Quality

[grand igloo]

true

I’m almost sad that the API makes everything so robust that there’s hardly ever any exploitable inter-mod jank

it either doesn’t run or it has no jank

but in all seriousness I kinda wish they kept some processes that’d be best controlled/optimized with circuits.

Doable without them, in a reliable way, but optimal with them, y’know?

[hoary current]

red belts is a good iron sink tbh

[daring siren]

Definitely, considering the free 50% prod.

[vernal bay]

If the green cicuit assemblers are Q2 and the copper wire assemblers are Q3 it will mess up the ratios. And the same problem applies up/down the chains, if whatever is consuming your green circuits (science of whatever) is designed to consome the exact amount of green circuits your Q2 circuit maker providedes upgrading it to Q3 won't provide any benefit, and if your smelting is designed to supply the exact amout of resources your Q2 maker consumes then it won't have enough iron/copper at Q3.

You can't really use prod bonus instead of quality bonus if you want anything other then commons anything with quality modues (even with common recepie) outputs all qualitys so you can't just "use prod mods instead" if your goal is to have a supply of Q2,Q3 or Q4.

[jaunty citrus]

I don't know why your goal would be to only have a supply of Q2 items. The Q3 items you get along the way are strictly better and can be used in places where they'd gain more value

Obviously mixing qualities changes the ratios, so does mixing the levels of modules. Don't mix qualities if you haven't designed the build to do so

[daring siren]

It's generally better to prod where possible, and only quality where you can't

[jaunty citrus]

And to DW's point if your production lines do not have the correct ratio of different quality levels, then you can remove some of the quality modules from the lines that have a disproportionately high quality level, to bring it in line with the other inputs

fwiw I think doing full quality production lines is a really bad idea, I'd just produce more of the items you want a higher quality of (modules, buildings, etc.) and recycle loop them to the desired quality level.

[vernal bay]

The point of having a Q2 only supply would be so you woudn't have to sort trough and proccess all the higher levels, if you make iron with quality moduel you need sort trough and proccess Q1, Q2, Q3, Q4 and Q5 seperatly which means anything that needs iron that you want in Q2 or higher now needs 5 assemblers. Randomly using Q3 machines in your mostly Q2 won't really help anything because ratios.

If you could just get Q2 you could get rid of a bunch of that complexity, which would make using cheaper Q2 much more worthwhile.

[jaunty citrus]

Why on earth are you "randomly" using Q3 machines???

just keep them for a later build

You're adding complexity and throwing away free value by wanting to process out the Q3 machines

[vernal bay]

Randomly using Q3 machines would be to use like a >= Q2 setting for placing your builds assuming that's possible.
If you store all the stuff for later builds you could in theory run out of storage and jam your builds even if it takes a long time. I personally like my builds to never jam even if it would take a long time for the storage to run full.

[jaunty citrus]

I would never have assumed that >= Q2 is a possible building option

Any production line that isn't recycle looping or has really high level and quality quality modules will produce less than 10% of the next level up than the lower one

so every 100 assemblers you make, using 4 quality 3 modules, you'd only expect 1 of them to be a Q3 assembler

You're not going to run into storage issues

and if you do, you can just up cycle them to Q4

and then to Q5, and at that point why are you even bothering with using Q1 or Q2 buildings? just use the Q3 or Q4 buildings

[daring siren]

With high mining prod, you don't care too much about recycling losing a lot of items

[distant eagle]

Only if it's a 2 step crafting chain that starts with all common
And only with 4 Q1T3 and you haven't unlocked Q4

[distant eagle]

https://factorio.com/blog/post/fff-375
"There are many internal changes we had to do...In some places, you just select quality, in others we had to allow the selection of a 'quality condition' ...This is a brief list of the places that come to my mind... assembler recipe selection"
Example picture is a requester chest, <rare, it's pretty easy to assume >=Q2 exists

[jaunty citrus]

Not for building things

We're specifically talking about building things >= some quality level... somehow

[half raptor]

Honestly I'm not sure what his goal is for quality, but I would advise anyone who's confused about quality to play with it but otherwise set up a quality loop at the last step and just be willing to sacrifice a lot of normal ingredients. Keep in mind that with high prod bonus, the "expense" is offset by the sheer quantity you're making

Great example of a final step quality loop, which will work for any machine production

[lyric yacht]

I don't think this design is that good tbh

[half raptor]

This loop will be more effective with higher quality modules, but that's true of anything

[half raptor]

It's a proof of concept design, not a fully efficient one

[distant eagle]

Oh. I see

[lyric yacht]

This example has maxed out modules

[half raptor]

Yes it does, but it'll work with rare modules too, just not as well. This path is more for people who don't want to mess with complicated quality builds

I'm personally going for a train grid that has production blocks for every quality combination of intermediates, even if the common and uncommon will run more often while others are mostly idle. Mostly for the meme and also just to see how well it works

bot based systems with circuit controlling recipe setting will probably be the most efficient

[half raptor]

One "bot mall" design I'm doing is that at my intermediate crafting blocks, the train loading chests will be with filter set to the item and quality intended for it, so all intermediates waiting to be loaded into a train is available for bots to grab.

At my bot mall, using Assemblers as an example, I'll have all 5 qualities of A1 set with requester chests asking for their quality ingredient. If it comes out the same quality, that A1 goes into a with filter and upgrades will go into a so that bots will sort them. Wire it up so it'll only load a couple stacks before turning off, then do the same thing for A2 and A3. I would not recycle anything until the A3 step. This way if I have a lot of materials on hand, it'll have 3 craft steps to upgrade naturally before any excesses go to recycle at a nearby recycler that goes to another active provider to sort it back out again for further crafting. The belts would just disable going to the recycler once I have my target amount of

By not recycling the A1 or A2, I'm only losing material at the A3 step if it still hasn't upgraded after 3 steps. Also since any upgraded intermediates are available from the rest of the base it's likely that every so often I'll be able to craft it from all legendary ingredients from the start

[distant eagle]

Requires lots of steel recycling, to match the ratio of A1 for A2 crafting. Though, youd have a foundry if you're doing legendary, so it's fine

Haven't done the math for recycling steel vs only using 2 modules for crafting gears, green circuits, and red circuits, and recycling the additional low quality A3

[daring siren]

I only eat A5

[half raptor]

My ratios across the rest of the plant are modular and they'll need to balance for every other request so honestly it's just a question for me of monitoring my demands

I'll be editing my interrupts schedule to either recycle excesses or sit on buffers. I need to play the game and test before I definitively decide. I'm designing based solely on speculation and guesses. There might be an awesome feature that blows my idea out of the water

[vernal bay]

I wonde how long it will take until calculators are updated to support quality

[frank warren]

It's kinda vague, it's not clear how they even would support it.

[daring siren]

Probabilistically

Putting into a calculation should change the recipe output to be fractional

Like any recipe with by-products

But it would be weird to click "how to make ?"... What options will it give you?

There are so many configurations of modules that can give that at different rates.

It should probably choose the same as input, and then if you change the modules and/or input quality, it could calculate it with that configuration.

[jaunty citrus]

"how to make " is a nonsensical statement to give to a calculator

probable answer it would give is 2

which wouldn't really be useful to the user

[daring siren]

Very useful, if you then let it be configured like I suggested

[jaunty citrus]

You'd pretty much have to do it through a matrix solve:

• craft gear from iron plate, using X modules
• recycle - gears, using X modules
• craft - from plates, each using X modules

[daring siren]

That's not the question. Of course you'll need matrix for it

[jaunty citrus]

That's simplified though, in reality there are 9 different operations happening in a legendary from normal recycle loop

[daring siren]

P.S. I'm talking Factory Planner / Helmod rather than Kirk / FactorioLab

[jaunty citrus]

Yeah FP would work fine for this, I just don't see the need for special consideration for quality

It should work fine with it's existing functionality + quality module support

[daring siren]

I want to be able to calculate like we did with gears in FP on JQ

Quality module support is what I'm asking lol

That's the post configuration I meant

[jaunty citrus]

Yeah so long as we get that it'll be perfect imo

[daring siren]

And an additional method to choose the input quality for the specific recipe so you could see e.g. how many you get from an input recipe

[jaunty citrus]

I would consider a to be a different recipe to a

They need different ingredients after all

So a quality selection similar to the existing quality selection interface would make sense

[daring siren]

Agreed

[half raptor]

Quality calculators are going to be interesting especially if you enable quality looping or if you designate no recycling until final step.

Also most calculations are going to assume you're using legendary modules for the math when in reality a lot of us will start with normal modules and upgrade planner them up gradually as we get better modules

[frank warren]

I'm under the influence that this is literally true now, quality select is almost just a UI thing.

[half raptor]

There was some exploit found using the mixed quality recipe do now they hard locked the recipes to require all the same quality for crafting

Not sure I understand the full exploit

[jaunty citrus]

charge the productivity bar crafting normal items with the crafting, then when the prod is about the fill up, put in legendary ingredients to guarantee you get legendary items for the prod output

basically you can get 100% prod on legendary items without needing 100% prod in the assembler

[half raptor]

Creative exploit. Too bad whoever figured that out didn't keep it under their hat

[daring siren]

Yeah crafting is no more

[wheat sluice]

pretty sure devs had it figured out way before us, IIRC one even made a pretty hefty base exploiting it.

[frank warren]

There is still a way to make it work. My new proposal is that recipe of Q(N) crafting should be able to use Q(N) or better items. It will always act as the Q(N) recipe though.

The thing is, quality any was not removed because of exploit, but because it's also a massive footgun. This does not aleviate that one bit.

[iron root]

The issue with that being default behavior is you can’t sushi your quality which is much more common than normal sushi

[frank warren]

That behavior could be a checkbox, that's an implementation issue.

You still could if you used a filtered inserter as well.

[half raptor]

That behavior is what was intended, but the exploit made it so they want to just set recipe and that's it

[iron root]

On second thought about not being able to easily weave excess quality stuff back into science
I guess it doesn’t matter much cause that excess is going to be such a small volume that voiding it or feeding it into science doesn’t make a difference

[half raptor]

I can't wait to see some interesting quality sushi behavior. I imagine if the whole belt reader reads a certain amount of on the belt, it'll add to the list of recipes the selector can set

[iron root]

Combining quality for circuit set recipes for stuff like that is…
An idea

I also wonder what the (fully vanilla) minimum factory size to launch a rocket will be now with circuit recipes

[wheat sluice]

Each quality being by an order of magnitude rarer that the previous means that very rarely you will have a deficit of a lower quality while also an excess of a higher one, and if you do ever run into that issue, you can just void the excess or recycle it into other stuff. Adding a checkbox for Q(N) or better would just convolute the system further.

[half raptor]

Just a mix of all recipes sorted by how much demand there is and how much ingredient there is, but the problem becomes if you accidentally put too many. I imagine you'd need a few sushi unloaders. Imagine if you set the limit for on the sushi network to be like 1k and the machine making gets lucky and adds a few more. You might need an inserter feeding a chest that picks up if it's over the limit and puts it back to the sushi belt once the buffer goes down

It would be a way to keep the belt from overfilling. Quality sushi belts are gonna be crazy tho

One reason the idea might work better via bots and logistics chests

[iron root]

The imbalances I’m talking about aren’t quality levels, but material types. With quality level imbalanced you can just recycle to kick the can down the road to
But I plan on mining quality ore for my mall, and shipping the normal stuff to science.
And science won’t exactly use copper/iron/etc in the same ratio as my mall

And once you’ve kicked down to or whatever the highest level is, there’s not much left to do but void if it’s still an issue

[grand igloo]

which I am still very sad about

[daring siren]

It was not a good feature

[jaunty citrus]

I have to agree with the foot-gun sentiment

[rotund egret]

it was a great feature for setting up dedicated quality lines

since you had a lot less clicking to do

[daring siren]

Param BPs

[lyric yacht]

I completely forgot about those

How useful are they actually

[iron root]

For train stations they’ll be incredible

For anything where it’s “just punch some stuff in a combinator and go” as well

[daring siren]

I use them all the time for a simple where the is filtered and the activates when logi amount is less than N stacks.

[lyric yacht]

wouldn't that be better with passive provider or buffer chests

[iron root]

Filtered storage is nice in that you don’t need to worry about checking the “request from buffers” box, and otherwise accomplishes their functions if you set it up that way immediately

[daring siren]

Also puts the items back rather than needing an extra for stuff. Doesn't really matter for , but helps with organization for - which take items from the closest rather than by priority.

[iron root]

take items back pretty well as is, but then you need everything to pull from buffers

[daring siren]

is earlier and don't need checkmark on

[half raptor]

I'm going to have a generic train request stations using parameterized blueprints. It'll just ask "what item/quality" "max input buffer" and the filters and everything else will just auto set

[iron root]

The only thing I don’t like about is that it’s horrendously scuffed if you don’t build it immediately, cause chests with matching items are prioritized over filtered chests

[daring siren]

Not a big deal either, it would just be like using .
That said, if you put your general somewhere reasonable, you could purge them into the other chests and then rebuild.

[half raptor]

Param bps will make it easy to set filters before they're even built. Gonna make design so much faster

[daring siren]

I use it so much that BP was on my quickbar

It could be more automated by connected a wire to an assembler, but there are so many different things to connect to possibly that I prefer to just select the item in the param BP

assemblers, belts, EMPs, etc...

[half raptor]

When a content creator talks about SA in the present tense... makes me so jealous. Do you live close to NC and are you willing to let me come over to hang out?

[jaunty citrus]

You could create a BP for everything you'd want to connect a wire to and put them in a book on your quickbar. Would shift the work to scrolling through the book to find the thing you want

[daring siren]

Yes, but it's just as easy to select the recipe

[jaunty citrus]

Yeah I would never make such a book, would be too restrictive on where you place the chest relative to the machine

That being said it would be quite nice for an assembler bot mall

[frank warren]

Most couldn't show you anything if they wanted :(

[wheat sluice]

not that much of a problem in 2.0
https://cdn.factorio.com/assets/blog-sync/fff-402-cut-paste-wires.mp4

[daring siren]

There's a mod for that https://mods.factorio.com/mod/reconnect-cut-wires

[neat shard]

Looking back at the quality FFF, the article claims that with a simple quality loop between assembler 3s and recyclers, with legendary qual 3s in everything, legendary items are 56x as expensive as normal items on average

My foggy sleep deprived brain wonders where that 56x figure comes from...

[sage vapor]

✨ math ✨

[distant eagle]

a year ago, no one could find a 56x

[neat shard]

Well, here's some math of my own, that takes into account the opportunity cost of not using prod mods instead...

Disregarding the chance of multiple quality bumps in one craft for the sake of simplicity, adding a prod mod 1 to your machine gives you 104% normal-quality products per batch of ingredients, while adding a qual mod 1 to your machine gives you 99% normal-quality products and 1% uncommon-quality products per batch of ingredients.

This means the choice is between +5% normal-quality products or +1% uncommon-quality products, implying that the uncommon product is 5x as expensive as the normal product.

Since qual mod 2s and 3s are 1.5x and 2.5x as effective respectively, just like prod mod 2s and 3s, this opportunity cost ratio is unchanged when you upgrade your modules.

So by this logic, for prod-compatible items, legendary quality versions are 5^4 = 625x as expensive as common versions.

[distant eagle]

Why are you doing 5^4?

[obsidian crescent]

This is the part where your math goes wrong. The question you're trying to ask is this:

If I have a bag of resources of X size, how many high quality parts Y can I get out of it?

[neat shard]

Why are you doing 5^4?

Each quality tier is a 5x cost jump, and there are 4 quality tiers beyond normal, so you apply the 5x jump four times

This is the part where your math goes wrong. The question you're trying to ask is this:

If I have a bag of resources of X size, how many high quality parts Y can I get out of it?

Assume the normal-quality outputs from the quality moduled assembler just get fed into science, and so are not wasted.

[obsidian crescent]

The thing about quality is that there are many ways of going about it, so their cost depends on the way you're doing it.

Feeding all of the non-quality stuff to science is one way to go about getting quality stuff, but the downside is that you'll always be limited in how much quality stuff you get by the amount of science you're making.

The advantage of a recycling-based setup is that you have both a 25% chance to get your inputs back to have another go and you get another set of quality modules to do that.

[iron root]

the funny thing about that 56x number
is not even the devs could reproduce it in the subsequent conversations

and i can see where youre going with that math, which is about the opportunity cost (lost prod)
that "56x" was rooted in the outright losses involved in recycle looping though.
And yes, prodding where you can and recycling with quality mods to quality boost and therefore exchanging the opportunity cost for the (possibly lesser) straight cost is... a thing for some configurations

[distant eagle]

If 1% are green or better, then 0.001% are legendary (if you've unlocked it)
5/0.001=5000x cost.
But this is for one step, you also produced green-purple which you can recycle and recraft, or send to quality specific crafting and continue rasing quality (if your goal item is more than one crafting step).
If they'd bring back mixed quality crafting, you'd only lose the opportunity cost 4% and the legendary 0.001%. 4.001/0.001=4001x

[fleet wyvern]

it's actually almost 31% more space because of the squaring. ( I know, this is a super important 6 months later response)

[kindred crater]

am i crazy or is it really strange that they decided to make fluids qualityless

i mean ig when they were deciding stuff about quality waifu was still clearly in the maybe stage

but it just seems really weird

[red kelp]

i dont think theres a nice way to have a quality fluid

[kindred crater]

especially with all the new fluids

[kindred crater]

why not

[red kelp]

for the same reasons people have touched on in #space-age pretty much

[burnt reef]

fluid mixing is very unsupported by the game

quality, by necessity, gives mixed outputs

[kindred crater]

does it

i thought each individual craft was a specific quality

[burnt reef]

the inputs are for a specific quality

the output is randomized

(if you're using quality modules)

[kindred crater]

yes i mean quality modules, output is randomized from craft to craft

[obsidian crescent]

You set the quality of the inputs and thus the base quality of the craft. But if the machine has quality modules, then it will randomly output qualities higher than the base quality selected

[burnt reef]

you could, in principal, have five different outputs from a machine, one for each quality

I do not believe this is a good idea

[kindred crater]

you can get 2 or 2 but not 1 and 1

[burnt reef]

right

[kindred crater]

so how would fluids break it

[chrome mauve]

it seems that someone didnt play Factorio before

[burnt reef]

but you can get 2 from one craft, and 2 from the next one

[kindred crater]

it might back up if the pipe has different quality fluid in it but i expect waifu makes fully emptying pipes quick and easy

[obsidian crescent]

Because you can't have 5 separate fluid outputs. Factorio does not really support multiple fluids flowing in the same pipe at the same time.

[burnt reef]

the machine's output needs to be able to handle both

[kindred crater]

you might need quality pumps to maintain high throughput

[kindred crater]

depends what you mean by that

[chrome mauve]

machine need to output up to 5 different fluids through same fluid boc to one pipe

[burnt reef]

... hmm. okay I see what you're thinking here. that could work. I still think it doesn't, but it could.

[kindred crater]

my idea has "concessions" (it's arguably a bit harder than quality items to understand) but qualityless fluids have very real concessions too

[chrome mauve]

which Factorio doesnt support rn

[kindred crater]

i don't like how it makes for instance so cheap

[obsidian crescent]

There are plenty of hard quality things to make. And remember, unless you have every intermediate of the same quality, you can't make stuff.

Quality LDS may be able to make quality spidertrons, but you also need quality fish. Which are... much harder to make.

[kindred crater]

i assume that could rests on the specific numbers involved in waifu?

which is fair enough

[burnt reef]

no, I mean, I think the fluid system connection rules still excludes jt

[kindred crater]

i'm like 99% sure pumps can have different fluids on both side

[burnt reef]

yeah but what about the machines themselves

[kindred crater]

wdym

[burnt reef]

the chemical plant, oil refinery, etc.

[kindred crater]

yes i know you meant machine

sorry that probably sounded rude

i just dk what you mean

[burnt reef]

can you hook up a water pipe to an oil output

if not, you probably can't hook up a rare oil pipe to a common oil output either

[kindred crater]

it'd be a hooked up to a output not a output

[burnt reef]

right, but a output is effectively a output most of the time

[kindred crater]

i think it has the potential to have very poor throughput

but i don't think the engine would prevent it outright

the machine would just get stuck

until the pipe emptied

which waifu makes both quick and easy

the biggest problem i forsee is offshore pumps would need to be rebalanced probably to consume power i imagine

or maybe not

hmm

i thiiink it would be okay if offshores didn't get module slots?

[burnt reef]

hm. casual accessibility? needing to give each qual modded chem plant five filtered pump outputs is a lot of work for players who aren't in deep

also needing to build five fluid networks sounds painful

maybe Pyanodon's will have it though

[kindred crater]

that's the thing i suspect modding this would be either super difficult or impossible

[burnt reef]

sadly so

[kindred crater]

admittedly i have 0 modding experience

it just seems tragic that vanilla has decided this is likely not allowed to happen

it'd probably break too many factories to have any chance of being in 2.1

i really think quality fluids are in vanilla in 2.0 or never ever ever, even in mods

it just seems silly with all the new spage fluids

[obsidian crescent]

I think we've seen... 5 new fluids.

• Lava
• molten iron/copper
• holmium solution
• electrolyte

[iron root]

this will severely kill throughput (and reliability) given the new fluid systems

[kindred crater]

and we've seen like 5 new resources if you count every plant as one

[kindred crater]

i think the new fluid systems would make it much better

[iron root]

because networks empty out much slower the closer they are to empty, and you need to fully empty the network to pump in a new fluid

[kindred crater]

ultimately we don't know until we have hands on it but i really doubt it's that drastic

[chrome mauve]

it wouldnt, it was changed cause people complain about pumps everywhere and you want to bring pumps everywhere back for little to no gain

[kindred crater]

throughput might be low but not intolerably low

both ideas have concessions and i really think quality fluids has less bad concessions than qualityless fluids

[chrome mauve]

just oil refinery with just 3 fluids is already bad when you think about quality fluids

[distant eagle]

oxidiser and fuel

[iron root]

every fluid network switch would cause... more or less cause a certain loss in time in the system (its at least capped by the 100%->0% time case)
And the more machines you have, the more theyre going to be sitting backed up

or, or, OR you have one machine that isn't quality moduled, and is feeding the network with fluid faster than it can be fully emptied, which makes every other machine with quality modules on the network jam

[obsidian crescent]

And 7 new resources (ice and carbon, though the latter can be manufactured).

[kindred crater]

i admit now that you point it out this specific sort of situation would actually be a silly knowledge check but i imagine players can connect the dots

[iron root]

situations like that lead to infinite numbers of questions of "why isnt this working"

we already have issues with 1.1 where people are trying to jam too much fluid through a pipe

[kindred crater]

would also be rare for people to only some of their machines

[iron root]

maybe, maybe not.
But say you miss a module insertion sometime in early game where you're doing stuff mostly by hand.
And even in the case where it works properly, the machines will slowly turn off one by one until theyre ALL off and the network can empty

and that process can take a long time and lead to a ton of lost productivity

not only that
its an unpredictable loss in productivity, or at least, its not immediately calculable with the resources in game

[kindred crater]

ig it seemed really easy for singular machines but does scale somewhat exponentially in complexity

[iron root]

yeaaaah
i spent like, 4-5 years dealing with the bullshit that happens when you get systems interacting

its fun to explore that stuff

[burnt reef]

even when the basic parts are simple, they can combine and interact in very complicated ways

[kindred crater]

ig i hope quality fluids are allowed in engine but disabled in vanilla if that makes any sense

[burnt reef]

and if the basic parts are complicated...

[iron root]

im pretty sure they arent even allowed in engine, like, the fluid prototype doesnt even support quality

(there is, of course, the janky fluid quality route. Make that mod at your own peril.)

[kindred crater]

yeah and we have no way to know how hard it would be to add that but we also don't know how hard quality was to add

i am Not A Programmer but i don't see how it would be so much more difficult than items

[iron root]

my experience with programming indicates that attempting to predict the difficulty of anything is a futile endeavor

[kindred crater]

true

[iron root]

but about 90% of the time its harder than expected

[kindred crater]

guess nullius will just have to be janky fluid quality mod

probably not fun to code

[iron root]

we already have the conceptual and high level framework for janky quality, because it exists as a mod now

[kindred crater]

true

[iron root]

for fluids, it'd be extending those patterns to, well, fluid objects
and iirc in janky quality all of the qualitied items are just wholesale different items, for fluids it'd likely be the same

[kindred crater]

what if you layer janky quality on spage quality

so everything has 2 different qualities

[iron root]

i mean
maybe
but janky quality will probably just have SA flagged as an incompatibility

[burnt reef]

https://xkcd.com/1425/

xkcd: Tasks

In the 60s, Marvin Minsky assigned a couple of undergrads to spend the summer programming a computer to use a camera to identify objects in a scene. He figured they'd have the problem solved by the end of the summer. Half a century later, we're still working on it.

[chrome mauve]

janky items are different from janky fluids, you dont have problem of putting differnet items into one fluid box

[kindred crater]

the

[iron root]

we were uhh
exploring the implications of attempting this with fluids 2.0

[kindred crater]

conclusion was it's too difficult for scrubs (/joke)

[iron root]

its definitely in a "mod only" territory given how jank itd wind up

[kindred crater]

i really don't think it'd be jank, just hard to build large scale and optimally for

each machine would need 5 pumps

would be too hard for vanilla spage though

[frank warren]

It's easy if each building has 5 fluid outputs for each output.

[kindred crater]

yeah i considered suggesting that they could add a giga- or such but i thought it would seem like a copout

[jaunty citrus]

And that would only work for recipes with one fluid output

[burnt reef]

is there any recipe in vanilla that gives multiple distinct fluid outputs?

[jaunty citrus]

No, but if you're making quality fluids a thing it needs to apply to all fluids, including oil processing

[kindred crater]

yeah

[jaunty citrus]

I suppose you could make quality modules incompatible with multi-fluid output recipes like with catalytic recipes

[kindred crater]

no

i have

an idea

[burnt reef]

so a giga-, a tera-, and a kilo-?

[kindred crater]

if you made this huge building with tons of fluid outs (5 suggested but could be more or less) and you made it so that when it finishes a craft it picks any output based on this:

1. if output connected to pipe with same fluid at same quality and with enough room in the pipe, output there
2. if not above rule, output to random empty pipe

[jaunty citrus]

There's also the issue that mods can add more quality levels

[burnt reef]

you'd need to make sure the output pipe system never ran dry

or else it'd get scrambled

[kindred crater]

or maybe it could support setting a filter on each output

or both

so if you set no filter you could use it like that

i understand this might just be too confusing

like for players

they obviously want to keep vanilla simple enough

but i think it would make it a lot easier to have high quality fluid throughput

[jaunty citrus]

You just can't marry quality and fluids and not have it be super jank

[kindred crater]

no, you can't marry quality and fluids and not have it be too difficult for base spage

[jaunty citrus]

It requires mixed fluid networks which is inherently janky

[kindred crater]

i hazard to believe it's not inherently janky

[kindred crater]

also you could do this with a really simple circuit

and this version, assuming your number of outputs is at least 5x your number of fluid products, wouldn't even need pumps i think?

[burnt reef]

you'd need pumps for the circuit condition

[kindred crater]

OH yeah

you right

still probably too hard for vanilla

just gotta hope mods can do it the non janky style

[jaunty citrus]

Mixed networks are janky enough that the devs have attempted to prevent players from accidentally doing it for years

But they're so janky that trying to prevent it doesn't always work

[kindred crater]

i think we have like 3 different definitions of jank here

fuckin' english

[burnt reef]

I think this is more a "limitations of language" thing than English specifically

[kindred crater]

i hope a janky fluid quality mod like the current janky quality mod will be possible but the more i think about it the less possible it seems

because for every recipe with fluid involved it'd have to like completely override vanilla quality behavior

or whatever

sigh if only i knew literally anything about modding

[daring siren]

There won't be a janky fluid quality.

Not with vanilla buildings

As you need 5x output pipes

[kindred crater]

see previous conversation, tl;dr with pumps you don't

[daring siren]

The game doesn't support outputting different fluids on the same output

[kindred crater]

oh

fuck

well this is the end of the line

i will never get to quality module advanced oil processing

[daring siren]

Unless you contort temperature into it somehow

[obsidian crescent]

You could always turn the liquid into a solid. Or multiple solids rather.

[kindred crater]

HMMMM

god i hope the modders figure it out

i guess i could use a combinator random number generator to determine my fluid's "quality" and playact at it

this is ADVANCED levels of cope you are witnessing right here

[fallen urchin]

They can use temperature, though it's hard to make temperature interact with modules

In Py, the concentration of UF6 is represented by the temperature

[chrome mauve]

there is a big difference between "simulation" of concentration with same fluid to make that work with something that its different fluid.

only way that would work is just same fluid with different temp and if that would be linked to quality of it there would be a lot of additional recipes - either for making that quality fluid in the first place and when using it as quality ingredient

[jolly swift]

you could also just add pipes that can carry 5 different qualities of the same fluid

and add pumps to filter them out into tanks

[jaunty citrus]

you can't connect 15 pipes to a refinery though

Oh, you mean using special pipes

[half raptor]

Or if a pipe system could hold multiple fluids, like if the higher quality floated on top

Or if making higher quality fluids take 5x as long to make

[jaunty citrus]

Lot of effort just for quality fluids, putting 5 fluid networks in each pipe

[dense kraken]

technically, you can

[daring siren]

That's not the question

[dense kraken]

just put all fluidboxes at the same place :p

[daring siren]

The problem is having 15 output fluidboxes

[dense kraken]

ah that's what you meant by vanilla buildings

thought you meant the engine, not the entity prototype

[half raptor]

Heresy, don't show that to the devs, might give them crazy ideas

[dense kraken]

don't worry, Klonan got mad at me already

[half raptor]

Even if you made it work, they like visually readable logistics systems and it's much easier to mess with one fluid per pipeline

[chrome mauve]

Ofc you "can" mix fluids in Factorio. That doesnt mean it is good to play with

[frozen rose]

sushi fluids 🤔

[jolly swift]

so for recipes that have fluids and items are the fluids just not counted for quality, i.e if I use and I’ll get batteries and the sulfuric acid doesn’t matter?

[obsidian crescent]

A more interesting example is the Foundry's LDS recipe, which was changed to take molten copper and iron instead of plates. As such, all you need for quality LDS is quality plastic.

[rotund egret]

correct, making quality LDS from qualitied coal was how I made a lot of my stuff for the second iteration of the space platform

[half raptor]

I wonder what rockets will yield. I suspect a quality rocket pad just constructs rockets faster

I hope increased cargo

[dense kraken]

+400% rocket per rocket

[rotund egret]

silo go brr

[obsidian crescent]

Even if they did, it would take a lot of item productivity researches before the increased cargo capacity of quality rockets would be worth the extra cost of quality rocket parts. Even if Q2 rockets could carry 2x the stuff of Q1 rockets, you still need to get to the point in your production chain where Q2 rocket parts are cheap enough to actually make.

Also, it should be noted that the UIs we've seen of rocket silos don't have a quality filter on them. Note that the silo in question uses Q2 prod module 2s, so it was taken from an SA run where quality has been researched. With "Any quality" no longer being a thing, it seems impossible to load a silo's construction inputs with anything other than base quality stuff.

[half raptor]

Well if Fulgora gives me a I can just recycle to get uncommon intermediates

Unless of course there's something worth using them on instead

[rotund egret]

sure, but you lose 75% for each recycling step

[half raptor]

Depends on if I have a use for quality LDS. If there's a use for it then sure I'll use it, but when I recycle scrap and I get a quality LDS I can't use, I'd rather get the quality steel copper and plastic out of it. We will see.

Odds are plastic is used in the lightning rods as an insulator or for other things so there might be use for it elsewhere. I'm only working on my 1.1 knowledge

[obsidian crescent]

Fortunately, all that means is that quality copper and steel plates costs 4x as much as they otherwise would. And technically... they may not even cost that much, thanks both to the Foundry's 50% prod bonus and to whatever recipe differences there are between casting LDS and casting the equivalent amount of copper plates.

[half raptor]

I'm talking about a hypothetical LDS coming out of scrap. If I don't have a need for quality LDS I wouldn't make it normally

Since it comes out of scrap, no initial investment went into it save the scrap and even then it was just a quality chance that caused the issue

So it's not a question of whether it's a waste or not, since we cannot mix and match qualities for production, if there is no tangible use for quality LDS, I would just use that random scrap as a way to get some quality steel and copper

Though if the EMP has LDS as a component, then the argument becomes moot since I want to make some quality EMPs. I'm basing my reasoning on 1.1 recipes

[dense kraken]

that's what so interesting about fulgora right there

it reverse the recipe tree entirely

expensive items become dispensable, basic resources become expensive

[half raptor]

Yeah, getting green circuits means sacrificing blue and red because the green doesn't come out of the scrap

[jaunty citrus]

You still get 5 green circuits from each processing unit, so you should be able to get plenty

Anything that you can't get from scrap recycling that needs something like copper plates though? Not so great

[half raptor]

Well cable comes out of scrap so that's not bad

[jaunty citrus]

Yeah but it's 1 copper plate from 8 cables

You could be off recycling LDS for 5 copper plate each

[half raptor]

I'm just going to be recycling excesses

Odd to think we'll have excess

[jaunty citrus]

I mean, everyone will be recycling excess

[dense kraken]

what do you need copper plates for anyway?

you can get circuits, LDS, copper cable and batteries from scrap already

ah of course, supraconductors

[solar osprey]

Quality nuclear components.

This is about quality copper plates.

[dense kraken]

you don't have to do that on fulgora tho

[solar osprey]

Don’t have to, but fulgora is the place where you’ll be building up stockpiles of quality items for free.

[dense kraken]

idk about that

you'll have not many quality steps

thus low quality output

[jaunty citrus]

Why would that be the case?

[solar osprey]

You get the quality items from quality modules in your recyclers.

[dense kraken]

you can do that on all planets

[solar osprey]

But on fulgora, everything is coming from recyclers.

[dense kraken]

which is bad

what's interesting about fulgora as a planet is its natural "mining productivity"

but the more crafting steps you have, the higher the quality of your output

[solar osprey]

How is it “bad”? Even if you just bring Q1 quality 1 modules to fulgora, 4% of everything you recycle will be uncommon or higher… just for free.

[dense kraken]

putting quality modules in your assembling machines also does that...

assuming you already put some in your miners in both cases

[solar osprey]

But you can use productivity there instead. Which will pay out more.

[dense kraken]

so it's better

if it "pays out more"

[solar osprey]

I can’t believe I’m having the “fulgora just creates quality items for free” discussion… again.

[dense kraken]

sorry I wasn't around during the first one lmao

also sorry if your point doesn't seem to make sense

[solar osprey]

The scrap recyclers on fulgora will create quality items if you put quality module in them, for free, as a natural byproduct of resource production.

If you craft using normal quality ingredients for most stuff, then Fulgora will gradually build up a stockpile of quality items.

[dense kraken]

yeah, and mining and assembling items using quality modules will also "create quality items for free"

[obsidian crescent]

You cannot put prod modules in recyclers.

[solar osprey]

I never said that.

[dense kraken]

they were replying about putting quality in assembling machines

btw prod modules aren't that good in the first steps of production

[solar osprey]

On other planets like Nauvis, producing quality items from ore requires trade offs in using quality modules in assembling machines or furnaces over efficiency or productivity modules and dealing with the increased logistical burdens of dealing with quality ore.

[dense kraken]

plus with SA there are buildings that have already a prod bonus, which lowers the effect of adding prod modules

[solar osprey]

And once you go to Vulcanus that gets worse as then the cost of starting quality early gets worse.

But on fulgora, quality items will just fall out of your resource processing, for free.

[dense kraken]

vulcanus is good for quality items

because all fluids are like legendary when mixed with other qualities

[solar osprey]

No matter how “cheap” quality is elsewhere, on Fulgora it can be “free”.

[dense kraken]

like everywhere else yeah

[obsidian crescent]

There is more to the opportunity cost of using quality modules in your scrap recyclers. Specifically, you need to sift through a lot of scrap. This means a lot of recyclers. Speed modules can compress this, but you can't use speed modules with quality modules.

So it's not "free"; it's very costly in space.

[dense kraken]

you want to be pedantic? it's not free on fulgora as well since you need to power at least your miners, recyclers and inserters

[solar osprey]

Power is free on fulgora. It’s lightning. If you want to be pedantic.

[dense kraken]

with solar it's free everywhere

[obsidian crescent]

The scrap recycling recipe is 5/s, and with quality modules, it'll be less than that. Sifting through 2000 scrap per second is a lot. You'd need 400 recyclers to do that, even if quality modules didn't have a speed penalty.

[solar osprey]

More looking at a stockpile over time.

[dense kraken]

anyway, who cares what's good or bad, just play the way you like the most

[iron root]

so everything you can do to quality module things on other planets, you can also do on fulgora. Fulgora also has the mandatory recycling steps that can be qualitied too, which lets you get more quality out of the same raw material, sure.
I think the space constraints might actually be a counterbalance to that though, because you do need to recycle a ton of scrap.
And if you're quality modding your recyclers, you cant speed mod (or speed beacon) them, so you need a ton more space to do things.
How this pans out in game I have no idea, but it's there. It'll require you to build a much farther reaching base than you otherwise would, which is a similar issue that building quality ore on nauvis will have.

and while power costs no resources to harvest via lightning, the amount of power each island gets will be constrained because they'll be too far to network until foundation

[daring siren]

And space is at a premium due to islands

[iron root]

pretty much. Quality on fulgora feels less about resource costs (opportunity cost or raw cost), and more about fighting the space/power constraints

(and the logistics design )

[obsidian crescent]

There's also the unfortunate fact that the first step in Holmium processing is to liquefy it. Which strips out quality from the thing that you most want to have quality.

Also, this liquefication step may involve a second resource (stone?); if it does, that's a big problem as, without any-quality, you now need quality stone equivalent to your quality holmium.

[solar osprey]

You should be producing quality stone in the same ratio as quality holmium ore

[obsidian crescent]

True, but it's still an irritation you have to deal with.

[jolly swift]

wait that message was an hour ago 😭

[obsidian crescent]

One idea to mitigate the space issue with scrap recycling is to have high-throughput processing and low-throughput quality processing. That is, you have one place for scrap to go that goes into speed-moduled recyclers, and another place that goes into quality-bearing recyclers.

That way, you can at least take advantage of some quality production without having to go too far with it.

[jolly swift]

my bigger question regarding fulgora rn is that, if the player goes to vulcanus first, and sets up logistics between fulgora and vulcanus, they can sidestep the recycling mechanic entirely and exchange finished intermediates from fulgora with what are basically free raw resources on vulcanus, with the only limiting factor being rocket cost

and rockets are basically free on fulgora, so it would be trivially easy to start that way

[jaunty citrus]

You don't need to go to vulcanus to do that

You can just send stuff from Nauvis if that's what you want to do

And you're glossing over rocket cost quite quickly there

[obsidian crescent]

Remember: at 50 LDS per rocket launch, one rocket on Fulgora costs about 5000 scrap. Productivity can bring that down quite a bit, but that's still a lot of scrap to churn through.

[jolly swift]

That’s a good point

[jaunty citrus]

Plus churning through all that scrap is going to give you a ton of other stuff. If you're just shipping in all your basic materials what are you going to do with it all?

So really making a rocket on Fulgora costs ~5000 scrap, with the byproducts of 50 processing units, 150 advanced circuits, 100 steel plates, 300 concrete, 100 batteries, 250 ice, 250 stone, 50 holmium ore, etc.

[half raptor]

The other side of that math is how often do we need to launch a rocket? Odds are most rocket launches will be modules and EMPs

[jolly swift]

basically what i was originally suggesting was creating some equilibrium where both vulcanus and fulgora have the correct amount of raw resources and higher-tier resources and then move forward with production as normal. I think the reason that wouldn't work is actually because of the rocket cost from vulcanus. The rocket cost in exporting tons and tons of low-quality resources from vulcanus would be ridiculous

[half raptor]

I want to make each planet self reliant and exports finished goods that makes sense for the planet, so foundries, belts, and Vulcanus science from and then EMP, modules, inserters, and Fulgora science from if we are right about prod modules needing tungsten then those modules will come from there too

[jolly swift]

i want to make interplanetary spaghetti

when i stop tesla turret production on fulgora I will be disappointed if my vulcanus science and nauvis ammo production do not also break

[half raptor]

If you do, make sure you post the results on Reddit/YouTube. I love seeing successful spaghetti

[keen igloo]

I want to bring back lightning towers from Fulgora to Nauvis for the style

first mod i'm requesting is lightning on nauvis(no actual gameplay effect, just visual

[merry spindle]

Vulcanus seems to be better place. LDS from plastic and 2 fluids means that quality coal easily becomes quality iron and copper

[obsidian crescent]

Easily becomes quality steel, not iron. You can't recycle your way back to iron.

However, you can make quality iron while making quality coal by quality cycling grenades.

[solar osprey]

That also requires legendary productivity modules and 15 levels of productivity research, or more levels of productivity research and lower quality modules. If you want to do it for free.

[obsidian crescent]

That's not what Advena's talking about. They're talking about how the Foundry LDS recipe uses molten metals (which don't carry quality), so the quality of the LDS produced is only affected by the quality of the plastic used. So if you can get quality plastic there, you can get equal quality LDS.

[solar osprey]

I did say, if you want to do it for free.

Which the 300% prod point is the free point.

[daring siren]

It goes free well before 300%, because of 50% built in prod and prod modules

[obsidian crescent]

You can't really get "free" LDS from the Foundry recipe, because recycling doesn't reproduce the fluids. You have to use an assembler, so the earliest you can get "free" quality LDS is +200% from research.

[grand igloo]

that’s so fucking clever

if it works

quality iron is easy quality steel too

then you just need a good source of quality stone and boom

quality factory

[uncut plank]

Take the excess stone from your lava process and cycle walls or landfill. It's wasteful, but you should have a ton.

[grand igloo]

can you make rails with prod?

[uncut plank]

you can'g in 1.1. so probably not in 2.0

[solar osprey]

It won’t. As recycling the LDS will only get quality steel, not iron. Good, but not as good.

[grand igloo]

hmm

grenades?

[solar osprey]

Besides, at 300% prod, if you put legendary plastic in, you can get legendary LDS out, then recycle that to get legendary copper and steel and the same legendary plastic you put in, out, to make further legendary LDS.

[grand igloo]

what recipes for molten iron exist

[solar osprey]

Gears.

Iron plates.

Concrete

[grand igloo]

…is concrete recyclable?

[solar osprey]

Steel.

Yes.

[grand igloo]

hm…

well then you can get quality iron from quality stone

[solar osprey]

But you have to smelt the ore into plates.

[grand igloo]

And?

Smelting doesn’t downgrade ore to base quality

[solar osprey]

An extra step.

[grand igloo]

It’s an extra step but an easy step

[solar osprey]

True

[grand igloo]

well actually is there coal available on vulcanus

or other burnable fuels

[solar osprey]

Yes.

Coal liquefaction is the basis of oil processing on Vulcanus.

Also, if you’ve got high quality iron ore, you may as well smelt it in an electric furnace so you can use productivity modules on it.

[grand igloo]

oh yeah those exist too

i don’t think you can prod concrete?

but either way you’re just using it to pass quality from an item to an item form of a fluid

[half raptor]

this is true, you can only use productivity modules on items in the Intermediates tab

[distant eagle]

if you balance the quality ratio of your ingredients, only 2 items can have more than 3 quality steps

[grand igloo]

so it looks like coal and stone quality are going to be the big chase targets

ironically

even though you wouldn’t initially think so because they’re used to pass quality to your main materials more cheaply

[dense kraken]

what do you mean? can have 7 quality steps

oh wait you right, plastic is the limiting factor

4 steps

same goes with , this time it's the bricks, also limiting the total to 4 steps

is exactly like , plastic-limited to 4 steps

ironically is the best with 5 steps

[half raptor]

I plan to not think too hard about it and just go for quality on every step, setting up automatic delivery to recycle stations if the request buffers for the item are too full

[distant eagle]

im not counting quality ore as a step, but if i do, green science is 4 steps

ore->plate->belt/inserter->green science

-> -> three

[half raptor]

the good thing about trying to craft is that you can use any normal rails for base construction

[obsidian crescent]

Yeah, but you can already get quality iron from quality cycling grenades, which is a good way to get quality coal.

[grand igloo]

maybe but qual grenades can’t be prodded can they?

[jaunty citrus]

It's unlikely anything both recyclable and made from coal can be prodded, at least of the current item list

[grand igloo]

explosives?

[obsidian crescent]

No, but coal isn't exactly the most important resource (even on Vulcanus), and iron is relatively plentiful. Of the available options, it's probably the best.

[grand igloo]

they’re an intermediate

[obsidian crescent]

They're a chemical plant recipe, so they can't be recycled.

[grand igloo]

oh, right

[jaunty citrus]

There's nothing to say that explosives would recycle into their ingredients

[grand igloo]

hmmm

[jaunty citrus]

I'm 99% sure the only reason batteries are an exception to the rule is because of fulgora

[grand igloo]

poison capsules maybe

[obsidian crescent]

The other thing is that plastic has a productivity research, so with lots of that and/or prod modules and/or a super-chemical plant, you can make your relatively small amount of quality coal go much farther.

[jaunty citrus]

Super chemical plant is very likely 👍

Looking forward to confirming everything in 18 days

[grand igloo]

poison capsules seems like they’d be good for cycling coal

tho ig no different than grenades

[jaunty citrus]

I'd say they're worse than grenades because they're more expensive per coal

[grand igloo]

well yeah but no different in speed

[obsidian crescent]

The main issue with poison capsules is that you need more stuff. Grenades are ideal because you only need one other thing: iron. But for capsules, you need iron, copper, and steel. That's a lot.

[dense kraken]

ah nvm, I hoped kirkmcdonald calculator was putting recipes in the order I wanted, there are items that are flowing next to other recipes

[frozen rose]

when the final product uses multiple intermediates steps i.e blue ciruits - is it worth it converting to a higher later or should you start this back at the smelting phase?

[dense kraken]

the less recycling the better

so better craft until the end product to maximize the chances of getting better quality

and only then recycle loop if you need to (or just skim what you need and leave the rest for the main factory, or a mix of both, etc)

[frozen rose]

so make everything in each step with and then recycle the final product until you get desired got it

[dense kraken]

uh I mean, if the first step outputs a item the second step with be like that

but if the first step gives you a item you better use it for crafting

just one assembler will do, it doesn't take much space

that way you get the most high-quality items at the final step, which reduces the need for the inefficient recycling

[frozen rose]

but you need quality modules in the intermediary steps then right?

[dense kraken]

right

[frozen rose]

is it preferable to have a mix of prod/quality modules in intermediary step or just one or the other (until the final step ofc)

[dense kraken]

I've heard 2x + 2x is the best, I'm not sure myself

I'll use only quality modules to make quality items

wait no, the 2Q2P was for recycling

I think going full quality is still the way to get as many legendary item per ore as possible (without recycling)

[jaunty citrus]

2x +2x is for getting the most higher quality outputs from the input ingredients

A lower % of your output is of higher quality, but the overall output growth from productivity counteracts that

[dense kraken]

does it stack tho

like if you have a recipe chain, is the 2Q2P really the best for all of the recipes?

[wheat sluice]

with modules it's best to use 4Q, same with , with 3Q1P, same with , only with modules 2Q2P is the best

And I would guess that that also also holds for chaining

[dense kraken]

ok that's interesting

[wheat sluice]

*assuming no recipe/machine productivity

[frozen rose]

but productivity modules will make it slower right? So you can trade resources for time if you want

[wheat sluice]

quality makes it slower too

[frozen rose]

ahh

[wheat sluice]

speed modules drop quality

[dense kraken]

time = space

[frozen rose]

oh no

[dense kraken]

and space is big

so not a problem in factorio

[frozen rose]

*sweats in quality prod space ship*

[dense kraken]

well

[wheat sluice]

but the quality penalty isn't even that big on speed modules

[dense kraken]

maybe not everywhere in factorio

[wheat sluice]

-8% from all of those beacons

[dense kraken]

yeah it's 1% per module

[frozen rose]

and with the changed beacons that's a..... good thing I think because less is more?

[wheat sluice]

with the beacon rework, more is still more

just less more

[dense kraken]

it's less more per more

but more it still is

[wheat sluice]

but less is more than what less used to be

[dense kraken]

right

also with quality more is more than what more used to be

[wheat sluice]

and 8 is just about as more as it was

why say more, when picture say morer

[frozen rose]

+fff 409

[eternal verge]

Friday Facts #409 - Diminishing beacons

Hello, Today we're going to take a look at a feature some of us have dreamt of changing for years now - the beacon transmission. The main purpose of beacons is to allow massively increasing your production in the late game while being more than just a module or a faster machine. To make use of beacons you need to adjust your building layout for them. Beacons succeed in this role, but...

[jaunty citrus]

This was caused by module effect clamping in the background. It would now show 0% because the clamping was fixed.

[wheat sluice]

Did they go with the clamping on quality? I thought that they decided against it.

[jaunty citrus]

Negative quality on a machine doesn't do anything so it should clamp at 0%

[frozen rose]

so is it "quality" subtracted by 8%, linearly?

[wraith moss]

Quality bonuses are clamped at 0%, yes. You can't have a negative quality bonus to downgrade items, instead speed modules only weaken the effect of quality modules.
https://discordapp.com/channels/139677590393716737/603392474458882065/1233870774683566134

[wheat sluice]

the effect yes, but kovarex said that they are keeping the negative value in the tooltip for transparancy

[half raptor]

It's good to know what the negative value is so that it's obvious how much quality effect you'd need to overcome the negativity

[solar osprey]

Though would you want to use speed and quality modules together?

[half raptor]

You wouldn't, but seeing the negative value makes that obvious to a new player

New player would see "oh, I shouldn't mix quality and speed"

[solar osprey]

I don’t know about that.

I recently watched a new player put 2 efficiency 2s into their miners.

And I had to explain that 3 efficiency ones would have the same effect, but be cheaper.

[wheat sluice]

yeah, but leaving the negative value in the tooltip has a better chance of communicating the quality penalty to a new player

[jaunty citrus]

I mean it'll literally tell you on the module tooltip

[half raptor]

Don't overestimates players ability to find the information they need to play well. Sometimes games need to have the text flashing and people will still complain "the game didn't tell me that"

Keep in mind some people are going in blind without reading the FFFs and the discord. The blog post explicitly explained it, but someone who didn't read it will discover it in their playthrough. I guarantee there will be a reddit post the first day "why do my quality modules not give me all of the percent chance they should?"

[red kelp]

people hate rtfm

[half raptor]

What's that?

[obsidian crescent]

An admonishment to "Read the Fine Manual". Feel free to substitute other appropriate words for "Fine".

[half raptor]

||Fine||

[jaunty citrus]

I feel great concern from the people who will read the module effect of modules but then not, in this case almost literally, put 2 and 2 together to understand that -% and +% means no %

But I suppose if one did not read the module effect and just used them, for reasons I suppose, then they could miss that detail

[iron root]

putting 2 and -2 together

[iron root]

that said
given the numbers we see here, qualitied does appear to be capable of climbing out of the hole that speed modules put the machine in.
a speed bonus of 600% (with prods in the machine!) is pretty substantial, and -8% quality isnt irreparable even at that level.
I can honestly see some speed+quality cases existing where a quality setup has to be jammed in a small space. You'll want, well, quality materials to make it work so you have more upsides for your downsides, but I can see it.

[obsidian crescent]

I'd say your best bet for that would be high-quality, low tier speed modules. Q5 speed 1s inside of a single beacon can give a +75% speed boost each. We'd need the exact numbers for the quality penalty from the various different tiers of speed modules though.

[iron root]

in fact
we can even roughly work out what the -quality module is on speed 3s from this ss

[jaunty citrus]

https://media.discordapp.net/attachments/1215073493142474833/1234846558198038568/image.png?ex=67008ab9&is=66ff3939&hm=ba51248dec369c6b62a9547bb8046b5e3573c87b695d6df91a9f180e51bfcf38&=&format=webp&quality=lossless&width=435&height=592

[iron root]

and that gives us... something to narrow the range, because these machines both are at -8% but have different amounts of speed module effect from the beacons (rounding's a bitch)

[jaunty citrus]

For the speed multiplier

Not exactly a nice number

[iron root]

okay
the foundry ss plays nicely with my logic.
Take 1916, add 60% (removing prod mods) to get 1976
Divide 1976 by 10 (# of beacons) and 0.3162 (relative effect of beacons), get 625%
Divide by 2.5 (Known single beacon effect) to get 250, then by 2 (# of modules in a beacon) to get 125% speed boost per module
which we know is also 50% ( ) x 2.5 ( quality modifier)
so that all checks out
now to apply it to this damn emp because this was weird

[jaunty citrus]

0.505964%

Definitely clamping going on imo

[iron root]

carrying out the same math on the EMP gives 80% speed boost per module if assuming the beacons are normal, which is possible with

[jaunty citrus]

https://media.discordapp.net/attachments/1215073493142474833/1234845216066769027/image.png?ex=67008979&is=66ff37f9&hm=472f0fc24bc0adfe1e00632f60b3d51bf732ef28c49b1c6e93051b1c46c3080d&=&format=webp&quality=lossless&width=629&height=592

It's all legendary in this one as well

That's all you really need to know that clamping is going on with the -8% quality value

Both have -8% despite one having ~13% more 'speed module'

If it wasn't being clamped you'd expect the foundry to have -9%, or the EMP to have -7%

[iron root]

i get 15.81 "speed modules" for the foundry and 14.14 for the emp, assuming the beacons are legendary in both

[half raptor]

this game is math

[iron root]

if the speed module gave -0.5% quality, and it rounded down to the nearest %, it'd be -8% in both cases

[jaunty citrus]

No it would be -7% for the EMP

[iron root]

EMP is 0.3535 * 8 * 2.5 * 2, which spits out 14.14 modules

if you rounded to nearest, it'd be 7%

but there might be reason to explicitly round down here

[jaunty citrus]

This is a big stretch, ngl

[iron root]

it might be but unless that clamp is actually meaningful, in that you only need to climb out of a quality hole of -8% even if combined speed effect was greater than that, then it'd be giving the player incorrect information
and wube loathes that

[jaunty citrus]

This was a screenshot taken while in development, and we already have precedent for effects to be clamped to -80%, see efficiency. (keep in mind the actual quality module effect is 10x greater than what is displayed)

[iron root]

...
if this is quality getting affected by the global -80% clamps (and that IS possible because I remember the devs talking about how those rank up chances are multiplied by 10 or 100 to get "quality values" under the hood)...

funny though that the rounding down assumption does make things come together consistently as well

actually

no it doesnt

[jaunty citrus]

Yeah then the foundry would be -9%

Nah it wouldn't

Not mathing correctly right there

[iron root]

because the beacons in THIS ss likely aren't qualitied, given that my math to calculate speed mod effects only worked out cleanly to beacon and
and i got 15.81 "speed modules" for the foundry, which at -0.5% each would still get an 8% when rounded down

but if the beacons aren't qualitied, then they should be applying a significantly lesser quality penalty to the building
and they arent, which is very strong support for clamping

[half raptor]

this is why I'm so glad the mouse-over tooltip offers us a rate calc so we don't have to do it ourselves

[iron root]

well. That was fun.
Shame that the possible exploitation may not be anywhere near as good as initially thought.

if it's even possible to begin with

[jaunty citrus]

That EMP uses and

[iron root]

yeah, that combination spit out very clean numbers while everything else was off by... more than rounding errors would suggest

[jaunty citrus]

With an expected quality% value of -4.24 assuming is -0.5%

There just has to be clamping going on

Otherwise they all couldn't be -8% total quality effect

[iron root]

yeah, even rounding doesnt overcome differences from that one

and we do have good precedent for the -8(0)%

[jaunty citrus]

We do have a known minimum (although I suppose it would be maximum) value for the negative quality % of a : -0.94%, call it -1%

Which if accurate would mean quite the significant throughput increase from a beacon on your quality setups with high quality modules

[iron root]

assuming its only -1% for a

[jaunty citrus]

Personally I believe that means the true value must be significantly more impactful, closer to -5%

[iron root]

yeaaaah

[jaunty citrus]

-2.5% would match 's 2.5%

[merry spindle]

Bright flashing text is filtered by modern people brains.

[half raptor]

Day 1 there will be someone either on the discord or the forum complaining about speed modules making quality "not working"

I don't care how many tooltips you put in the game, some people just won't get it

[iron root]

for what its worth
its a pretty reliably good filter

[jaunty citrus]

They'll be told it's intended and they can accept it or be upset I guess

You can't help everyone, unfortunately

[frank warren]

nobody reads - but I also don't feel bad for those who don't read

so it really goes both ways

[grand igloo]

anyway I think gleba might have just usurped vulcanus for quality component manufacturing

[jaunty citrus]

I curious how catalytic spoilables will be handled for quality

If you have to maintain some base input for the catalist to work, (or it just eventually spoils) then there's no real way to maintain a quality stockpile of them

So the only way to get something like quality prod3 is to loop the product itself

[iron root]

yeaaaaaah
quality p3s look like theyre gonna be locked to recycle looping

[small egret]

Cool idea. There are some interesting ways to mix these features.

[obsidian crescent]

My understanding is that catalytic recipes cannot use quality modules. So this is never a question.

[obsidian crescent]

It would be interesting if captured nests generated quality eggs based on the quality of the bioflux they receive. But that's unlikely; they probably treat bioflux like a fuel.

That being said, I forgot that on Aquilo, you gain the ability to build captured spawner nests like buildings. Maybe these ones have module slots and/or treat eggs like a recipe.

[grand igloo]

once again, will specify: this is likely not an API limitation and is just a balancing convention if I had to guess

the devs seem to be keen on letting people make questionable decisions when modding if it doesn't break the engine

[obsidian crescent]

When the developers (temporarily) stopped prods from supporting catalytic recipes, did they do it in a way where mods could put it back in? I don't think they did.

[daring siren]

Very likely they just stopped the recipes from accepting prods

[grand igloo]

I seriously doubt they hard-coded per-recipe limitations like that

recipes already have a way to specify "these effects should not work on this recipe"

[obsidian crescent]

Why wouldn't they?

[grand igloo]

It's horrible practice.

And probably slower performance-wise than just marking the recipe as "prod effect no work"

(since it's adding an additional check that consists of multiple lookups)

[obsidian crescent]

Why? If the developers consider quality modules as applied to catalytic recipes to be broken, why would they force themselves to find every catalytic recipe in the game (even new ones they add) and specifically say "does not support quality"? It'd be far more ironclad to just code a thing into the engine saying "catalytic recipes can't handle quality."

[obsidian crescent]

It's a check that happens once when a recipe is set.

[grand igloo]

Because, for one, there's no "catalytic recipe" prototype in the game. If the game were set up that way, then we'd see it in the prototype structure.

Secondly, the developers could quite easily set up a script to manage their prototypes to keep all the benefits of hard-coding while taking none of the downsides. (in fact I bet they probably have some helper scripts to batch-change prototype properties conveniently)

[obsidian crescent]

No, but the engine clearly knows when a recipe is catalytic, as it must make adjustments for how productivity works. That is, when the engine is loading recipes, it checks the inputs and outputs of the recipe for catalytic components and sets up some data accordingly.

I'm saying that the way they turned quality off for catalytic recipes is by adding a flag to this catalytic detection code that says "this recipe cannot use quality".

[grand igloo]

It's likely that the 'catalytic detection code' is baked into the prod system

i.e. they changed the prod math from "output" to "output - input"

so

[obsidian crescent]

Why compute that every time instead of just computing it once when you initially loaded the recipe?

You read the recipe's data and convert it into the internal data structure. Part of that computation means detecting catalytic recipes and adjusting some things accordingly. Why not just add a thing there that turns off quality?

[grand igloo]

Because, again, it's bad practice (leads to unclear prototypes)

unclear prototypes as in

if I set the 'allows quality' flag to true, I fucking expect it to allow quality

If it overrides to false because I included an item as both an ingredient and an output... that's bad practice

[obsidian crescent]

Again, I go back to the point of when productivity was turned off for catalytic recipes. Do you think they did it at the prototype level or the engine level?

[grand igloo]

Prototype level.

[daring siren]

Prototype level.

[grand igloo]

If catalytic recipes did have this sort of distinction, then guess what

we'd see it in prototypes

[daring siren]

If we have a portable version where it was disabled available to download, it's easy to check

[grand igloo]

We'd have a NormalRecipePrototype and a CatalyticRecipePrototype

[obsidian crescent]

Do recipes even have a setting to allow/forbid quality?

[grand igloo]

YES

I am almost certain there is a way to do that

actually apparently the restrictions are machineside

...THAT'S RIGHT

yeah okay so it might be hardcoded

though it might also just be that the qual roll doesn't occur for the catalyst type

which would make a lot of sense

@daring siren no need

I went into the docs and recipes actually don't have module filtering abilities

so it has to be hard-coded

Got confused with the module limits on machines themselves

though I will say it's probably not a universal 'no quality on catalytics'

if I had to guess it's 'the catalyst doesn't roll quality'

since that's partially in line with how prod works now

[daring siren]

It's not hard-coded, it's on the module prototype side, under "limitation"

[grand igloo]

WAIT A MINUTE

No I was right

And yeah, there's a module-side limitation

but for prod there's also

on ingredient and product prototypes

"catalyst_amount"

it's auto-calculated on recipe load by default as @obsidian crescent expected

however it can be overridden

https://lua-api.factorio.com/1.1.110/types/ItemProductPrototype.html#catalyst_amount

[daring siren]

That is correct. But this was added for this reason.

It works for prod because it's a single item stack type, and just having a larger number.

It doesn't work for quality because there are several possible output stacks, where the catalyst remains in the same quality, and the non-catalyst side can increase.

[grand igloo]

unless the code is

"if catalyst_amount >= 0, don't roll"

which it could be

[daring siren]

I'd wager it's not.

[grand igloo]

true

[daring siren]

But I don't care for this enough to discuss. It's fruitless

[grand igloo]

I'd also make that wager considering uh

recipe limitations on modules

prolly easier for them to just have a script for qual modules that sets an exception whenever they create a catalytic recipe

[grand igloo]

https://forums.factorio.com/viewtopic.php?p=615425#p615425

actually this might indicate that there's an "ignored_by_quality" boolean property on recipe products

[iron root]

as far as prod and catalysts go
my understanding is that each recipe has separate "normal" outputs and "prod" outputs, and for most recipes theyre the same

and while that works for prod, quality would require more separation

[grand igloo]

nope, each recipe has outputs, and there's a property for those outputs (that's calculated at the end of data stage if it's left unset) that deducts from the output amount for the purposes of prod

and production statistics

but they're apparently split in 2.0

[iron root]

huh
i guess the format i saw was just for one particular exporter for calculators

[grand igloo]

and if I had to guess, they may just have
ignored_by_productivity (int)
ignored_by_stats (int)
ignored_by_quality (bool)

[small egret]

My guess is that catalystic recipies support quality as normal by default. Based purely on my conjecture that it will let you make ledgendary fish by fish breeding to use in spidertrons

[iron root]

I think the devs have said that the only way to get quality fish is by recycling spidertrons

[burnt reef]

can you at least breed quality fish once you have them?

[obsidian crescent]

Yes. While catalytic recipes cannot use quality modules, they still operate under the normal rules of quality. So if you get X quality inputs, you get X quality outputs.

However, if the fish breeding recipe requires an ingredient other than fish, you also need to make that ingredient in equivalent quality. Also, fish spoil, so you need to take that into account (although higher quality items have a longer spoil timer).

[iron root]

fish spoil now?

[obsidian crescent]

I think one of the LAN people accidentally let that one slip.

[iron root]

i guess it makes sense
still sad

[obsidian crescent]

Has anyone done calculations on the cost of raising a non-recyclable item from Q1 to Q5? That is, if all it recycles to is 25% of itself, and you have no other way to do it, how many items do you need (given a particular arrangement of quality modules in the recycler) to make a Q5 item?

Also, do we have any information on whether any of Gleba's fruit-based intermediates are recycle-able? I know you can't undo baking or mashing (the former is a furnace recipe, the latter has a byproduct), but can you undo, say, Jelly-yum manufacturing?

[iron root]

iirc we did a lot of stuff like that back when quality first dropped, just raw recycle loop efficiencies

and it was something around 40x if you could prod whatever was being recycle looped

if theres no crafting back and forth between an item, youre just trying to self-recycle to get to then the math should be fairly simple

[obsidian crescent]

I've more-or-less forgotten how accumulated probability math like this works.

[iron root]

i just threw it in a spreadsheet and iterated till it mostly converged
I think i got something like 83x for a no-prod case

[obsidian crescent]

That's a lot of fruit.

[iron root]

with no back/forth crafting its 4:1 both for the recycle ratio and the quality up ratio (4x ), so 16 items in per qualitied up item out
course the next step is to handle the chance for multi-ranking

[grand igloo]

According to the linked thread, auto-detection is removed for catalytic recipes.

[grand igloo]

it was around 40x if you could prod exploit

it was somewhere around 70-80x iirc if you couldn't

[iron root]

hm.
Yeah that might be the case, cause iirc i was saying something along the lines of the 56x relying on prod effects we didnt know of yet

so the baseline case would be worse than the 56x stated in the fff, which mean the 80x case is probably more correct

[grand igloo]

that was bc it was a snapshot estimate

[grand igloo]

iirc

[distant eagle]

cant remember, no productivity in recyclers, correct?

cause i get 2727 initial items recycled till only 1 Q5 is left

with 24.8% quality

[iron root]

yep, cant prod recyclers

[jaunty citrus]

Yes, but I was more wondering about the issues with maintaining a quality stockpile of spoilables to maintain a catalytic loop of.
If the product is always as spoiled as it's least spoiled ingredient, then any catalytic loop you have where the catalyst/product spoils will eventually all spoil, and you'd loose your catalyst.
The only way to maintain the catalyst stockpile would be to continuously feed in fresh catalyst to maintain a higher average spoil%

And maintaining quality with all that would be even more challenging, and most likely impractical if you wanted just the highest quality

[obsidian crescent]

Given the fact that the catalytic loop for ore bacteria functions, it seems likely that a recipe can override the normal rules for freshness. This would likely use the same tool that allows spoilage->nutrients to specify that the nutrients are half-spoiled.

So catalytic recipes around spoilage can just state that the output will always be 100% fresh. Which makes sense with fish breeding, ore bacteria growing, and egg manufacturing if that is indeed a catalytic recipe.

If that is the case, then the main issue for maintaining quality will be getting any other ingredients needed for the recipe. That is, you can make a single higher-quality ore bacteria ready for catalysis. But how do you make the quality Bioflux to go with it? Every single time you need more quality ore.

[jaunty citrus]

I'd guess you're correct on catalytic spoilable recipes resetting the spoil, it would make the most sense given what we know about how recipes and spoilables work.

[obsidian crescent]

This is why I'm wondering if any of Gleba's biochamber recipes will be recycleable. Because if they're not, then there's basically no way to use ore bacteria culturing to make quality ore. It's just too expensive to make quality bioflux if you can't recycle it into its inputs and try again.

Not to mention the fact that the Biochamber only has two module slots...

[jaunty citrus]

But how do you make the quality Bioflux to go with it? Every single time you need more quality ore.
It's questions like this that make me challenge the viability of making quality intermediaries long-term.
How many machines/products that you want have intermediaries similar to bioflux that seem they would have no good way to make large quantities of in high quality?

[obsidian crescent]

It depends. Fluids aren't a problem, so things like holmium solution and electrolyte are fine. Holmium plate comes from a liquid, so it's non-recycleable. But superconducting wire and supercapacitors are probably recycleable.

[jaunty citrus]

It's all well and good to get circuits, module 2s, plates, etc. But if you can't get biter eggs in the same ratio then most of it will just sit there doing nothing

[obsidian crescent]

Tungsten carbide is, shockly, an assembler recipe. So I can't see a good reason for it to not be recycleable. Tungsteel is a different story.

[jaunty citrus]

I'm not brushed up on the Vulcanus recipes, what was tungsteel used for again?

[obsidian crescent]

Almost certainly BMDs, but probably also green belts. We actually don't know much about what other items are on Vulcanus.

There are clearly some final products where direct recycling is the most effective way to make them. The main off-planet buildings being good examples of those, as well as modules. But things like beacons, platform components, chemical plants, even assemblers, those all seem like good candidates for production from intermediates.

[jaunty citrus]

Hmmm, really the only things I can feel confident saying are definitely recyclable are end products and anything in the crafting chain for stuff that comes from scrap. (not necessarily recyclable into their ingredients)
I'll give leeway to assembler recipes because of how it was described the recycler recipes were generated.

[obsidian crescent]

As for quality biter eggs, we do eventually get the ability to manufacture a captured nest building. Presumably, such a building doesn't need a constant infusion of Bioflux, but that means it would likely have something else: a recipe (presumably taking Bioflux 😉 ). And thus, if you could get quality inputs, you can make quality eggs.

Or you can just quality cycle biolabs. Which... sounds way easier and you can do it way earlier. You don't need Q5 right away; a few +60% speed labs would be great on their own.

I mean, what else are you going to do with all those eggs you're not sending back to Gleba for mysterious uses (or turning into prod module 3s).

[kindred crater]

they said the captive spawners you make do get a little bit quirky at night

so they might need a constant source of something even more expensive than bioflux

[iron root]

all i found is just "and the captive spawners you make will have their own quirks..."

so we know normal captive spawners begin making biters again if they arent given bioflux
but placeable captive spawners you make are .... ???

unless the nighttime thing was from an outside of fff bit

[kindred crater]

it was a shitty fnaf joke that i for some reason expected people to get

[iron root]

ah yes
well, given the devs literally used "quirks" in their own descriptions, well, yeah theres a few collisions there
either way, the actual crafted captive spawners apparently are going to be different from "natural" captive spawners

[kindred crater]

maybe i should've just passed it off as being an attempt to intentionally spread misinformation on the internet

[iron root]

nah, spreading misinformation is overrated

[kindred crater]

i will automatically believe anybody that says something was spotted in the background of b roll or said by a dev in a reddit comment

anyways

i can't imagine what quirk they might have other than needing fuel

like there's just not many ways for a factorio assembling-machine-like-thing to be quirky

[obsidian crescent]

Every assembly craft has a random chance to spawn a biter.

Or, more to the point, it spawns eggs that have a random freshness. That can sometimes be zero.

[iron root]

well, we'll find out soon enough

[kindred crater]

that would actually be pretty good

you're allowed to put the artificial spawner wherever you want which means it should be much easier to stock turrets around it

and it would make having a large field of them all speed moduled be... difficult

i assume the eggs spawn biters based on evo

so eventually you'll be getting a significant amount of behemoths

[obsidian crescent]

Um... why? Assuming that artificial spawners even can use beacons, there's still plenty of room for laser or gun turrets. They're only 2x2 buildings after all.

[kindred crater]

idk

[half raptor]

if they are random, I hope they cap it to be at least 50% fresh

[grim maple]

It's interesting how quality became the most contentious feature in space age

I still absolutely hate it.

[obsidian crescent]

At this point, there's basically a large group of people furiously debating about how best to take advantage of quality and when its worth investing in, and a small group of people who just hate it for whatever reason.

[lyric yacht]

good thing its fully optional so the people who keep going off about how much they hate it can just not use and stop constantly bringing up how much they hate it

[obsidian crescent]

I kinda doubt it. In part because whenever person A says that they have a problem, and person B suggests using a higher quality product as a solution, there is a non-trivial chance that person A is one of those people who hate quality and they weren't the one who brought it up.

Like, imagine some thread about end-game megabasing in SA and someone shows off their base and asks how to improve it. If everything is base-quality, someone in that thread is going to suggest improving the quality of some of the machines in question.

It's likely going to be more like people who hate beacons and/or modules. They kinda have to tell you going in that they're one of those people or else they'll be inundated with suggestions that they don't want. As such, they're going to have to keep bringing it up.

[jolly swift]

I don’t think nearly as many people are going to forego quality as we think

people dislike things they aren’t used to but I think after trying it most people will adapt

[obsidian crescent]

I expect that you'll probably have more people forgoing quality than we do forgoing beacons... but not by a whole lot.

[iron root]

we'll also probably have less people forgoing beacons in SA than in vanilla, just cause the beacon changes let you get a lot more out of the builds that arent annoying to make

[burnt reef]

one beacon on one machine is emotionally rewarding now

[frank warren]

You're seriously kneecapping yourself if you don't think about beacons in 2.0

I really really liked the new balance when I heard about it, but testing it made it clear that it was an even better change than I thought.

It's even nice for megabasers because it is much more viable to forgoe 1-2 beacons to do a more optimal layout. Getting 'every last beacon' isn't so impactful anymore, to the point that it is above all other concerns.

[half raptor]

yeah, even if you're just going for quality modules in things, putting efficiency modules will help a ton

[kindred crater]

the only actual "argument" i've seen from people that don't like quality is muh rng

[red kelp]

engineer who hasnt done uranium processing

[jolly swift]

Quality is just such a well made feature in so many ways

[jaunty citrus]

I think it fits Factorio very well, giving direct incentive to build massive factories for everything you need to build the factory. No more getting away with just one assembler making beacons/assembling machines/furnaces etc.

[grand igloo]

Nah one assembler making things will still be a thing, we’ll just make better intermediates :3

[jaunty citrus]

Sounds boring, I want a river of inserters!

[cold wadi]

In my base there's gonna be river of inserters feeding AIO assemblers making better intermediates

[jaunty citrus]

Well that's not a river if they're not flowing 😔

[jolly swift]

There’s just so, so many design choices that come with quality

[half raptor]

That's a good thing though. I'm bored of games that have one ideal tired design

Everyone is going to claim to be the best for the next year or so

[grand igloo]

Ehhh I feel like community will have a consensus for the most part in like a month

[steel falcon]

Seems like you either do a "top down" and roll the dice on final products and recycle, or "bottom up" and max base materials and build up from there, time will tell I guess.

[jaunty citrus]

The answer is pretty clear imo: top down is far less complex and easier to scale, bottum up is potentially more resource efficient as you get more recipe prod

[half raptor]

I like the idea of rolling for quality on each step and recycling excesses toward legendary. I plan to have inserters check for demand for current quality or higher before loading. If I don't have demand for legendary (probably because it just unlocked and I haven't built any stations for it yet) then it won't load and it'll wait for a demand to open up

If it measure a demand for legendary, it makes all the other inserters activate to load but if the destination is full for that quality, go to a recycle station

[grand igloo]

See I’m personally probably going to skip quality til recyclers at least

[half raptor]

Until I get recyclers I'll just put quality modules in the last step assembler and just be happy for a 2-4% chance to get an uncommon or rare product out

[grand igloo]

i’m not even gonna bother with that

[half raptor]

It'll be nice to have a few rare solar panels on hand, but yeah until we can do a fully scaled up recyclable operation, just doesn't seem important to worry about quality

[steel falcon]

Skimming enough base iron/copper/plastic to guarantee a blue armor seems worth before you blast off

[stray marsh]

A red belt of iron smelting with t1 qual is 33 uncommon and 3.3 rare irons per minute

[steel falcon]

I figured Quality in the miners and Prod in the rare-only smelter early on when you just need specific finite quantities

[half raptor]

You can put mk 1 quality modules on everything and skim quality off into storage chests for sure

[keen igloo]

quality labs and solar panels early are nice

[half raptor]

Quality panels are useful for platforms

[daring siren]

Quality to me is like modules. Not necessary for an enjoyable time and beating the game, but adds a ton of longevity and early power

[iron root]

and on the flipside
it takes a fair amount of design effort to leverage, but also has some (very) small investment options that are pretty easy

...similar to modules/beacons

quality modding only end product machines would be like the equivalent to doing un-beaconed moduled setups (that 1s+3p kinda thing idk)

[daring siren]

Speaking of modules/beacons, with the stronger beacons in 2.0, it's quite easy to replace old with and in the exact same footprint with similar output

In general going from to with center feeding is actually 1 wide thinner. The problem is length, and that's where come into play

[half raptor]

Also electric furnaces are nice if you're going for quality plates

Probably foundries for processing normal ores, electric furnaces for processing quality ores

[daring siren]

Speaking of, we would probably have a reasonable amount of and from research

Dunno if it's worth putting quality in though, just for recycling back for guaranteed

[steel falcon]

have we seen the actual stats on a Qual1 module

I thought it just had an energy penalty but I've seen it floated they also have a speed malus

[half raptor]

I believe they've said a is 1% chance for quality and we surmise that a base is 2.5%

all will be revealed when the media embargo is lifted

[wraith moss]

four Quality 3's have a 10% chance of upgrading, so 2.5% each sounds right

[kindred crater]

remember on that graph the chance is >= that tier and all the blank spots should say 100%

[half raptor]

well this chart is where we get the assumption that 4 is a 25% chance , though there's some question about floating decimal point that might acutally make th is 24.8% instead because Factorio tends to only keep 1 decimal point instead of 2

[kindred crater]

basic inserters from is another good one to use for higher tier inserters

[distant eagle]

Also, the 10% is really 10% or better. And is actually 9%

[half raptor]

indeed

I shorthand it to 10% chance to upgrade when talking to people

I don't say 9%, 0.9%, 0.09%, 0.01% I just say "yeah, a 10% chance to upgrade" and what it upgrades to is up to RNGesus

[distant eagle]

Unless the question is if they changed it to not be 24.8%, then it's still 24.8%.
#friday-facts message

[half raptor]

yeah.... hate that rounding error. I wish they'd just hard code it to 25%. I'm just gonna refer to it as 25% because it's close an easier

[jolly swift]

is there a reason we can't put in foundries?

[jaunty citrus]

No

It's just not useful for melting ore, as it's a fluid product

[jolly swift]

wait i get it

hes talking abt quality ore, not quality plates

gotcha

[half raptor]

Right. Since using the foundry to make plates can give you 4 modules, you can get more quality out of that step than using an electric furnace, but if the ore already has quality, you'd lose it when making it into a liquid and it would return to normal. You'd be better off taking the and putting it into an electric furnace for a chance at upgrading to rare or better plate but you would at minimum make uncommon plate out of it

[daring siren]

Putting prods to make more out of is also a reasonable option

[half raptor]

Im reserving prod modules for legendary

If you get a decent quality chance over 4 to 5 craft steps, it's worth the lost prod potential. It being RNG is unfortunate but if you want to mass produce finished goods to recycle loop, that's fine, but I think researching prod bonus and using special buildings will make it not too bad to go for quality on every step

You just have to set up some train logic to go recycle if all the input buffers are full

[solar osprey]

Still don’t think quality modules in miners are worth it once you’ve got the foundry. At the rate you’ll be able to consume materials, and the transport capabilities of molten metals in fluid wagons, mining directly into foundries could be the way to go.

[distant eagle]

Do we know recipes in the foundry?
If the green ore came from quality module miners, you might get more quality output if you use productivity mods in the miners and send everything to the foundry.

[obsidian crescent]

You should use speed modules in miners. You can use quality modules, but that requires sending the quality ore through a traditional furnace stack. But prods are never the right answer for miners, especially if BMDs are on the table.

[half raptor]

Mining productivity is one of those unlimited prod bonus researches you can get. Also you can always set more lines up. if you want to process some quality, for faster results without more miners, don't use because it adds to your researched prod bonus and it's just an additive result that only makes the mine last longer

[distant eagle]

The point would be that using no modules in miners (and not having to set up a foundry for each quality) might be better than electric furnaces

[obsidian crescent]

You wouldn't need a "foundry for each quality" because fluids have no quality. Putting quality ore in a Foundry makes no sense because you won't get anything different out of it. If you want to preserve quality from ore, you must use furnaces.

But yes, by the time the Foundry is on the table, you probably will have graduated from quality mining as your means to get quality stuff.

[distant eagle]

Right. The point I'm trying to make: if you have produced quality ore from a miner, and have then locked yourself into using an electric furnace, due to the fluid output from a foundry having no quality, your total output of quality gears,for example, might be lower than if you had used a foundry to process the ores instead.
You could mitigate the issue by using a foundry to process the higher quality ore, sacrificing the quality you already had, to produce a non quality liquid.
I say this because,to my knowledge, we do not know the production rates of ore-> molten metal -> plate/gears

[half raptor]

You say it like it's a bad thing. I suppose it is if you're trying to use the foundry prod bonus but I'd assume that if you were mining for quality ore, you'd be happy to have quality ore to start with

A at minimum will craft into with no other lucky crafts

[distant eagle]

I assume the goal would be a quality item made from ore, not the ore itself.

[half raptor]

Well yes

The whole game is built around turning ore into items

[distant eagle]

Common ore in a foundry with quality modules, might produce more legendary copper wire than quality ore->quality plate from a furnace->quality wire in an assembler.

[jolly swift]

The math doesn’t support this

A 50% bonus on a foundry doesn’t make up for losing the initial quality

I think the best outcome is to filter out quality ores and send those to a separate quality smelting place

Even in a huge base you would really only need a few furnaces since the vast majority of ores would be sent to foundries

[half raptor]

Not necessarily. Molten copper either turns into copper plate or copper cable, so if you go -> -> vs -> with a foundry step on both lines, the plates then cable will have a 24.8% chance to quality then an additional 24.8% chance on the assembler step

Both ways gives you the 50% prod on the foundry step

[distant eagle]

Do you know how much molten metal is produced from ore, and how much molten is needed per gear?
I started by asking if we knew the recipes in a foundry

[daring siren]

If you have that is

[jolly swift]

For the foundry to produce more quality despite losing the initial ores quality, the recipe would need to be around 6x cheaper

So unless one copper ore is turning into 24 wires its quality-negative to use a foundry on quality ore

also just from a design perspective it would make no sense if quality iron and copper ore were useless

[daring siren]

+50% built in prod, and 2 more modules slots is quite a big bonus

[jolly swift]

I’m actually not so sure about the math, how many module slots does the foundry have?

[wheat sluice]

4

EMP 5

[half raptor]

You'd only want to use the electric furnace for quality ore. Foundry can have a 24.8% chance on normal ore into molten into plate and electric furnace only has 12.4% but if you're starting with then you already have that 100% chance to get at least uncommon and now you're rolling for rares

My end game base is going to be modular and devoted to a one step craft for each quality level. Having dedicated areas to handle quality ore is part of it because big miners have 4 module slots. The odds of getting quality ore (post game) is going to be pretty good.

[distant eagle]

1 ore mined with 3 quality modules produces 0.81 common, 0.17 uncommon, 0.017 rare, 0.0017 epic, 0.00019 legendary ore.
Smelting the ore with 2 quality modules (and 2 prod for legendary) produces 0.71 common, 0.23 uncommon, 0.04 rare, 0.006 epic, 0.0004 legendary.
Crafting the plate into gears, with 4 quality produces 0.27, 0.17, 0.05, 0.01, 0.0007.
If you do 2 quality 2 prod you get 0.47, 0.22, 0.05, 0.01, 0.0006

If the recipe is 1 ore to 1 molten, then 1 ore in a foundry with 4 productivity modules produces 2.5 molten. If it takes 2 molten per gear, then 2.5 molten in a foundry with 4 quality modules will produce 1.41 common, 0.42 uncommon, 0.042 rare, 0.0042 epic, and 0.00047 legendary.

The productivity from the foundry, producing molten from ore, gives you more ingredients to roll quality on when you convert the molten to gear

[half raptor]

Yes but did you math in the potential quality plates to gear?

And yeah turning normal ores into molten is definitely better

I'm only talking about the quality ores

[distant eagle]

The math smelts all the qualities of ore into plate, and all plate into gear. With the stated modules

[half raptor]

Prod while turning the ore into molten is a good thing I forgot about, yeah

[distant eagle]

Maybe it's better. As I said, we don't know the recipes. But if it is better to use the foundry, then it's better to waste the quality of the ore and use it to produce common molten

[half raptor]

I just see going to plate then to gear as an extra step to roll for quality, though the extra productivity might be worth it too

I'm also going all in on quality for the meme and just to see if my end game idea will work

[distant eagle]

Maybe 1 molten produces either 1 plate or 1 gear, but plate to gear is still 2:1

[obsidian crescent]

FWIW, the last time we got numbers, the ratio of ore:gear through molten metal was still 2:1. At that time, 1 ore made 1 molten metal, and 2 molten metal made 1 gear. Of course, both of those steps still get 50% and 4 module slots, so it'd still be a 2.25x improvement in output even without prod modules. But it does mean that two-step fabrication (make plates from molten metal, then craft gears) gives you greater productivity if you use modules for everything.

The downside is that Foundries have crafting speed 4, while the assembler can only do 1.25. And the speed of the gear metal casting recipe is almost certainly the same as the assembler version, so one Foundry can do it way faster than a Foundry making plates + assembler.

While the amount of molten metal produced and consumed has undoubtedly changed, I doubt that the ore:product ratios have. That is, now you may get 3 molten metals from one ore, and it takes 6 molten iron to make one gear. But the overall recipe ratio of ore:gear is the same.

[gleaming quartz]

Thinking on the idea though, say you have a blueprint where one of the components requires less than 1 assembler as part of the targeted production. You can essentially place quality in there for free if its addition doesnt put the assembler count over 1.

[daring siren]

Or in general if the ratios match

e.g. if you need 3.1 machines and put 4

You can qual it down almost 25%

[distant eagle]

You'd reduce the output to be <75% of an assembler without modules. Since quality also lowers speed

[daring siren]

That's exactly the point. If you only need 3.1, then reducing the speed doesn't hinder the production

[distant eagle]

Just pointing out that although 25% of the output is uncommon or higher, the output of common is not 75% of normal.
Or did you get 25% another way?

[grand igloo]

Yeah, but you could also just loop intermediates until you got high quality ones

[grand igloo]

quality smuggling from plastic absolutely makes this the case

[earnest oar]

what is quality smuggling from plastic?

[obsidian crescent]

I believe that's taking advantage of the fact that fluids don't impact quality and that the Foundry's LDS recipe uses molten copper and iron instead of copper plates and steel. Because the only solid input is plastic, if you feed it quality plastic, you get equal quality LDS. Which can immediately be recycled into... 5 copper plates and 0.5 steel plates.

So if you can find a way to get high-quality plastic (perhaps from quality-cycling grenades, which makes high-quality coal to make plastic from), you can make high-quality copper plates at less than 4x the cost of regular copper plates.

[earnest oar]

Seems interesting. Are you sure the devs won't disable quality outputs from foundry or something?

[obsidian crescent]

That would be really weird. No other building has quality specifically disabled for it. If they decided that this was some exploit, they could just go back to having the Foundry use plates.

[iron root]

quality smuggling from plastic is no less exploitative than printing material from a 300% prod point if you ask me

[obsidian crescent]

It should also be noted that it took the collective wisdom of this Discord less than 24 hours from being told that LDS took molten metals to figure out this exploit. So I'm pretty sure the developers have been aware of it for some time, yet they've kept it this way.

[earnest oar]

Without any exploit and using regular assembler, recycler, modules on intermediate products without prod research, does it cost ~80x for legendary? Is that confirmed or just speculation?

[iron root]

it was mathed out

[earnest oar]

Didn't the devs share a 56x number in the FFF?

[iron root]

yep
and no one was able to reproduce it with the info we had

not even the devs

[obsidian crescent]

... that's a very particular question. Yes, if you take away all of the really cool tools for reducing the cost of quality goods, quality stuff can be very expensive. Which is what makes quality an interesting mechanic: you can find ways to game the system, to produce quality cheaper than you might otherwise.

If the developers just wanted a Q5 recipe that was 80x more costly than a Q1 recipe, they could have done so much more obviously.

[earnest oar]

I like the quality mechanic in general. The optimal method being complex / different for different items, different in late / mid / early game would be cool. But one simple way being better than all other ways would be a shame.

[obsidian crescent]

One simple way to do... what? All this method gives you is quality coal, iron, steel, copper, and plastic. It doesn't handle stone, tungsten, holmium or any of Gleba's products. So while you can make module 2s, you can't make any module 3s with this method alone.

Oh, and it can only be done on a planet with a lot of coal reserves and no real use for them (ie: not Vulcanus, Fulgora, or Gleba).

That's not to say that it's insignificant. It's very significant. But it doesn't undermine the mechanic.

[iron root]

hes just saying it would be a shame if there was a simple optimal way for quality in general

and consensus is right now there really isn't

the most logical thing i see is doing a skimming setup until lategame where you can get prod juiced really high and then start doing recycling for
but there's still a lotta wiggle room in that

and it could be the "most logical" way in the same way main bus/cityblocks are the "most logical" ways to do factorio bases
which is to say theyre not hard and fast rules

[main ruin]

I bet it will be like modules, a lot of variations of optimal paths based on your choices / constraints

with end state being the same for all, but way, waaaaayy later than modules due to spaceage's length

[grand igloo]

So. The Foundry can make certain products from a combination of solids and molten metals. Fluids aren't factored into the quality calculation, so you can use quality plastic, for example, to turn molten iron and copper into quality steel and copper at .25x-1x efficiency. (after recycling)

[grand igloo]

That was a snapshot estimate

[main ruin]

Finally a coal sink for solar bases

[grand igloo]

which proved to be inaccurate

[grand igloo]

Eh. I think skimming setups are going to be more trouble than they're worth considering most quality products will be one-time expenses.

[iron root]

possible. I'm wagering on there being enough of a benefit (and a reasonable supply) to just build my entire factory from material created via quality modding miners

[gleaming quartz]

Yeah that works too. Implementing some quality machines into the usual science production prints will be interesting. Wonder if it could change the calculus for speed running

[jolly swift]

We’re talking specifically about quality ore from quality miners

[grand igloo]

I don't think it's worth it when you can tier skip straight up w/ quality smuggling

[jolly swift]

You can do both

I.e filter out uncommon+ ores for separate processing

[grand igloo]

why would you want uncommon ores when you can tier skip to legendary

[obsidian crescent]

How do you plan to make quality tungsteel?

[grand igloo]

Dunno what the recipe is

[iron root]

given that tungsten likely has no liquid form, you wont be able to quality smuggle it upwards

[obsidian crescent]

Tungsten ore + carbon+molten iron in a Foundry

[grand igloo]

then you'll have to quality cycle something made with tungsteel

[iron root]

yep

[jolly swift]

Presumably you dont have infinite quality plastic

[iron root]

or just cope with only and via quality modules in the tungsten miners

[jolly swift]

And you can use the common ore for quality smuggling

[grand igloo]

Eh. Coal's cheap enough and as LDS productivity approaches 300%, the input requirement will drop to 0

[iron root]

provided we have a LDS recipe prod

[grand igloo]

true

[iron root]

which...
there's a good chance we will I think, but it's still only an "if"

[obsidian crescent]

Coal is cheap enough on Nauvis. Notsomuch for Vulcanus, where you need a bunch of it for liquefaction, regular plastic making, and tungsten processing.

[iron root]

we also dont know coal patch sizes on vulcanus

all stuff that we just have to get a feel for by playing