Modular Inserters

[just_dont]

After playing a couple dozens of hours, I had this notion that there's a huge "tech" gap between low-tech belted assemblies and hi-tech logistics network assemblies. After you research long-arm inserter, your options for building a belted assembly pretty much peak out (you can squeeze some more efficiency out of smart inserters, but that won't change general layout). Afterwards, you get a few tiers of "faster, faster, and even faster" production, but there's no options to improve layouts.

Then you get bots, and it's pretty much over for belted assemblies, because bots allows SO much better modularity and scalability that belted designs (which are prone to maintenance hell) are simply outdated.

It seems like many people come over this problem, and there are various suggestions here, like "chest on belt" and such. However, I've decided to take a different look on it.

The "bottleneck" issue with belted designs isn't about belts. Belts are simple, well-functioning, and they perform admirably in their scope. The bottleneck is always about inserters -- there simply aren't enough options for them that would allow some advanced designs.

These limitations are (again) already recognized by players, as there are several mods that provides a much wider selection of inserters. However, here comes another problem -- we already have quite too many "inserter" items, and having more just adds unnecessary clutter.

So, instead of having a separate item for each inserter option, we should have a basic inserter plus several special inserter modules that'll change its functioning. To ease the pain of having to left-click each inserter, some universally useful variants, like fast inserter, should stay as separate items available for production.

The lesser used options will all become "place basic inserter, then add modules". I.e. long-arm inserter will be "basic + long-arm module", smart inserter is a "basic + speed + logic module", and so on.

So, how this addresses the belted vs. botted assemblies issue? We can easily have more advanced inserter modules for cleaner layouts. For example: an "adjustable pickup/dropoff point" module(s), which would allow much more flexibility for inserters' placement. Of course, such modules should be appropriately costly, so that it won't become an "overpowered" option.
Given the extra freedom with assemblies' layouts, these advanced inserters could become a nice "middle ground" between low-tech belted designs, and hi-tech logistics networks.

[ssilk]

Hm. I read that now some times. I don't think that inserters are the key. The key for me is the inserter stack-size-bonus. Which means for me that the belt technology from which the inserters can pick up/place only one item and the belts itself, which have a somehow very limited capacity and cannot react to peaks have a gap and that is the reason for all this thinking about "chests on a belt" and all the other ideas around it:
- automatic carts,
- streets,
- one-rail train,
- funicular, etc.

So the target for me is, to replace the inserter/belt combination (not inserter from/into chest or something with stacks, because that is fast enough!) with something, which provides stacks, but can move. That's currently the car and the train. Which brings us automatically back to the above list.

Well, hmmmmmm, maybe a faster inserter would be helpful? I'm not sure. So let's restart!

I think the part till you introduced the inserters is correct and I can sign it.
But I don't think, that introducing more types of inserters solves that problem. The reasons are written above and I think they are correct.
And I don't think, that the problem with the diversity can be solved with combinations of modules. Because then we have a diversity of modules. If you follow your own thoughts, you can see that clearly. And this idea leads to more micromanagement, which must be avoided, because instead of just placing an inserter, I need to place also the modules.

But I think the basic idea is good. If I need to make a headline it would be: Gain diversity by combination of items.

And then we came back to the already existing ideas in this forum, for example the idea to have a car, which can have added turrets. Or better motor, or more armour, or more wheels... Etc.

So here my vision:
When I would be an engineer in this game and I need to repair an inserter, I need to loose all screws and it would reveal the inner parts. If I do that with many inserters we'll see not so many different parts: we have currently two types of motors, two types of arms, some types of circuits, eventually a boiler. And the logical conclusion is, that we can create from that simple parts all existing types of inserters, if we plug the parts back into the basic inserter frame.

No, not that I want to change the complete inserter production, because that means to have new graphics etc. and this will take minimum one version-step (=2-3 month). But what I mean is, that we already can think in that parts, we could say: ok, what is needed to make an inserter faster (for example). We can plug in a faster motor. Or a second! Or more arms. More logic? Why not? Longer arms? No, because they don't match onto the frame...

The difference to the module idea is, that we don't have to click into the inserter, you see the configurated from outside. Every part in/on the frame can be displayed and can be seen when zooming in a little. So expanding it is as easy as building the inserter, but to avoid micromanagement we can build ready-to-use inserters, which are already advanced with a second motor. Or other stuff.

Again, this is much work, of course, and again, I wouldn't change the existing inserters (despite smart inserter) too much, cause it shifts the balancing. My idea is having more or less 4 basic frames, with all needed parts (=already fully functioning) included at fixed positions, instead of one universal frame, which needs to be filled with all the parts. I think this is a good compromise between playability and complexity.

[just_dont]

Well, I can say that I don't like stack-size-bonus tech at all.
For a seemingly "minor" thing it gives such a huge boost that can't be beaten at all energy-wise, or can be beaten just barely with quite a few extra pieces of equipment (which also take space and complicate the layout) speed-wise. The basic idea of "chest + inserter" becomes so good (both compact AND energy-efficient) that other designs fall off quite a bit. Then it's further exaggerated by the fact that this basic "chest + inserter" layout is fully compatible right off the bat with logistics bots.

Stack-size-bonus can be "nerfed" in various ways (I don't think the problem lies with other approaches being too slow, the speed you can achieve with belts and inserters is quite sufficient for all kinds of assembly lines) to give other solutions a competitive edge. It was a train-specific solution, anyway -- it just happened to be ALSO overwhelmingly useful in assembly lines.

But I think the basic idea is good. If I need to make a headline it would be: Gain diversity by combination of items.

Yup, as a sort of grand idea, I'd like to see more of that approach in the game. Especially seeing this entire game is about making complicated systems out of small parts. So it would be extremely cool if some of those parts could be modularized as well (modular defense towers is another idea lying on the surface).

But having all that will likely require quite a few deep changes in the UI (to make object management less of a chore, as you described in your post).

Hm. I read that now some times. I don't think that inserters are the key. The key for me is the inserter stack-size-bonus. Which means for me that the belt technology from which the inserters can pick up/place only one item and the belts itself, which have a somehow very limited capacity and cannot react to peaks have a gap and that is the reason for all this thinking about "chests on a belt" and all the other ideas around it

I also can't help but notice that you still describe a problem with inserters
I.e. it's not like belts are bad, slow, or whatever. It's the fact that an inserter can't pick items off the belt in stacks. So, for example, assuming we have a unit to dump items from belt to storage without an inserter (I don't think we need such a unit, though) -- it won't be a problem any more.

[ssilk]

Well, I can say that I don't like stack-size-bonus tech at all.
For a seemingly "minor" thing it gives such a huge boost that can't be beaten at all energy-wise, or can be beaten just barely with quite a few extra pieces of equipment (which also take space and complicate the layout) speed-wise. The basic idea of "chest + inserter" becomes so good (both compact AND energy-efficient) that other designs fall off quite a bit. Then it's further exaggerated by the fact that this basic "chest + inserter" layout is fully compatible right off the bat with logistics bots.

Stack-size-bonus can be "nerfed" in various ways (I don't think the problem lies with other approaches being too slow, the speed you can achieve with belts and inserters is quite sufficient for all kinds of assembly lines) to give other solutions a competitive edge. It was a train-specific solution, anyway -- it just happened to be ALSO overwhelmingly useful in assembly lines.

Well, you put here some assumptions as fact, as like you have programmed it.
a) The speed with belts and inserters is NOT sufficient enough. To keep up with a fast belt you need about 4-6 fast inserters, but the fast belt transports only about 1200 items per minute. I have built setups, where I need to transport more than 15,000 items per minute.
b) The inserter stack size was not just introduced for the trains

Gain diversity by combination of items.

Yup, as a sort of grand idea, I'd like to see more of that approach in the game. Especially seeing this entire game is about making complicated systems out of small parts. So it would be extremely cool if some of those parts could be modularized as well (modular defense towers is another idea lying on the surface).
But having all that will likely require quite a few deep changes in the UI (to make object management less of a chore, as you described in your post).

You don't need to program it, right?

it's not like belts are bad, slow, or whatever. It's the fact that an inserter can't pick items off the belt in stacks. So, for example, assuming we have a unit to dump items from belt to storage without an inserter (I don't think we need such a unit, though) -- it won't be a problem any more.

Yeah, there have been many suggestions into that direction.

[just_dont]

Well, you put here some assumptions as fact, as like you have programmed it.
a) The speed with belts and inserters is NOT sufficient enough. To keep up with a fast belt you need about 4-6 fast inserters, but the fast belt transports only about 1200 items per minute. I have built setups, where I need to transport more than 15,000 items per minute.
b) The inserter stack size was not just introduced for the trains

a) It's not the atomized assembly line that required 15,000 items/min you describe, right? It's probably a (rather large) assembly cluster of multiple lines. So no wonder that a single belt can't feed it.
b) It is quite indispensable for current trains (but not so for other things). When we'll get cargo-based conditions (instead of simple wait time) for trains -- it won't be required even for trains.

You don't need to program it, right?

Pardon me for being a programmer (though not a game developer), it seems that I tend to look at feature requests with "effort vs. effect" mentality

[bobingabout]

in my opinion, we need a fast long arm inserter. when you get to the high speed stuff like express belts, the long arm inserter becomes useless, which limits your factory layouts a fair bit.

in my opinion, fast long arm inserter would be a good thing, long arm smart inserter too, but also options for long in, short out, or short in, long out, or options for which side of the belt it places items would be useful, I often find myself making some crazy loops in my belts, just so I can move things from the far side to the near before passing another set of factories that put another goods item on the other side of the belt.


You've used smart inserters right? an extra option for "Near side output" would solve one of these problems.
and on a new Smart long arm inserter, not only near/far side of a belt, but also near/far belt for both input and output, would solve everything else.

[ssilk]

in my opinion, we need a fast long arm inserter. when you get to the high speed stuff like express belts, the long arm inserter becomes useless, which limits your factory layouts a fair bit.
in my opinion, fast long arm inserter would be a good thing, long arm smart inserter too, but also options for long in, short out, or short in, long out, or options for which side of the belt it places items would be useful, I often find myself making some crazy loops in my belts, just so I can move things from the far side to the near before passing another set of factories that put another goods item on the other side of the belt.
You've used smart inserters right? an extra option for "Near side output" would solve one of these problems.
and on a new Smart long arm inserter, not only near/far side of a belt, but also near/far belt for both input and output, would solve everything else.

Well, I'm not sure, if you read that: https://forums.factorio.com/forum/vie ... f=6&t=3394
So maybe from that side: Have you ever played with logistic bots. Or do you know about the inserter stack size bonus?

[Cadetlink]

I personally think that we should go with the changing of the different components. it adds an extra bit of customization that can fit with your specific factory build. i think that for speed however, you can get a logic motor that allows for the changing of speed, and a logic chip, that makes it the same as the smart inserter. you could also have a logic splitter sensor, so that you can select what item you want taken (ex. iron plates) and then have the arm filter out 1/3 of all plates to go elsewhere. with this, you could also slide to select how much you want taken. the coolest thing we could do is have huge customization for this thing, that costs very little to unlock, except for things like super fast motors, dual motors, splitters, ext, ext. or have it unlock with the inserter equivalent (long arm enhancement comes when you unlock the long arm inserter, fast motor with the fast inserter, splitter hand sensor with the fast splitter, ext, ext.)

[Garm]

I think we should not.


Want faster inserting speed? build more inserters. Simple.