roboports, smelters, etc. Then remove bits of your old base as the new base fulfills that function.