Dual inserter mechanic

[zOldBulldog]

It has always bugged me that train tracks are two wide while inserter are one, making it impossible to transboard goods directly from one wagon to another.

In other words, there is no way to use a stack inserter between two rails, and the only alternative - the long inserter - is so much slower that it is worthless for the purpose.

I suppose that some of the warehouse mods would do the trick (train-stack inserter-warehouse-stack inserter-train) but that is not an ideal solution.

Trying to come up with a good solution I realized it could be so simple:
- Just make two inserters that are placed pointing in the same direction one after another with no space between them still pickup the item from each other.
- Pretty much like what happens when you place a long inserter right next to an assembler and it still picks up/drops the items fine even.

That way when you want to bridge a two space gap (like that between tracks) with stack inserters you just use two stack inserters.

The alternative would be creating a long stack inserter, but the new mechanic seems simpler and cleaner.

[bobingabout]

Try using a long handed inserter. Wagons are 2 tiles wide, so the pickup and drop locations will be inside the wagon for both. You can even double them up like your suggestion of the inserters grabbing items from each other to make them move items twice as fast.

Alternatively, you could use mods.
Bob's Inserters mod will allow you to configure a stack inserter to be "Long handed".

[zOldBulldog]

Try using a long handed inserter. Wagons are 2 tiles wide, so the pickup and drop locations will be inside the wagon for both. You can even double them up like your suggestion of the inserters grabbing items from each other to make them move items twice as fast.

Alternatively, you could use mods.
Bob's Inserters mod will allow you to configure a stack inserter to be "Long handed".

Long handed inserters are slow as molasses compared to stack inserters. They are not suitable for loading/unloading trains, except at the most those trains we use to carry outpost construction supplies.

DEVs chose to hold (completionist) players hostage in vanilla by disabling steam achievements if you use mods. So, while mods might do the trick, they will do so only after many hundreds of hours of play. Again, not adequate.

[Nefrums]

Use two stack inserters with a car between them.

[eradicator]

In other words, there is no way to use a stack inserter between two rails, and the only alternative - the long inserter - is so much slower that it is worthless for the purpose.

Orly?

cars.jpg (37.85 KiB) Viewed 5857 times

From a modding perspective it's much easier to implement a long stack inserter (many mods already do) than a dynamically adjustable normal stack inserter. It's also less UPS cost to have one inserter. As a vanilla mechanic inserter "chaining" might be interesting, but probably too powerful, as you can't just make it an "works only for this edge-case with trains" thing.

[zOldBulldog]

Use two stack inserters with a car between them.

LOL, that is exactly the problem I'm trying to solve, to move items from one car to the next.

Since train tracks can only be placed zero or 2 spaces apart... a stack inserter will not reach.

Oh, LOL... you mean a wheeled car. Not a train car.

But I think we still need something for Vanilla, maybe a "long stack inserter" then, or the ability to place track 1 apart.

[Zavian]

That solution is vanilla.

[n7m6e7]

I admit, I dont like using cars as storage containers, it feels like im using an exploit more than a feature...


But even so, couldn't you just use blue belts? If you are really obsessive about transfer speed, you can even unload and load from both sides with belt spaghetti magic.

[zOldBulldog]

I agree, using a car = automobile (not to be confused with a car = railcar = wagon like I did earlier ) does feel weird.

So, at least until Vanilla gets some inserter mechanism, since I already got done with the Vanilla achievements, I will probably bite the bullet and use the Warehousing mod (https://mods.factorio.com/mod/Warehousing) and do wagon-6 stack inserters-warehouse-6 stack inserters-wagon to avoid the throughput loss of going from wagon to belt to wagon.

The more I think about it the more I like the idea of using that mod. Train to belt unloading has always been an irritant to me, and this mod seems to solve it neatly: wagon-6 stack inserters-warehouse-# stack inserters-# belts... where # is whatever number of belts I want. Boom! No need of fancy balancing mechanisms. And if the stack inserters don't fully saturate the blue belt, I just merge a couple of belts right after the exit from the warehouse.

[eradicator]

So, at least until Vanilla gets some inserter mechanism, since I already got done with the Vanilla achievements,

Hope you're prepared to wait a good while.

I will probably bite the bullet and use the Warehousing mod (https://mods.factorio.com/mod/Warehousing) and do wagon-6 stack inserters-warehouse-6 stack inserters-wagon to avoid the throughput loss of going from wagon to belt to wagon.

For your "might become a megabase in the future"-base-thing i recommend you design the whole thing in a way that doesn't require inter-train-loading in the first place. It's generally a waste of UPS to stuff things in and out of containers :p.

[zOldBulldog]

For your "might become a megabase in the future"-base-thing i recommend you design the whole thing in a way that doesn't require inter-train-loading in the first place. It's generally a waste of UPS to stuff things in and out of containers :p.

Yep. If feasible I will avoid transbording between trains. If not, use a warehouse.

For train unloads... I am leaning more and more to use a warehouse instead of chests. Will test it this week or next.

[eradicator]

For train unloads... I am leaning more and more to use a warehouse instead of chests. Will test it this week or next.

Chests with very high slot count (aka "warehouses") are a well known UPS drain btw, due to insertion cost being (linearly?) correlated to slot count. I.e. 10 inserters inserting into a single chest with 100 slots, is 10 times as expensive as 10 inserters inserting into 10 chests with 10 slots each (very very roughly speaking). So if you just want to balance stuff get a warehouse with not-so-many-slots.

[zOldBulldog]

For train unloads... I am leaning more and more to use a warehouse instead of chests. Will test it this week or next.

Chests with very high slot count (aka "warehouses") are a well known UPS drain btw, due to insertion cost being (linearly?) correlated to slot count. I.e. 10 inserters inserting into a single chest with 100 slots, is 10 times as expensive as 10 inserters inserting into 10 chests with 10 slots each (very very roughly speaking). So if you just want to balance stuff get a warehouse with not-so-many-slots.

I would have never guessed it. Small warehouses it is.

[Koub]

I would have never guessed it. Small warehouses it is.

viewtopic.php?p=129829#p129829
Now, you even have the source

[zOldBulldog]

I would have never guessed it. Small warehouses it is.

viewtopic.php?p=129829#p129829
Now, you even have the source

Sounds like they never fixed it. But at least it is clear that the problem is the number of slots, not the warehouse itself. Go for the same size (6x6) and as many slots as 6 chests... and I hope it will behave similarly without to horrible a hit. Of course, it all depends on whether I can choose that.

Horrid chest design though.

[eradicator]

I would have never guessed it. Small warehouses it is.

viewtopic.php?p=129829#p129829
Now, you even have the source ;)

Sounds like they never fixed it. But at least it is clear that the problem is the number of slots, not the warehouse itself. Go for the same size (6x6) and as many slots as 6 chests... and I hope it will behave similarly without to horrible a hit. Of course, it all depends on whether I can choose that.

Horrid chest design though.

The last post in that thread clearly states that there isn't anything to be fixed. Iterating a list of length n is obviously O(n) :p. (Short of redesigning the whole system to not use slot iteration obviously.) Also using a warehouse with 240 slots will still be 6 times slower than using 6 chests with 40 each, as i said above. As it's insert_operations * slot_count_of_target_inventory.

[zOldBulldog]

I would have never guessed it. Small warehouses it is.

viewtopic.php?p=129829#p129829
Now, you even have the source

Sounds like they never fixed it. But at least it is clear that the problem is the number of slots, not the warehouse itself. Go for the same size (6x6) and as many slots as 6 chests... and I hope it will behave similarly without to horrible a hit. Of course, it all depends on whether I can choose that.

Horrid chest design though.

The last post in that thread clearly states that there isn't anything to be fixed. Iterating a list of length n is obviously O(n) :p. (Short of redesigning the whole system to not use slot iteration obviously.) Also using a warehouse with 240 slots will still be 6 times slower than using 6 chests with 40 each, as i said above. As it's insert_operations * slot_count_of_target_inventory.

That is what I meant. It is what it is and they decided to not do anything about it. What is horrid is the design of interacting with every slot at every add/remove... even though I can see why they might have chosen to go that route.

In any case we must live with it whether we like it or not.

[Lubricus]

You guys, there is an simple vanilla solution. Use diagonal tracks... and you can fit stack inserters between the wagons.

[zOldBulldog]

You guys, there is an simple vanilla solution. Use diagonal tracks... and you can fit stack inserters between the wagons.

Bingo!

And that is certainly original.

Otherwise I am thinking that an easy fix to the the core problem could be a "single item type" checkbox that could be applied to chests. When the flag is checked, only one kind of item is allowed and then the chest no longer needs to check every slot... just the total quantity. UPS problem... gone.

[eradicator]

You guys, there is an simple vanilla solution. Use diagonal tracks... and you can fit stack inserters between the wagons.

That's nice, thought it does mean you can't incrementally unload a large train into several small ones due to lack of diagonal stations. Also you only get four inserters per wagon. I like weird mechanics though, so i'm probably gonna use it somewhere just for the sake of using it :D.

Otherwise I am thinking that an easy fix to the the core problem could be a "single item type" checkbox that could be applied to chests. When the flag is checked, only one kind of item is allowed and then the chest no longer needs to check every slot... just the total quantity. UPS problem... gone.

If you can single-handedly reduce the cost of chest insertions to O(1) i'm sure the devs would be very interested. Alas i doubt it's quite as simple as you think. But that's a discussion best left for other threads.