Fixing NPE - Full Burner Start

[Rythe]

Extension of this thread where the problem with the burner phase was brought up.

The problem with Factorio's beginning is that it gives the player a critically flawed and incomplete toolset in regards to burner powered buildings. A new player starts with burner buildings so that the simplest power mechanic is their introduction: Put coal or wood fuel into building -> building works. Simple. Direct.

The incomplete toolset is that you can't use this burner powered mindset to automate production of anything nor can you do any research. Both can only begin when you build a steam engine electrical plant (water + fuel -> steam -> electricity -> building works). So to complete the burner phase toolset, we'd need to have Assembly Machine 1 burner powered and have a Research Lab variant that was burner powered. One building changed and one new building variant.

The critically flawed part is that burner inserters aren't quite where they need to be to be fueled by automation. I'd thought this was fixed when burner inserters were able to grab fuel for themselves, but the problem still exists in that they don't pull fuel from a building's fuel stockpile for themselves. So the burner inserter that pulls a product out of a smelter/assembly can't fuel itself in a reasonable way (you'd need a separate inserter and fuel line for it, which is silly.) The fix here is to simply let a burner inserter fuel itself from the building's stockpile it's pulling finished products from.

Now the Assembling Machines are slotted to lose their ingredient limits, which removes the 2 item limit on the AM 1. Good in that the AM1 stops being barely useful, bad in that it opens up a lot of more complicated recipes on the player pretty much immediately. This can be ameliorated by adjusting the initial recipes to use a simpler ingredient set, by which I mean that burner phase items should not use the green Electronic Circuit (Copper -> Copper Wires + Iron Plates -> Electronic Circuit -> Intended Product). Which leaves us with Iron/Copper Plates, Gear, Iron Sticks and the Stone Furnace.

To separate out the AM1 and Burner Lab recipes from Burner Drill (Iron Plate, Gear, Furnace)...
Burner Lab: Pick 2-3 from (Iron Sticks | Copper Plates | Iron Plates | Gear), Furnace
Assembling Machine 1: Iron Sticks, Iron Plates, Gears, and Furnace

Which would make the AM1 the most complicated recipe to automate and thus how a new player 'wins' the Burner Phase. Here's how complex that would be if we didn't use the splitter because electronic circuit requirement. Easier if the first splitter's recipe changed (gears instead of circuit). Slightly easier if we reasonably expected that the new player would just have multiple coal lines instead of using inserters to split a single one, in the off chance that the first splitter remains circuit locked. But really, swap circuits for gears and then the entry level of logistics (inserter, belt, underground, splitter) becomes complete for the burner phase too.

BurnerFactorio.png (971.06 KiB) Viewed 11852 times

The side benefit is that Iron Sticks are suddenly more used. The major benefit is that this level of complexity is a lot less than powering and automating the AM1 recipe currently in the game.

[planetfall]

This discussion is fascinating to me because I always see the burner phase as pure bootstrapping, something intentionally janky and needing significant manual intervention so that the value of "true" automation is made more clear. Hence why it's obsolete so quickly. Personally, I've never seen value in the burner inserter at all, outside of the tutorial that has to explain rudimentary automation to the player without first explaining electricity. To each their own, of course; arguably, since my first world had me handcrafting blue science for more than half the play time, the existing burner phase wasn't enough to goad me into setting up proper assembly lines.

Adding a burner assembler begs a rather nasty question, though: why is it easier to make an electric assembling machine than an electric furnace? And why does the assembler use basic circuitry but the furnace uses advanced circuitry? Surely, given the complexity of the tasks involved, they should be switched?

Yes, yes, obviously there's the whole pesky game mechanics angle to it, but I would raise my eyebrows at this as a new player. It seemed intuitive at first that an assembler would have to be electronic by its very nature, and the electric furnace is as gated off as it is because you need significant technological improvements in order to get the temperatures needed for smelting out of a heating element. Part of my objection to the burner inserter is that it feels like something like that shouldn't exist without electronics, and as a result it feels a bit gamey and immersion breaking, but again, to each their own.

(Though, to take the "pesky game mechanics" perspective, what exactly does it add to make players design assembly lines with dedicated fuel inputs that will be obviated by early red-science tech?)

But if the furnace, inserter and assembler are all things that get upgraded from burner to electric versions, I seriously balk at the idea that it's red science to reinvent complex and delicate machinery, but blue science to reinvent a box that gets hot. There's some precedent for this: Minecraft tech mods (of which Factorio is somewhat of a descendant) frequently make an electric furnace one of the earliest machines you can build. That, plus the inherent nature of crafting vs. smelting, plus the fact that from the very outset crafting is something that requires manual intervention while smelting can be left in the background, combine to make the thought of a burner assembler feel rather implausible.

[darkfrei]

It looks like very good three tiers progression:
First tier of entity has burner and needs coal. Speed = 1
Second tier has also burner, but speed = 2, now we have it by steel furnace.
Alternative second tier is electric powered entity but also slow, speed = 1.
And third tier has electric power source and speed = 2.

It can be applied to mining drills, furnaces, assemblers, inserters and may be also cars, trains, offshore pumps when it is possible.

[bobingabout]

I would not be opposed to a burner assembling machine to start. I would not be opposed to what is current Assembling machine 1 being burner powered and available from the start with 2/3 being what you have to research. Lets face it, the tree wouldn't really suffer if you just knocked out assembling machine 1 from the start of it. It's literally the cheapest technology in the list, and the first thing most people research.

[eradicator]

Hi, it's me your friendly Advocatus Diaboli from around the corner, let me tell you something :p...

Imho anyone demanding burner powered assemblers should play at least the first 5 hours of Xander Mod. While not a perfect comparison due to increased recipe complexity it does offer a lengty burner-only starting phase. While this is great fun for experienced players i think it is easy to realize that the increase in routing-complexity is non-trivial. The starting part of factorio is supposed to be simple, but short. But you still need to use asm1 to automate some basic components before you can get asm2 reserach done, and requiring players to competely rebuild their base after that because all those fuel lines are no longer needed is far from making the game more fun for beginners. Double-sided belt-use is also something that most people only learn to do after some time, thus making it an absolute requirement to do anything at all is a bad idea.

[Mike5000]

The starting part of factorio is supposed to be simple, but short.

Sandbox mode currently offers options to start with full research and various resources. These options would be extended to freeplay and would allow a choice from several starting tiers - e.g. stone age, iron axe (as in 0.16), electrical, construction bots, logistic bots, and all finite research.

Burner phase assembling would not be long rows of belts and assemblers but something simpler - typically a belt assembler, a red science assembler, and eventually a generator assembler together with one or more gear assemblers and a few burner labs. Those who wish can figure out fun uses for belts here but the only important use of belts in this phase is for coal. Most players would feed their AM1s from wooden boxes of mixed coal, iron, and copper.

In addition to being burner-powered, AM1s (like furnaces) would be able to accumulate a full stack of product items internally.

As a gate to the electrical phase, generators would be non-craftable (like engines).

[Rythe]

This discussion is fascinating to me because I always see the burner phase as pure bootstrapping, something intentionally janky and needing significant manual intervention so that the value of "true" automation is made more clear. Hence why it's obsolete so quickly. Personally, I've never seen value in the burner inserter at all, outside of the tutorial that has to explain rudimentary automation to the player without first explaining electricity. To each their own, of course; arguably, since my first world had me handcrafting blue science for more than half the play time, the existing burner phase wasn't enough to goad me into setting up proper assembly lines.

...combine to make the thought of a burner assembler feel rather implausible.

The burner phase currently is pure bootstrapping. But this means that you start your game with some burning drills, stone furnaces and a whole lot of hand mining/transporting until you have the resources in hand to manually craft an electric plant and research lab. Then it continues to be a whole lot of hand transporting, possibly a little more mining, and hand crafting everything until you've bootstrapped straight into where automation can actually start.

This is, indeed, why you spent your first run handcrafting blue science for more than half the playtime. The gap between game start and the start of automation (incorrectly) tells you that handcrafting is how much of the game works.

And there's always going to be some implausibility in a game where one dude hand crafts his way up to a nuke plant or space rocket. The question is whether the game mechanic consistency/smoothness is enough to have you naturally gloss over most/all the implausibility. Factorio isn't quite there yet.

Now to pull in the "Doesn't this make the transition to electricity harder?" question.

Yes, but you're overthinking it. For the new player, they'll be loving not having to run coal everywhere for their inserters, and the new player isn't going to have a whole lot of factory to fix unless they're really digging it. The first thing is they'll replace burner inserters and clean out most of the coal from their initial little makeshift factory. So coal will be eliminated for everything but the AM1s. For AM2s, they'll most likely just start a new factory expansion that is pure electricity. It's more organic than most are thinking, I'm pretty sure.

For the veteran player, a burner start means:
You can automate iron/copper plates immediately, which saves a lot of the hand carrying and gives an incentive to burner drill it up a little more. Particularly for iron.
Replacing burner inserters for standard electric ones in the smelting lines doesn't require the redesign/placement of anything nor removing excess coal. It's all the same unless you've got a long-handled inserter design you start with.
You can automate belts, underground belts, and red science research before building a power plant. You can do this without having to mine up a bunch of copper for circuit cards and power lines.

So the extended hand mining/hand transport awkward bootstrap phase can be broken up for the veteran where they create a mini factory that produces 3 or so beginning essentials and does starter research before they run off for the electric plant and the start of their first electric factory which would be the start of their 'real' factory.

The wrinkle is the jump from AM1 to AM2, because the veteran would have to deal with a few coal lines going into their first electric factory if they start it with AM1s. That's where a full burner start is a little more annoying. ...Ignoring that AM2s are an 80 red science research right now so skipping to them is already trivial.

[Oktokolo]

The critically flawed part is that burner inserters aren't quite where they need to be to be fueled by automation. I'd thought this was fixed when burner inserters were able to grab fuel for themselves, but the problem still exists in that they don't pull fuel from a building's fuel stockpile for themselves. So the burner inserter that pulls a product out of a smelter/assembly can't fuel itself in a reasonable way (you'd need a separate inserter and fuel line for it, which is silly.) The fix here is to simply let a burner inserter fuel itself from the building's stockpile it's pulling finished products from.

I fixed that in Inserter Fuel Leech. All inserters also fuel their targets with fuel from buildings and burner inserters also fuel themselves from buildings.

[Rythe]

Hi, it's me your friendly Advocatus Diaboli from around the corner, let me tell you something :p...

Imho anyone demanding burner powered assemblers should play at least the first 5 hours of Xander Mod. While not a perfect comparison due to increased recipe complexity it does offer a lengty burner-only starting phase. While this is great fun for experienced players i think it is easy to realize that the increase in routing-complexity is non-trivial. The starting part of factorio is supposed to be simple, but short. But you still need to use asm1 to automate some basic components before you can get asm2 reserach done, and requiring players to competely rebuild their base after that because all those fuel lines are no longer needed is far from making the game more fun for beginners. Double-sided belt-use is also something that most people only learn to do after some time, thus making it an absolute requirement to do anything at all is a bad idea.

There's two assumptions here. One is that the new player won't just dump 50 coal into an inserter and then be able to forget about it for a very long time. Burner inserters don't use a lot of energy, so their coal stack lasts a long while. The coal + component dual belt line is an advanced trick. Giving a new player a reason to learn and use it early on is a good thing rather than a bad thing.

The coal + component belt discovery early on will tune their design thinking so they'll more naturally run belts with two components as they progress through the game.

[Rythe]

I've been thinking about the burner -> electricity transition a bit, and it could step out with some tweaks to research and recipes that would make some sense one way or another.

Two techs:
Steam Power would unlock Copper Cables, Small Electric Poles, Steam Engines, Boilers, Pipes, Pipe to Ground, Offshore Pump, and Electric Mining Drill*.
The Offshore Pump and Electric Mining Drill recipes would lose Electronic Circuits for Copper Cables. (It was always weird that a simple pump had a circuit card, where Copper Cables makes more sense given wire coils in electric motors).

Circuitry Tier 1 opens up after and would unlock green Electronic Circuits, Electric Inserters, Long Handed Inserters, and maybe Splitters.
I can still vaguely remember how much of a thing it was for standard Splitters to be added to the game once upon a time. Might be worth having the same sort of effect of giving them to the player early into electricity too. Or not. Either way.

There's one not insignificant problem here, and that is a player ends up with a power plant that has nothing to power. Makes some sense from a research advancement perspective. Not great for gameplay, but with one little caveat that a player would be figuring out/setting up their electric plant while researching stuff to use it for. The other slight bonus is that this gates more items so that a player doesn't have all these options right off the bat. Helps focus them in incremental steps.

Haven't figured out a clean way to do this. Either a player ends up with a whole bunch of things dumped on them at game start/Steam Power research, or you get an odd little gap at the very start of electricity.*

Edit: Oh! And underground belts would be shifted to Automation 1 with the AM1.
*Edit 2: Duh

[BuilderOfAges]

There's one not insignificant problem here, and that is a player ends up with a power plant that has nothing to power.

You could unlock electric labs together with the steam engine, but keep the electric AM2 locked behind automation research. That way the player can set up electricity, add a lab and hand-feed it bottles, or re-route existing red science automation to the new lab. That has two advantages: the labs don't need to be refueled anymore, and you can get the AM2 research done faster (assuming the electric lab is faster than the burner one, which is what I would expect).

[BuilderOfAges]

Imho anyone demanding burner powered assemblers should play at least the first 5 hours of Xander Mod. While not a perfect comparison due to increased recipe complexity it does offer a lengty burner-only starting phase.

Perhaps AAI Industries would be a better proxy for the experience being described here. It has burner labs and assemblers, but doesn't complicate recipes (except by adding motors, but you can easily see how the game would play without that change).

[Rythe]

There's one not insignificant problem here, and that is a player ends up with a power plant that has nothing to power.

You could unlock electric labs together with the steam engine, but keep the electric AM2 locked behind automation research. That way the player can set up electricity, add a lab and hand-feed it bottles, or re-route existing red science automation to the new lab. That has two advantages: the labs don't need to be refueled anymore, and you can get the AM2 research done faster (assuming the electric lab is faster than the burner one, which is what I would expect).

I'd prefer the electric lab to be its own research so the research speed boost of the electric lab over the burner lab could be a thing, and in that way, replace the research speed boost generic tech tree powerup. Thinking about the research speed boost research, it's kinda funny that you have a game about building a factory in order to do research, but within there is a research tech that's basically allowing the player to get by with a smaller factory.

I'd say the answer is to have Steam Power unlock the Electric Mining Drill, which turned into a bit of a duh moment for me. Swapping circuits for the lesser Copper Cables in the Electric Drill recipe feels like cheating a little, but at least we'd end up with some synergy with the Offshore Pump that way.

[Rythe]

Some additional thoughts:
The intro having janky/worthless automation didn't teach you to appreciate it when it eventually worked. It told you that automation was more trouble than it was worth so you learned the habit of hand crafting all the most complex things because that's what you did for the first couple hours because that's what you had to do for the first couple hours.

Another plus to AM1s being burner powered: Unlocking plastics use to be this super weird thing where you suddenly needed coal for a crafting recipe, so you'd need to interrupt your midgame to go connect a coal patch to wherever your oil processing ended up. This could happen earlier in gameplay when explosives became a thing, but AM1s being burner powered means your factory starts with coal being fed to it so explosives and plastics won't cause such an interruption anymore.

It's a bit more helpful dynamic than the steam plant or smelting area being coal fed too. Sometimes the water + coal of steam power meant that your production area was a ways from your power area so it could end up that coal was also far away from your production area, particularly since it's easier to use power poles to cross distances than belts. Smelting areas need coal, but your factory only needs the copper and iron plates, which meant that it seemed fine for the factory to end up on the far side of your smelting zones from the coal patch. Given all that and plastics needing coal, AM1s being coal fired organically nudges a player into smarter initial factory placement.

[eradicator]

@Rythe:
I'm slowly getting the feeling that what you really want isn't more burner stage, but less handcrafting (possibly without noticing yourself). Which is a goal that sounds easier to achieve by...limiting handcrafting. Disallowing handcrafting of intermediates would be a good start. In a game about automation that doesn't sound too unrealistic. Ofc it'd require some recipe tweaks to work, but if handcrafting simply never was an option players would likely feel much less resistance to non-handcrafting.

The coal + component dual belt line is an advanced trick. Giving a new player a reason to learn and use it early on is a good thing rather than a bad thing.

The coal + component belt discovery early on will tune their design thinking so they'll more naturally run belts with two components as they progress through the game.

If you can successfully make them learn it easily. Yea. But that's a tutorial problem, not a content problem. And i think you're overestimating how many tricks you can stuff into a new player in the first hours.

[Rythe]

@Rythe:
I'm slowly getting the feeling that what you really want isn't more burner stage, but less handcrafting (possibly without noticing yourself).

1)AHAHAHAHANO
2)Hand crafting is needed for recipes a player uses rarely, particularly complex ones. And given the rarely, it only works well if the player can chain through most of the intermediates from base components themselves.
3)"Hand crafting a pickaxe first thing is how a player understands their factory is an extension of them." - paraphrased. I enjoy Factorio more when the factory is an extension of the player rather than the other way around. Hand crafting tanks helps keep that theme intact.
4)I'm exhausting the argument as a way of exhausting the dynamic system analysis in play here. The goal of which is to go from a rough bootstrap intro that dissuades a player from automation to a gentler curve that encourages automation from the get go. Because Factorio.

If you can successfully make them learn it easily. Yea. But that's a tutorial problem, not a content problem. And i think you're overestimating how many tricks you can stuff into a new player in the first hours.

It's more than a tutorial problem. A neat trick isn't a neat trick unless it's actually useful in the actual gameplay (or for showing off). The sooner an opportunity for that presents itself, the better.

This is also why I barely use logic networks, because I've only ran into three things where they provided enough of a benefit to bother with.

And it really depends on the player for how fast they can absorb things. Given Factorio, there's probably a lot of players that this neat trick is pretty simple and quickly integrated. The greater point is to give the player a reason/opportunity to feel a little clever early into the game.

[eradicator]

2)Hand crafting is needed for recipes a player uses rarely, particularly complex ones. And given the rarely, it only works well if the player can chain through most of the intermediates from base components themselves.

Take a close look at all the buildings, they're mostly made of green circuits, gears and metal plates. Enforcing machine crafting just on the green circuits would have a large impact on the beginning of a factory, but almost none on later ones, because you don't hand-craft circuits, they're the "basic" stuff that you carry around anyway. More complex intermediates (red/blue circuits, engines) are already machine-only.

It's more than a tutorial problem. A neat trick isn't a neat trick unless it's actually useful in the actual gameplay (or for showing off). The sooner an opportunity for that presents itself, the better.

As long as it is an actual opportunity and not a requirement. This is the part that i find difficult to balance.

This is also why I barely use logic networks, because I've only ran into three things where they provided enough of a benefit to bother with.

I think logic networks are quite fine like that. Having a "thing i don't understand yet, but maybe later" isn't a bad thing, it adds a bit of a "mistery" factor to the game (for lack of a better term - things one fully understands in every aspect are more boring than those that have unknown parts left). Most people build really long belts before they understand trains, and when they start understanding trains they love them. It's imho just a kind of scaleing mechanic that you can decide to master or not. And circuits are like that. You *can* use them, but you're also just fine without. They're a precision tool if you want a precision solution. Not something that should be used all over the place.

And it really depends on the player for how fast they can absorb things. Given Factorio, there's probably a lot of players that this neat trick is pretty simple and quickly integrated. The greater point is to give the player a reason/opportunity to feel a little clever early into the game.

Maybe the best way would be to have the very first tutorial/mission start with a sushi belt, due to broken machinery or whatever. Then run it through a splitter or something to be "sorted" into a dual belt. In "first contact" situations people always make assumptions during the process of understanding. So if we simply prevent invalid assumptions from occuring in the first place. Maybe they'll still start with one-item-belts, but they will inheritely know that there are no item type limitations. So, yea. I still think it's a tutorial problem.

____
Btw, i'm not against a longer burner stage. I like more content. I'm looking at it and wondering if it's the appropriate solution for the problem at hand. And i hope wube has enough data on how big the "problem" (i.e. new players getting frustrated) actually is.

[AileTheAlien]

I wouldn't mind a direct limit on hand-crafting items, but I think burner assemblers and burner research are better, because it doesn't force the player to play a certain way. They can come to the realisation on their own, that some recipes are going to be made in bulk, so they should automate them and not waste their own time. Additionally, the burner items in the game right now are only useful for a very brief time, before you get your electrical things researched and set up. That makes them fairly pointless, and clutter the game's menus. So they should either be removed from the game, or improved; I vote for improved. :)

[featherwinglove]

So, I've read the whole thread, and it's pretty interesting how opinions contrast with my masochistic mod packs, most of which disable both hand crafting and hand mining (yes, this is the same "featherwinglove" who has kinda appointed himself the axe's champion almost by accident.) Despite having an "infinite" ingredient burner assembler in these mod packs, I usually ditch it for electric ones as quickly as possible, mostly because I've designed the packs to encourage that. (It's not all that hard to go into mod recipe files and tweak things.)

With hand crafting completely disabled, I carry around a bunch of assembly machines, including one capable of making everything interesting at that point in the game, and if I need to "hand craft" a bunch of somethings, I set up a workshop and do it, much as others (I've seen AntiElite do it a lot in his speed runs) grab a bunch of intermediates so that they're only hand crafting final products.

In my very first free game, the first thing I felt compelled to automate was green circuits, because I noticed I was spending most of my hand crafting time making green circuits, and they were used in pretty much everything in the early game. I forgot exactly what I was going for, but I did try to automate level 2 science before researching Automation 2, and was setting the recipes and then, "Wut, I can't set the inserter?" I think I switched it over to Automation 2 at that point just so I could get the machine that would make the inserter.

To the best of my recollection, I have never hand crafted a blue science pack, even in my first game. I was too busy hand crafting things like AM2s and oil refineries to bother, and I realized that my labs were going to be consuming a lot of SP3, and I also realized that I could finish my SP3 build long before I ran out of SL2 research, although I did get into relatively unimportant weapon upgrades.

Rythe talks about that transition point where you seem to become the slave in the factory rather than the factory being an extension of your capabilities. I felt this when I saw the first rocket parts go into the silo, having no idea how many they would need. My rocket part factory was tiny, probably just one or two AM2s per part type. And 1%. UUUGGGHHH. It really did feel grindy, like it was demanding of me like a starving puppy at that point, not as much like my own creation. But understanding the problem, I didn't get that feeling in future games, I just built an appropriately sized factory for the pace that I wanted.

[T-A-R]

Tldr: I suggest to allow AM1 to run on coal or steam.

Although i really like the idea (What about steam powered bootstrappping:p), the burning phase probabily won' t get changed in this state of finishing up to 0.17. Since there are good arguments for that, a full burner start (phase) would maybe fit better as mod-content since devs did not mean it to be a phase. You need electricity to run your factory.

They are just tools to set up a electric line easier, learning you non-trivial skills and bad habits. The burner miner is obsolete in minutes and cant be recycled. (You should dump burner mining asap). The game would not change much if you would start with a single steam plant and some poles.

There are seven (7) types of inserters but only two types of assemblers left to fill you infinite playground. And of course a chemical assembler for mixed fluids. (why are batteries and e-engines not produced in the same machine??) Why so many inserters and so few assemblers? If you want to guide the player by unlocking tech i would love to see the differences between AM I, II and III would be increased, instead of simply removing a assembler and ingredient limits. It would be cool if a obsolete AM still has some niche-functionallity, like the burner inserter does now. There should be a decision, a reason for different machines to form a complex factory.

If the ingredient limit is dropped, burning/steaming could be the unique feature to give AM1 unique value. It encourages different approaches of solutions. It would make electricity a true decision instead of a obvious choice.

Keep it electrical, but allow it to run off grid on coal or steam. You'll go electric most of the times, but when you want no poles, you can. This would also help speed up bootstrapping, preparing for proper electric AM2 lines

Not needing a life line, or a second electrical network for your outposts, could lead to interesting designs. To keep three types i support the idea of burner assemblers, hybrid assemblers or steam assemblers as replacement. With the removal of the ingredient limit, we could theoretically burn-assemble a satellite;)
I think the labs should require electrical energy, to not alter the intended gameplay too much. But electricity should not be manditory everywhere while there are burner instruments available.

AM1 running on steam power would be a cool feature to explain fluid handling with a single fluid before learning fluid ratios in the refinery.

Additionally the idea of burner inserter speeding up by type of fuel, could keep AM1 relevant by upscaling speed without modules.