Factorio Forums

Feedback on Factorio: Space Age and Suggestions for Hi-End Content

1. Introduction

I want to express my gratitude to the developers for an amazing game. I really enjoyed playing and exploring it. I finished the game recently, but I started back in 2019 and have logged nearly 13,000 hours. I have all the achievements that can be obtained without starting a new game.

2. My Factory and Links

My factory is very old and huge:
https://factorio.com/galaxy/Iron%20I:%20Alpha5-7.D7X1

It’s so old that there’s hardly anything left near the starting area. Due to the 1000-screenshot limit, only a small portion of the main Nauvis base is visible.
I kindly ask the developers to fix this inconvenience so that large factories can be properly uploaded to the “Galaxy of Fame.”

I also wrote about my factory on the forum here:
viewtopic.php?f=204&t=109072

3. The Hi-End Content Problem

3.1. Impossible to Keep the Factory Running Autonomously for Long
At the late stage of the game, I discovered that I cannot leave my factory running overnight or for a week without intervention.

• The main issue is Gleba science packs, which expire after a certain time (e.g., Legendary packs last about 2.5 hours).
• If the current research doesn’t require these packs, they go to waste.
• To avoid losing them, I have to manually switch research every hour or two or hire someone to do it. But Factorio is about automation, and I dislike such micromanagement.

3.2. Possible Solutions

1. A combinator that lets you select the current research.
2. Allow us to set research in the biolab via the circuit network (at least for infinite researches, so it doesn’t break the balance).
3. Give us the option to run different researches in different labs, so we can plan around the short lifetime of Gleba packs (e.g., keep one lab busy with those packs while another lab works on something else).

4. Infinite Research

They are now much more interesting and varied than before, and essentially remain the only end-game content, as before.

• Having a large factory made it easier to explore space and planets — which was a very positive experience.
• A negative point: some infinite upgrades got reset to zero. Please don’t do that in the future. If you need to rebalance, it’s better to increase the “cost” of upgrades than to take away players’ existing bonuses.

5. My Experience with Space Age

Here is a personal summary of my journey and impressions.

5.1. Dismantling Part of the Base to Improve UPS

1. My UPS dropped to 6, which was painful. I had to dismantle half of my enormous base (reducing drones from 25 million to 10 million), which got UPS up to 30.
2. Taking apart such a huge setup was quite the challenge. Using nuclear rockets causes heavy lag, so I dismantled it piece by piece.
3. I fed chest contents into recycling chains using Fulgora resources, which was another interesting task.

5.2. Exploring Space and Planets

1. I built a “flying square” to travel around, fully autonomous in logistics.
2. I unlocked Fulgora, quickly set up mining, then moved on to Vulcan and Gleba.
3. I really enjoyed the asteroid mechanic — you can create entire autonomous systems for complex logistics challenges.
4. I didn’t make white science packs (I already have 10 million in storage). Instead, I focused on Legendary Asteroids to produce Legendary Rockets and Railgun ammo.

5.3. Railways and Their Specifics

1. I enjoyed using a second rail layer as a concept for higher transport density.
2. On Nauvis, my system is that each facility produces only one product at maximum efficiency; trains are 4 wagons + 1 locomotive. I avoid enabling/disabling stations, so trains don’t get stuck.
3. On Fulgora, the islands are too small for my usual design. I tried two-way trains, but it was inefficient. I had to enable/disable stations, and every 10 hours or so a train would get stuck.
4. On Vulcan, I’m testing a system of separate depots with 1-wagon-1-loco trains. The problem is that we can’t transmit two different signals via radar, e.g., “where to deliver” plus “what cargo.”
5. For a huge base, multiple depots are needed. It’s unclear how to make trains leave the nearest depot rather than a random one. There may be a solution, but I haven’t bothered yet because Gleba packs still require manual research switching anyway.

5.4. Worms on Vulcan and Enemies on Gleba

1. I killed small worms with 20 turrets in the worm’s “tail”, thanks to my high-level upgrades.
2. The medium worm went down just as easily.
3. The large worm got one-shot by my Legendary handheld railgun, though apparently even a blue shell would suffice.
4. Spiders and biters on Gleba are no threat with my advanced upgrades. One artillery in the center of my base is enough to prevent them from nesting.
5. If shell deliveries fail, the base can still get wrecked at night.
6. On Nauvis, once I removed water artillery, I had to bring back thousands of turrets and shells, while the bugs kept breeding at the map’s edge.

5.5. Resource Quality

• A very interesting mechanic: I set up production and reached 700k Legendary plastic.
• It’s hard and forces a major shift in usual strategies — quite a deep engineering challenge.

5.6. The Ruined Planet

• Traveling there feels somewhat unfinished, but potentially a good spot for Promethean Meteorite farming.
• I could make a platform that flies there, encounters random events, and returns to produce black science packs.
• Building such an autonomous system is a cool challenge for automation.

6. Main Suggestions

Below are ideas that I believe could make Factorio’s hi-end content even more exciting:

1. More DLC
   I’m willing to buy additional content that doesn’t break the balance but extends the game — like a 5th planet with a different combination of existing resources.
2. Fully Autonomous Factory Without Manual Research Switching
   Recall the issue of “rotting” Gleba packs. I’d like to leave my factory for a week and have it run without constant micromanagement.
3. Trains Getting Stuck
   When stations are disabled, trains lose their routes and freeze. This mechanic could be improved.
4. Planet Diversity
   Add mountains, deserts, polar caps, oceans, and alternate versions of existing planets (e.g., multiple types of Vulcan or Gleba).
5. Travel to Other Star Systems
   A warp drive allowing us to leave the solar system could reveal binary stars, black holes, unique asteroids, and exotic resources.
6. Deep Drilling
   Expand resource mining under the surface or beneath oceans. Consider gas giants with “cloud cities”.
7. Higher-Quality Rails
   Let upgraded rails increase train speed, providing another motivation to build them.
8. Electric Trains
   An alternative to diesel locomotives — maybe requiring different infrastructure but offering variety.
9. Isotopes
   A more realistic approach to uranium-235/238, half-life decay, helium-3 for fusion reactors, etc., adding complex production chains.
10. Expand Calcite Mechanics
    Introduce reagents that boost old recipes, offering explosive increases in productivity or unique products.
11. Extended Turret Arsenal
    Tank-shell or grenade turrets. We have rocket turrets for space — why not broaden the lineup?
    Spidertron became less relevant with powerful power armor — maybe it needs a unique ability?
12. Moving Resources Between Platforms
    Why can’t we transfer cargo if two platforms orbit the same planet?
13. Variety of Rockets
    Why only one single-use rocket? It’d be logical to have small or large rockets, including reusable ones. Different planets could have different gravity levels, changing launch difficulty.
14. Competitive Events, Rankings, Rewards
    Multiplayer events, ranked servers, periodic challenges, and prizes. Factorio is about automation, so it’d be fun to compete on who can build certain projects the fastest.
15. Unique Planets & Special Conditions
    For example, a planet with a rare material for “flying saucers” that simplify orbital launches. Or a gas giant with destructive storms requiring special defenses. Or a radioactive planet demanding shield modules in every building.

7. Conclusion

I absolutely love Factorio and appreciate any new content. The ideas introduced in Space Age (item quality, asteroids, planets) greatly expand the gameplay, making it deeper and more engaging.

However, there are a couple of major pain points, particularly:

• “Rotting” Gleba packs and the need for manual research switching.
• Trains getting stuck when stations are disabled.

Solving these issues would significantly improve the late-game experience. As a loyal fan, I’ll gladly buy new DLCs and enjoy the expansion of the Factorio universe. Thank you for your excellent work!

Original text before chatgpt translate is in attach.

Could you expand a bit about how the Gleba science packs prevent automation? You can just toss the spoilage into a furnace so they can be handled automatically. The waste kinda feels bad but they're produced renewably so you're not really losing much. You could certainly turn off the gleba factory when you're not researching something that uses it to save UPS too.

I'll try to explain it in simple terms. My Gleba packs are regularly delivered to Nauvis, where they're evenly distributed among the labs. If any of them spoil at any step, it's all handled correctly. That's not really the issue. The problem is that I have to manually switch research every hour (or even more often). Needing to check in every single hour is unacceptable for me.

On the scale of going AFK for a week, I'd lose about 99% of the Gleba packs to spoilage. Even briefly logging in would reduce that loss to virtually zero. It feels like the kind of pressure you'd get from a hardcore online game, and I don't want that.

Whether the packs spoil or I shut down their production doesn't make much difference at my factory's scale, and it doesn't affect UPS either. I have about 10,000 drones on Gleba, which is negligible compared to the 10 million on Nauvis. Even the idea of stopping or starting Gleba pack production automatically can't be done, because there's no way to feed current research info into combinators.

But you're saying that you have continous supply of the packs setup already. In what way does it matter if they spoil? Either you consume them or you don't, but like you said you get a fresh delivery anyway. Or is this simply about the resources getting wasted by spoilage? I don't mind that so much, as they are free (infinite) to produce anyway.

Banks do spoil, yes, and they are infinite, yes. They’re always in the labs, ready to be used, yes. The problem is that I know if I open the game every hour for just a few seconds to switch research, I can save a huge number of packs from spoiling. Then 10 minutes later, I’d switch back to the original research. It grabs my attention and keeps me from sleeping well. That’s the real issue, not the fact that the packs are going to waste.

Different question about your base: you said in the OP that upon installing Space Age your base ran at 6 UPS. What was the UPS of your massive base before Space Age?

jaylawl wrote: ↑Sun Dec 22, 2024 10:21 pm Different question about your base: you said in the OP that upon installing Space Age your base ran at 6 UPS. What was the UPS of your massive base before Space Age?

That’s a minor translation inaccuracy. I took about a year-long break, and from what I recall, the UPS was around 6 both before and after installing Space Age. From an IT specialist’s perspective, it seems there might have been a performance leak related to pollution. In version 1.0, I had a huge amount of pollution, and even leaving the factory idle for several weeks didn’t help clear it up. After Space Age, within a couple of days, pollution levels dropped significantly, and once it fully dissipated, my UPS went up a lot (maybe other factors also helped, but it’s unlikely).

I think you are making Gleba a bigger problem than it actually is. My research array is set to only insert packs when they are needed, but even if research is not using a bio-pack from Gleba and it spoils in a Biolab, I have an inserter to remove it, put it in a logistics chest, and have a bot come take it to a furnace or to convert it to nutrients depending on what is needed at the time.

I also don't know what you meant about legendary asteroids. I haven't seen any asteroids of any varying quality at all. If that was a typo, then I am not sure what you meant to say.

I started over from scratch on SA rather than trying to fix any issues with my old world map. Everything I did runs automated just fine for even days at a time without intervention. Lately I have been letting it run just to do infinite research, but I just check on it every few days to see how it is doing.

Anlide wrote: ↑Sun Dec 22, 2024 8:44 pm The problem is that we can’t transmit two different signals via radar

You totally can send an arbitrary amount of differnt signals via radar. For simplicity, you could send signal A, if the tick is odd and B, if the tick is even using one control signal (ie a virtual signal not in use) to tell the connections, which signal is currently send. This can be done for any number of signals, as long as an N tick delay is fine, if you use N different signals.

Apparently, its called multiplexing and someone explained it in this post: viewtopic.php?f=193&t=61722&p=373141&hi ... ex#p373141

ichVII wrote: ↑Sun Dec 22, 2024 11:59 pm

Anlide wrote: ↑Sun Dec 22, 2024 8:44 pm The problem is that we can’t transmit two different signals via radar

You totally can send an arbitrary amount of differnt signals via radar. For simplicity, you could send signal A, if the tick is odd and B, if the tick is even using one control signal (ie a virtual signal not in use) to tell the connections, which signal is currently send. This can be done for any number of signals, as long as an N tick delay is fine, if you use N different signals.
Apparently, its called multiplexing and someone explained it in this post: viewtopic.php?f=193&t=61722&p=373141&hi ... ex#p373141

I made a version to showcase the use of radar for this usage specifically : 112723
At the time it was a modded radar meant to mimick the current ones but it is functionnaly the same.

Anlide wrote: ↑Sun Dec 22, 2024 9:41 pm On the scale of going AFK for a week, I'd lose about 99% of the Gleba packs to spoilage. Even briefly logging in would reduce that loss to virtually zero. It feels like the kind of pressure you'd get from a hardcore online game, and I don't want that.

I don't get how it works with the later:

Anlide wrote: ↑Sun Dec 22, 2024 9:41 pm Whether the packs spoil or I shut down their production doesn't make much difference at my factory's scale, and it doesn't affect UPS either. I have about 10,000 drones on Gleba, which is negligible compared to the 10 million on Nauvis.

Then why not just let them run all the time and you won't lose 99% ? Only when you don't reseach you loose the pack but it shouldn't be a problem is it ? like the ressources on Gleba are infinite, and if they spoil or not it makes no difference you said. I don't understand what's the problem with "not researching" and having them spoil, if it doesn't make much difference on the factory.

Anlide wrote: ↑Sun Dec 22, 2024 9:41 pm Even the idea of stopping or starting Gleba pack production automatically can't be done, because there's no way to feed current research info into combinators.

You can detect the speed at which the number of pack diminish in Nauvis though by comparing the content of the network at 2 different successive moment. This way you know if you science pack are "eaten by labs fast" or "spoiling slowly". It is also possible to detect the frequency at which the inserter feeding agricultural science pack in a lab is going. There are ways to achieve the stop/pause for Gleba's science , but it is more difficult than just letting the pack spoil.

MoDon wrote: ↑Sun Dec 22, 2024 11:44 pm I think you are making Gleba a bigger problem than it actually is. My research array is set to only insert packs when they are needed, but even if research is not using a bio-pack from Gleba and it spoils in a Biolab, I have an inserter to remove it, put it in a logistics chest, and have a bot come take it to a furnace or to convert it to nutrients depending on what is needed at the time.

I don’t have any technical issues with Gleba packs spoiling; it’s more of a psychological issue for me.

Listen, the game is about automating actions. Right now, the “correct” method is to log in every hour and manually switch researches — something you cannot automate. That’s the core problem I’m referring to.

MoDon wrote: ↑Sun Dec 22, 2024 11:44 pm I also don't know what you meant about legendary asteroids. I haven't seen any asteroids of any varying quality at all. If that was a typo, then I am not sure what you meant to say.

Regarding “Legendary Asteroids”, that’s not a typo.

MoDon wrote: ↑Sun Dec 22, 2024 11:44 pm I started over from scratch on SA rather than trying to fix any issues with my old world map. Everything I did runs automated just fine for even days at a time without intervention. Lately I have been letting it run just to do infinite research, but I just check on it every few days to see how it is doing.

Unfortunately, Factorio doesn’t give you a quick way to dismantle a mega-base. It’s a painstaking process that can take well over 100 hours. Parts of it can be interesting, but overall it becomes quite tedious.

ichVII wrote: ↑Sun Dec 22, 2024 11:59 pm

Anlide wrote: ↑Sun Dec 22, 2024 8:44 pm The problem is that we can’t transmit two different signals via radar

You totally can send an arbitrary amount of differnt signals via radar. For simplicity, you could send signal A, if the tick is odd and B, if the tick is even using one control signal (ie a virtual signal not in use) to tell the connections, which signal is currently send. This can be done for any number of signals, as long as an N tick delay is fine, if you use N different signals.

Apparently, its called multiplexing and someone explained it in this post: viewtopic.php?f=193&t=61722&p=373141&hi ... ex#p373141

Thank you, I understand your idea. I was able to improve my depot system on Vulcan based on this concept. Now, "extra" trains no longer leave to fulfill a single delivery request.

However, this doesn’t address the issue of how to fairly distribute rare resources among different consumers. Rare resources will always go to the closest consumer. On Nauvis (where I have 3,000 trains), I’ve implemented a balancer where trains wait at unloading stations. But with the depot system, I still can’t figure out how to achieve something similar.

Additionally, in the depot-based system, it’s unclear how to dispatch a train from the nearest depot rather than a random one.

Anlide wrote: ↑Tue Dec 24, 2024 7:08 pm However, this doesn’t address the issue of how to fairly distribute rare resources among different consumers. Rare resources will always go to the closest consumer. On Nauvis (where I have 3,000 trains), I’ve implemented a balancer where trains wait at unloading stations. But with the depot system, I still can’t figure out how to achieve something similar.

Supposedly the addition in Space Age that allow to deal with this is the "priority system" on train stop.

mmmPI wrote: ↑Mon Dec 23, 2024 1:35 am

Anlide wrote: ↑Sun Dec 22, 2024 9:41 pm On the scale of going AFK for a week, I'd lose about 99% of the Gleba packs to spoilage. Even briefly logging in would reduce that loss to virtually zero. It feels like the kind of pressure you'd get from a hardcore online game, and I don't want that.

I don't get how it works with the later:

Anlide wrote: ↑Sun Dec 22, 2024 9:41 pm Whether the packs spoil or I shut down their production doesn't make much difference at my factory's scale, and it doesn't affect UPS either. I have about 10,000 drones on Gleba, which is negligible compared to the 10 million on Nauvis.

Then why not just let them run all the time and you won't lose 99% ? Only when you don't reseach you loose the pack but it shouldn't be a problem is it ? like the ressources on Gleba are infinite, and if they spoil or not it makes no difference you said. I don't understand what's the problem with "not researching" and having them spoil, if it doesn't make much difference on the factory.

Anlide wrote: ↑Sun Dec 22, 2024 9:41 pm Even the idea of stopping or starting Gleba pack production automatically can't be done, because there's no way to feed current research info into combinators.

You can detect the speed at which the number of pack diminish in Nauvis though by comparing the content of the network at 2 different successive moment. This way you know if you science pack are "eaten by labs fast" or "spoiling slowly". It is also possible to detect the frequency at which the inserter feeding agricultural science pack in a lab is going. There are ways to achieve the stop/pause for Gleba's science , but it is more difficult than just letting the pack spoil.

Does no one realize that the real issue isn’t the technical side of handling spoiling items? The real problem is that the most efficient way to use Gleba packs right now is to log into the game every hour, day and night.

Anlide wrote: ↑Tue Dec 24, 2024 7:19 pm Does no one realize that the real issue isn’t the technical side of handling spoiling items? The real problem is that the most efficient way to use Gleba packs right now is to log into the game every hour, day and night.

I don't understand this, you only need to log to queue another research ? What is that you do every hour when you log to be more efficient that if you had not logged ?

mmmPI wrote: ↑Tue Dec 24, 2024 7:15 pm

Anlide wrote: ↑Tue Dec 24, 2024 7:08 pm However, this doesn’t address the issue of how to fairly distribute rare resources among different consumers. Rare resources will always go to the closest consumer. On Nauvis (where I have 3,000 trains), I’ve implemented a balancer where trains wait at unloading stations. But with the depot system, I still can’t figure out how to achieve something similar.

Supposedly the addition in Space Age that allow to deal with this is the "priority system" on train stop.

That’s a helpful feature — I really like it. I used it on Nauvis, where I effectively have unlimited 1.0-quality resources, with instant restocking of everything consumed. So I simply lowered the priority of new stations that produce the same resources of all qualities, and it worked perfectly.

However, if you build a new mega-factory that uses a depot system, this feature won’t resolve the issue I mentioned. For example, suppose a facility produces five different grades of tungsten ore. We have two consumers: “Tungsten carbide” and “Tungsten plate”. After some time, we accumulate enough legendary-grade ore for just one train. How do we balance that ore among all the consumers? And how can we scale the system so that we can have ten facilities of each type — both producers and consumers — and still have the rare legendary ore distributed evenly to everyone? Meanwhile, the consumer setups themselves should also be something we can drop in via blueprint.

mmmPI wrote: ↑Tue Dec 24, 2024 7:24 pm

Anlide wrote: ↑Tue Dec 24, 2024 7:19 pm Does no one realize that the real issue isn’t the technical side of handling spoiling items? The real problem is that the most efficient way to use Gleba packs right now is to log into the game every hour, day and night.

I don't understand this, you only need to log to queue another research ? What is that you do every hour when you log to be more efficient that if you had not logged ?

Let’s say I’m currently researching drone speed. It requires a large amount of packs from Fulgora, so if I just leave it running, my labs will start to accumulate packs from Vulcan, Aquillo, and black science packs. However, in the meantime, the Gleba packs spoil. To prevent that, I log in and switch to any research that requires only Gleba packs (which takes just a few seconds). Then, after about ten minutes, I switch back to the drone-speed research (in the meantime, I do non-Factorio things). Again, it only takes a few seconds, but it distracts me from fully focusing on other tasks and forces me to plan my day around returning to my computer every hour to tweak the research. Fifty minutes later, I repeat the process.

Realizing this made me compare Factorio’s late-game loop to the gameplay of an online game.

Anlide wrote: ↑Tue Dec 24, 2024 7:30 pm For example, suppose a facility produces five different grades of tungsten ore. We have two consumers: “Tungsten carbide” and “Tungsten plate”. After some time, we accumulate enough legendary-grade ore for just one train. How do we balance that ore among all the consumers? And how can we scale the system so that we can have ten facilities of each type — both producers and consumers — and still have the rare legendary ore distributed evenly to everyone? Meanwhile, the consumer setups themselves should also be something we can drop in via blueprint.

I think the answer is generally "if you don't have enough for all the consumers, make more", or "make it so its not problematic that some consumer are starved due to the shortage". Because the "solutions" to split evenly a ressources as you describe require complex circuitry. It would require i think implementing a mechanism to measure your production of that legendary thing( couting items), and divide by the number of consumer, to load train with the appropriate fraction of production, ( production measured at 1000 / minutes, 5 consumer, load 200 per train every 12 seconds and make sure the train is told to leave the load area when ready ).

That wouldn't be enough though, it would also require that once a train arrive at a unload the unload turn off, until they are all off, and then they all open again when they are all off.

Very complex imo x)

Let’s say I’m currently researching drone speed. It requires a large amount of packs from Fulgora, so if I just leave it running, my labs will start to accumulate packs from Vulcan, Aquillo, and black science packs. However, in the meantime, the Gleba packs spoil. To prevent that, I log in and switch to any research that requires only Gleba packs (which takes just a few seconds). Then, after about ten minutes, I switch back to the drone-speed research (in the meantime, I do non-Factorio things). Again, it only takes a few seconds, but it distracts me from fully focusing on other tasks and forces me to plan my day around returning to my computer every hour to tweak the research. Fifty minutes later, I repeat the process.

Well i think the "efficent" way would be to not produce Gleba's science pack if you are researching drone speed, because it's not required. You cannot have all the science pack producing at 1000 per minute and consume them all with labs that can only consume 1000 science per minutes. I think that's the thing that annoy you, because you see them spoiling when they are not used, and you can "prevent" that by queuing a research quickly, because the other science pack can accumulate / delayed in consumption, but not Gleba's.

I have had a game where it wasn't megabase, but i still tried to let it run 24/7 to test the reliability. All my Gleba's science pack were spoiling when i researched drone speed, i had not setup anything to prevent them from being produced in that time, that would have been more difficult and smarter but it wasn't "necessary". When i logged i could just queue any receipe, even if it had Gleba's pack in it, during the night maybe 1 research with Gleba's pack from the queue would be done without me having to do things, it would change from a state where they go to waste , to a state where they are consumed and maybe back to a state where they go to waste. While the fulgora science pack would go from a state where they are backed up, to a state where they are consumed , to a state where they are backed up again when a research from the queue would require some of them.

mmmPI wrote: ↑Tue Dec 24, 2024 7:54 pm

Anlide wrote: ↑Tue Dec 24, 2024 7:30 pm For example, suppose a facility produces five different grades of tungsten ore. We have two consumers: “Tungsten carbide” and “Tungsten plate”. After some time, we accumulate enough legendary-grade ore for just one train. How do we balance that ore among all the consumers? And how can we scale the system so that we can have ten facilities of each type — both producers and consumers — and still have the rare legendary ore distributed evenly to everyone? Meanwhile, the consumer setups themselves should also be something we can drop in via blueprint.

I think the answer is generally "if you don't have enough for all the consumers, make more", or "make it so its not problematic that some consumer are starved due to the shortage". Because the "solutions" to split evenly a ressources as you describe require complex circuitry. It would require i think implementing a mechanism to measure your production of that legendary thing( couting items), and divide by the number of consumer, to load train with the appropriate fraction of production, ( production measured at 1000 / minutes, 5 consumer, load 200 per train every 12 seconds and make sure the train is told to leave the load area when ready ).

That wouldn't be enough though, it would also require that once a train arrive at a unload the unload turn off, until they are all off, and then they all open again when they are all off.

Very complex imo x)

For a mega-base, it’s enough to design one extremely complex system and then replicate it thousands of times. But you can’t rely on shutting down stations. If you disable a station while a train is en route, it might stop anywhere — potentially on an intersection — and block everything. Likewise, if you disable a station while a train is refueling, it will just stay there and prevent other trains from refueling. I think it should be different: consumers should somehow “register”, and the system would serve them all in turn. After every consumer is served once, it repeats. Theoretically, for balancing, you’d send one train from each depot at the same time so the nearest depot handles each request. But to do that, you need to know in advance how many trains are going to arrive, otherwise you end up with too many. It’s probably solvable in theory, but it’s unclear how to handle that “registration” process.

mmmPI wrote: ↑Tue Dec 24, 2024 7:54 pm

Let’s say I’m currently researching drone speed. It requires a large amount of packs from Fulgora, so if I just leave it running, my labs will start to accumulate packs from Vulcan, Aquillo, and black science packs. However, in the meantime, the Gleba packs spoil. To prevent that, I log in and switch to any research that requires only Gleba packs (which takes just a few seconds). Then, after about ten minutes, I switch back to the drone-speed research (in the meantime, I do non-Factorio things). Again, it only takes a few seconds, but it distracts me from fully focusing on other tasks and forces me to plan my day around returning to my computer every hour to tweak the research. Fifty minutes later, I repeat the process.

Well i think the "efficent" way would be to not produce Gleba's science pack if you are researching drone speed, because it's not required. You cannot have all the science pack producing at 1000 per minute and consume them all with labs that can only consume 1000 science per minutes. I think that's the thing that annoy you, because you see them spoiling when they are not used, and you can "prevent" that by queuing a research quickly, because the other science pack can accumulate / delayed in consumption, but not Gleba's.

I have had a game where it wasn't megabase, but i still tried to let it run 24/7 to test the reliability. All my Gleba's science pack were spoiling when i researched drone speed, i had not setup anything to prevent them from being produced in that time, that would have been more difficult and smarter but it wasn't "necessary". When i logged i could just queue any receipe, even if it had Gleba's pack in it, during the night maybe 1 research with Gleba's pack from the queue would be done without me having to do things, it would change from a state where they go to waste , to a state where they are consumed and maybe back to a state where they go to waste. While the fulgora science pack would go from a state where they are backed up, to a state where they are consumed , to a state where they are backed up again when a research from the queue would require some of them.

For me, “not producing” something isn’t an efficient solution. In version 1.0, I could just store science packs, and I had storage for about 15 million packs of each type, which was very effective.

Yes, I also left the game running overnight to research something that required Gleba packs. So you do understand the problem I’m describing — it’s just that psychologically, it doesn’t bother you as much as it does me.

I hope the developers notice this issue and come up with some improvements. Here are a few ideas:

• A storage facility that severely slows spoilage, allowing you to keep spoilable items for a long time (maybe modules that consume a lot of electricity).
• A Biolab that outputs logistic signals indicating which science packs are needed for the current research.
• Infinite researches that have icons, allowing us to read the “active research” from the Biolab and set a new research via that interface.
• The ability to run different researches in different labs, so excess packs on certain planets could still be used for other tasks.

Anlide wrote: ↑Tue Dec 24, 2024 8:34 pm But you can’t rely on shutting down stations. If you disable a station while a train is en route

My wording was bad, i meant reducing the limit to 0. 'turning the request off' in my mind, not the train stop itself. As if it was like 10 little lamp and each train goes for one, turn it off, when no lamp are lit anymore, it means 1 trains was dispatched to every lamp. I haven't achieve this in game, i see this only in my imagination x)

I think the idea of having the consumer register can work too. It's slightly different. In both case it is a very complex setup I understand for a megabase it is more worth investing the time to make the system. I would spent probably as much time for it than for a full 1.1 playthrough.

Note that this is not considering the interrupt system. This could also be a direction to look after for a consistent system. Things like "train in the depot move when their interrupt trigger". And then such interrupt being a circuit condition, that you get from a radar, that receive some information from consumer themselves connected with radar or something. I haven't used this for the particular use case of splitting ressources evenly in case of supply tension. ( i use this to have a smart train that fetch science, or dispatch ammo to defense )

Anlide wrote: ↑Tue Dec 24, 2024 8:34 pm For me, “not producing” something isn’t an efficient solution. In version 1.0, I could just store science packs, and I had storage for about 15 million packs of each type, which was very effective.

Yep i understand well the problem x) i think though it's on purpose that Gleba's science pack works this way. To a lesser extend, the same apply to Fulgora and Vulcanus science packs, if you decide to research only artillery range, i think it's one that consume only Vulcanus science pack, you "risk" having the Fulgora science packs "not producing" because you have filled up your storage of 15 million. ( the extreme case where you let it run a week in Vulcanus science). The Fulgora science pack once their buffer is full aren't produced anymore so their infrastructure is not efficient. Now with Gleba's science pack it's worse, because you can't hoard. But i think it boils down to the fact that you have "more" different types of sciences packs than those that can be used at the same time. So without a clever tech queue that drain the buffer that constituted during a research. You will have to suffer the inneficencies