Remove or Miniaturize Circuit Networks

[steinio]

I for one like the circuit network idea very much especially if they are adding more things to wire into. I would really like the option to fiddle with optimizing and directing some aspects more freely. Of course they should be optional, clearly not everyone likes them.

That said the miniaturization is something I would like to see. It feels very awkward when even some rather simple calculations take the space on the field of a basic oil refinery! The container with its own building grid for integrating some combinators between a pair of inputs and outputs would give the best flexibility in my mind. I even put together a simple picture of one:

Clearly the combinators for this stack-computer should look more like circuit boards/chips than the buildings now in the game but that's the idea anyway.

This looks really great and i would like that.
Looks like an integrated programmable controller and anyone can try and error easy because of rearrangement is easy.

Very good idea.

[ssilk]

To come back to the topic:
There is a good argument, which speaks against this idea: Music.

In Music we have the same problem: There are instruments, that are by far much more complex to play than Factorio. Take for example a harp.
What must be said more? Some like triangles, some like harps. Both make music.

And: If you want to make really good music, use BOTH.

[Zeblote]

To come back to the topic:
There is a good argument, which speaks against this idea: Music.

In Music we have the same problem: There are instruments, that are by far much more complex to play than Factorio. Take for example a harp.
What must be said more? Some like triangles, some like harps. Both make music.

And: If you want to make really good music, use BOTH.

How's that on topic?

[Neotix]

ssilk quote

To come back to the topic:
There is a good argument, which speaks against this idea: Music.

In Music we have the same problem: There are instruments, that are by far much more complex to play than Factorio. Take for example a harp.
What must be said more? Some like triangles, some like harps. Both make music.

And: If you want to make really good music, use BOTH.

How's that on topic?

It'a analogy.

For me that miniaturization is good idea but ONLY on paper. It will be hard to wire more complex constructions, hard to debug, hard to show to others and explain how it work. You will have only box that do something. This idea is only to save some space.
But like ssilk told, to make even better "music" we could have both elements: standard combinators and that from proposition.

[Kirvesnaama]

Certainly I wouldn't propose removing the regular combinators either, they are useful and a few won't take that much space. On the debugging and showing to others: I would argue that a mess of 20 combinators with green and red wires here and there is not very presentable as is either without explanation. At least with the boxes (that you can separately open and show off) you can procedurally explain from overall to smaller details under the hood.

At least after putting together a 5-digit number display I was a bit sad that the operations took more space than the number lights themselves... A separate advanced technology to allow building boxes and new integratable combinators shouldn't remove anyone elses fun either.

[civ00000004]

I for one like the circuit network idea very much especially if they are adding more things to wire into. I would really like the option to fiddle with optimizing and directing some aspects more freely. Of course they should be optional, clearly not everyone likes them.

That said the miniaturization is something I would like to see. It feels very awkward when even some rather simple calculations take the space on the field of a basic oil refinery! The container with its own building grid for integrating some combinators between a pair of inputs and outputs would give the best flexibility in my mind. I even put together a simple picture of one:

Clearly the combinators for this stack-computer should look more like circuit boards/chips than the buildings now in the game but that's the idea anyway.

Suggest this to become more avaliable tech name "Micro Circuit" that allow player craft 'Circuit board Combinator' and use old Combinator to make the new micro Combinator one to use inside 'Circuit board Combinator'.

[xuhao]

Hey

Im gonna post these thoughts here, since this is the closets i could find to this topic using the search function. Pls move this(or ask me to move it) if its out of place.

Ive played a few hundred hours of factorio so far, and I find that the circuit network has an annoying setup.

- I love the idea of a circuit network that you can program to do all kind of stuff, but I must admit that I dont like the current system for the following reasons:

1. The base layout uses resources as symbols, which can/will be confusing to some. It is perhaps a sensible system to an engineer or a programmer, but to the average joe I think its far from intuitive.

2. the system deals with base algotithm tools only! This resembles, at least in my oppinion, the core parts of a processing unit and ONLY the core parts of a processing unit. It means that you can basicly do anything!!! which sounds cool, but I dont think it is as cool as it sounds. You start from a complete scratch. Only the simplest networks can be accomplished without using multiple combiners and such. And the simple networks are imo easier and more cost efficient (both in ressources, space and energy) with smart inserters and boxes with a set limit.
More usefull network functions can only be done by creating a string of calculations/ algoritmh, which means many machines, lots of space needed and increased energy consumption.
the actual setup of such a network brings me to the 3rd reason of why i find the circuit network annoying in the game:

3. Gameplay. When i want to create a network the chain of thought becomes the creation of an actual algoritmh. To me this is mathematics and programming brought into the game without much addaption. I can do things that is completely out of sync with the rest of the game, just following standard algorithms. Hell, I could build a Texas Instrument TI 40 calculator, or an old ibm computer like the 3/84... its theoreticly possible because I have the basic tools needed to do so. But to what purpose???

Edit:

When i play my purpose of using the network is to create an efficient factory. I dont think you should remove the cirquit from the game, I really like the idea. And Im sure some players like to build crazy stuff, like the good old ping-pong tennis game made of flashing ligths and square miles of algorithms . What I would like is something more than the current basic tools, so i can create something factory-efficient a little more easily. Herpaps some tools that felt more adapted to the game.

[ssilk]

Basically I nod to this, but things will get (hopefully) much better with v0.13, which will be released soon. See latest FFF.

And yes, this would be better discussed in General, not here, but I have no time, need to go.

[Harkonnen604]

My 2 cents on why after 120 hours of gameplay I haven't touched circuit networks yet:

1. It's strange that it appears as some very early tech, but actually you can use it only when you can produce smart chests, and they require steel, and by this time you already have at least belts/inserters production automated. So main use of circuit network as of production buffer limiter is actually unusable until some not-so-early stage, and you already stick to red cell marking in target wooden chests instead of putting red wire and smart inserter. Perhaps if chests would have no red-cell marking possibility, more players would be enforced to use circuit network and love it. Afaik some of that comes in 0.13 with everything becoming circuit-connectable, that's good news if wooden chests are now CN-connectable.

2. By the time you get logistic network, you start caring more about limiting whole reserves in your logistic network instead of caring for just one target crate on circuit network. The reason - when you disassemble something or use logistic trash slots, all those products go to some random storage chest instead of passive provider chest near factory output. So again smart inserter near factory starts operating on logistic network conditions instead of circuit network conditions.

3. I once though where circuit network would be useful in my cases - that was to supply remote outposts with repair packs, turrets, walls, etc... without wasting 50 turrets in dedicated train slot. Circuit network would come in handly, but it would require me to drag red wire through all those poles down to the outpost. And if I have two outposts along that single power line, I'd have to put a green wire. And remember which wire for which outpost. Or reassign turrets to locomotives via some combinator to have it on same red wire and remember that locomotives are actually turrets for outpost B. Or even drag another power line just for the sake of extra signal with correct name on it (Turrets). That appeared too quirky to remember and manage, so I didn't even try

What I would propose regarding circuit networks:

Compared to logistic network they are not that useful. Really. I mean as a separate bus. Why need red/green wires at all while everything in factory is already covered with power poles with copper wire on them - can it be high-frequency information transport medium instead of that wiring? Because in early game you care about some particular crate. Later you care about logistic network, and frequently about REMOTE logistic network. So I'd replace all that wiring with one simple decision:

a) Allow to name chests (via chest interface).
b) Allow to name logistic networks (via roboport interface).
c) Consider everything to be on the same red wire if it's on the same electric network.
d) Red wires may still remain to connect distinct electric networks to be on the same circuit bus.
e) Now - the final part. Allow assembly machines and inserters to have arbitrary formula conditions, e.g. "Outpust1.Walls < 50" or "(Trains+Wagons)*Beacons > 20" written with text. Clicks on icons may substitute names or even embed icons in text editor with autoreplacement.

That will solve combinator compexity problem as well. For when I see screenshots full of combinators that reminds me of blueprints from Unreal Engine where editing shaders was simplified so that artists could do that, and eventually stuff like "A + (B+C)*D" takes the whole screen like a block schema, and becomes even harder to perceive and draw than original formula would be, even for artist, not programmer.

The point here - giving a full programming language is certainly overkill, that's why UE, Maya and other software all have of blueprints to tune materials by non-programmers, but still any non-programmer is capable of writing a formula and actually does that in his head before starting to drag arrows in such editors. That actually was resolved in UE later as "math expression" node which binds inputs and outputs, schemes have become much better since then. And I think this is what Factorio actually needs. The format may be as simple as "ItemType" (signal type) or "ChestName.ItemType" or "LogisticNetworkName.ItemType" - perfectly readable, editable and resolves a lot of things. And everything is on the same bus already through power poles with sort of high-voltage DSL.

Sorry for writing a lot about thing I did not even try to use yet, but that sorta explains why I didn't.

[ssilk]

TL;DR: many decisions to this subject have been already made.


I just repeat some very old arguments, cause the idea with the programming language and/or the minaturization (factory in factory) has been discussed a while ago.

The results for the circuit network was:
- No more artificial border between circuit and logistic network. This is done now with v0.13.

- Also the ability to have more networks. There where some threads about "radio network" for example. There is also a mod for this, but I think the current solution is not really useful, cause you need to have more than one "channel". I think v0.13 enables that way, but is not implemented yet.

- The questionable need for red and green wires. viewtopic.php?f=6&t=14465 Make colored wire "free"
and others ( search search.php?keywords=free+wire*&terms=al ... mit=Search ) There is just no real reason for them to need them to craft.

- If you need complex function there are two ways: Use a blueprint or write a mod that creates a new type of combinator. It makes no sense to have that inside of another combinator. One reason is for example that it is (slightly) more afford to simulate that inside of an own surface (that is the factorio-term for an own factory in factory), instead of simulating it in the same surface, cause you need to transfer the input and results between the surfaces. Sorry, if this is a bit technical. See in the wiki about "surface" or just ask.

But I would prefer to have this: A combinator, which can load the program running on it, cause that would be the most elegant way to "miniaturize circuit networks". I mean: Things like NAND, NOR etc. are nice, but in Factorio you don't need that very much. What you really need is stuff like a regulator (https://de.wikipedia.org/wiki/Heizungsregler), a timer, a counter that can compare, a controller, that is able to calculate an alarm-signal due to the current power-state (including the accus) and so on.

Not those simple binary logic stuff, more like macro-programs. Something, that you really want to write in a higher programming language.

I mean it like so: In real industrial plants (and I think Factorio should be compared a bit with it) they also don't "program" every switch, they use some kind of controller for that. And the controller switches normally on/off as a switch, but not, if the factory is for example in alarm-state or if we have lack of power. That is some complex logic, which I don't want to program X times, cause I really want, that the factory works some kind of "standardized".

[Aatch]

I'm here for the "miniaturisation" aspect, as I was essentially going to make Kirvesnaama's suggestion myself. While 3/4 combinators aren't too bad, and you can do a fair bit, as the complexity rises, so does the space and that can be incredibly frustrating when you find that you don't have quite enough space and need to rethink everything.

By making the "micro combinator" optional, and require its own later research, you don't hurt newer players by throwing the extra complexity at them while allowing more experienced and/or advanced players to build more complex systems a little easier. If the micro-combinators are just built from regular combinators (essentially), then you also have the nice property that micro-combinators are technically unnecessary, anything you can put in a micro-combinator, you can build in full from the composite parts.

[British_Petroleum]

A better solution is your miniaturisation suggestion. Have one building that is multipurpose, then when you open it, have a grid (similar to the modular armour) where you can add different gates, and draw your circuit in there. That way with just 1 tile, you can create what you'd currently need a few dozen to do.

This would be so awesome

[Xeteth]

I don't really like the miniaturisation to be honest, it seems like a whole lot of work to change something that I don't really think is broken. Area to build is one of the most free things in Factorio and honestly making something like this I believe is actually really awesome to look at in your base, especially seeing things like the data busses when you mouseover the combinators and it highlights the wiring

[Rseding91]

How about an official response to this idea/suggestion:

No.

[siggboy]

I don't really like the miniaturisation to be honest, it seems like a whole lot of work to change something that I don't really think is broken. Area to build is one of the most free things in Factorio and honestly making something like this I believe is actually really awesome to look at in your base, especially seeing things like the data busses when you mouseover the combinators and it highlights the wiring

What about the demo save / blueprints of your train scheduler Xeteth? Is that ready to be released by now?

How about an official response to this idea/suggestion:
No.

Please tell me it didn't take you 16 weeks to come up with a one-letter official response .

I'd like to be able to move combinator around, without having to deconstruct/reconstruct them. Right now, refactoring circuits is very clumsy.

I don't see how the miniaturization would make things worse, at least it would make it easier to provide an appropriate UI / quality of life for more complex circuits.

[ssilk]

No.

I really like short messages. But that is overdone!
Something, you can stop, please, cause it makes more problems, than it solves.

[siggboy]

No.

I really like short messages. But that is overdone!
Something, you can stop, please, cause it makes more problems, than it solves.

Give the man a break, he has clocked almost 600 hours in 2 months.

Wube Software needs a community manager.

[Nexela]

Wube Software needs a community manager.

$$ spent on a community manager is less $$ spent on factorio development. Besides, No with cherries on top is still the same as No.

[siggboy]

Besides, No with cherries on top is still the same as No.

That's what Wube appears to think as well. I could not disagree more. But maybe we'll get DT as the next POTUS, and the paradigm will get accepted far and wide .

Also, the community manager could still be a developer, it just would have to be one with social skills (maybe even a woman, *gasp*).

[Harkonnen604]

Cherries update player game model, so instead of arguing and feeling shut up s/he starts agreeing and saying those same clues to others in viral form. But yeah, initial cherry is sometimes long to write