Quality

[final nimbus]

Quite a lot.
(it's not worth it for any of them except maaaybe agriculture science)

[quick rampart]

i believe consumables should be designed like this, more durability

[frank warren]

It would be the case that none of them would be remotely worth making.

[quick rampart]

well when it comes to consumables, i think the optimal way is to plugin in some random quality modules and don't recycle the output, when you get high quality modules the prod and quality will multiply

i might do a calculation later

[final nimbus]

For consumables like ammunition, you should always use
Trying to get quality science packs makes a lot less sense than just increasing the amount with

[quick rampart]

no, tier 1 prod +4% while tier 1 quad +1%(10%*0.1)

[gusty trench]

Gray science too. Quality walls and quality ammo are pretty useful

[young bison]

I've thought a bit about recipes plus productivity. It feels like "if you put items in for the crafting cycle that triggers an additional prod, you get legendary items out" would be an emergent puzzle. It's automateable using circuit networks, after all. It rewards advanced players who design a circuit to generate legendary items much more quickly, and allows less skilled players (like myself) to just brute force quality by quantity.

[frank warren]

Don't you think that behavior is a problem?

[daring siren]

Yep. It's going to work very nicely specifically with the +50% prod buildings

However, there's some chaotic behavior that makes it not consistent

[jaunty citrus]

Feels far more like a hack than an intended mechanic though

Like how you can stop all expansions by killing all biters on the map using artillery/spidertons

Plus it has the potential to be a massive noob trap

[half raptor]

Better to have production lines for each quality in my opinion. Might feel wasteful to have a whole line dedicated to making but as you scale up (especially if everything is delivered by train) that line will get busier as you go without having to build another line

Probably don't need to worry over much about end game designs until you unlock the 3 intermediate planets since we need cliff explosives, recycler, and whatever benefit from Gleba (what exactly I don't know) before we start reshaping the worlds to our benefit

[frank warren]

Why do we need anyways? What good does it serve?

[fallen urchin]

More durability means you can just replace them with quantity, which makes them pointless

A repair pack costing 56x but only having 6x durability is not going to do any good

[quick rampart]

in situations where amount of items is restricted for example launching a rocket, i believe it makes sense to compress them

and in rocket each item counts their weight separately, so the problem that different quality items occupy different slots doesn't exist

and it's not necessary to recycle them

[daring siren]

Sometimes quality exists but irrelevant i.e. , but removing that option is more expensive than just letting it exist

[quick rampart]

buildings last forever so it makes sense to recycle them to make , but consumables are not

[fallen urchin]

For rockets, I think the cheap consumables can be mostly produced locally, and expensive consumables will have costs outweigh than the rockets

[fallen urchin]

I think the main motivation for quality is , which simply can't be replaced with quantity

And can improve the utilization of those , the same does and

And buildings being permanent means they will eventually pay off, while consumables don't

[chrome mauve]

How many of consumables can be prodded? Cause if they cant be, its not like you really lose anything by putting qual mods there

Price is that you dont get that many of them

[quick rampart]

most of them are under the intermediate products category, and ammunitions are under combat, but most of the time, insert prod modules if you can. #1215078107334057984 message

just realized cannot use prod, so if you want to get more, you have no choice...

[gusty trench]

Quality consumables can offer gameplay because resource scarcity is not the same across all planets. On Fulgora, you might have a consistent excess of LDS or whatever, so turning that into quality consumables makes them easy to ship while using something that would have been voided otherwise

[half raptor]

Only intermediates can be productivity boosted but if you prod the ingredients, consumables are effectively cheaper

[merry spindle]

It's like upgrading to T3 modules from T1 but better and cheaper

[frank warren]

I mean 'any quality' recipes (the thing being discussed)

[frank warren]

see here

[obsidian crescent]

I'd say the main reason it exists is to reduce the need for learning circuits. Kinda.

It also acts as a way to get rid of higher quality stuff if for some reason you're fine with using to create lower quality goods. It's an extremely rare need, but its not unreasonable.

[tall musk]

How is this thread still going??

[frank warren]

Why wouldn't it be?

[tall musk]

There's not that much to the system

[half raptor]

Yeah but people come with occasional questions and didn't know the thread is here

If moderators want to delete the thread, those questions will be out in space Age chat

[kindred crater]

also discussion of how quality affects x new item(s) in FFFs

[frank warren]

I’m not entirely sure how you can reason that. Can you defend your justification?

[tall musk]

No

I already have half a headache

[frank warren]

I mean, I’m just saying we had to write code to even remotely understand the implications of quality cost, so it can’t be that simple :P

[tall sandal]

we love quality

[half raptor]

every so often we get a new believer of the faith and they come with questions. They've been lied to by the pushers and must be shown the life of the quaint

[uncut helm]

wat about e-quality?

[tall sandal]

epic quality? its nice, but not as good as l-quality

[dense kraken]

what do you do when you have too many high quality copper wires / too many low quality iron plates to make circuits?

answer in less than 2000 characters

you have 4 hours

[half raptor]

Not all items are made equally, some are made like all the others and some are a higher class. Do not worry though, as all are made equal in the eyes of the recycler who will cleanse their lack of quality

[dense kraken]

only 3/4 of the time :p

[half raptor]

First of all, you send excess to recyclers to go for and beyond. Second, if you wind up with excess you can substitute your excess legendary in the epic and down recipes since the recipe relies on the lowest quality item and they allow you to substitute one of the items with higher quality. Thirdly, if after all that you STILL haven't fixed the issue of too much quality cable, then you're not paying attention to your plate production

[dense kraken]

no way I'm using recyclers

[daring siren]

Personally I'd sink in science packs

[dense kraken]

I'd rather short-circuit assembling machines with ones to feed more low quality items where it's needed

[daring siren]

That's great too

[half raptor]

Those ideas work too

So glad there's not just one "correct answer" I hate games that have one linear answer

[dense kraken]

that's where the fun of factorio is

reminds me when they mentioned how they purposely made nuclear ratios not even for this very reason

[half raptor]

Also consider that if you do recycle your excess cables and put the resulting copper plate on trains some of the quality plate might go to make low density structures and the like

It doesn't necessarily stay in the copper cable route if you keep your train bases modular

[dense kraken]

right, the logistics gonna be insane

most sane quality enjoyer:

now that I think about it, quality is like a difficulty setting without being one

[half raptor]

It is and one issue you might have to solve is how do you distribute your limited between your uncommon green circuits and steel and I think the answer is that if you're going to set up automatic recycling you'd need to be selective on what items are allowed to recycle so that product will build up until needed and not kill your supply of ingredient

I plan to start with just end products and pick and choose strategically what items are clogging the system due to over production and Imbalance

I feel it's fine to have one intermediate that takes plate is fine to auto recycle if you get too much as a means of keeping the flow of iron plate running but if I'm recycling all excesses from iron plate then I might wind up with a huge shortage. As long as I have one pressure release item then the plates can continue to flow. Even if it's wasting normal materials, it allows the machine to keep rolling for quality with each craft

So probably stuff like since that releases pressure on iron, copper, and plastic

[jaunty citrus]

Not a huge fan of any quality crafting because you can exploit how it works with productivity

[half raptor]

If you research closer to the 300% cap, it definitely is an exploit since you can get back all (on average) of your starting materials and just upgrade via recyclers

[dense kraken]

I'm taking two weeks off for the SA release, I'll have time to figure out

[half raptor]

Wish I could

By the time I get home from work you guys will have theory crafted the best blueprints

[dense kraken]

I have enough to take 7 weeks rn

[half raptor]

I'm going to avoid blueprints from the community at least for the first week with the exception of rail

[dense kraken]

yeah, better enjoy the expansion to the fullest

[half raptor]

Can't wait to have reversible refineries

[dense kraken]

can't wait to play for more than 36 hours straight

just to get an orange spider

[half raptor]

It'll feel so satisfying to have a legendary spidertron with legendary equipment grid

[jaunty citrus]

% speedrun

[dense kraken]

the size of legendary grid omg 😵‍💫

[obsidian crescent]

Honestly, I'm not sure if a Q5 Spidertron will matter that much. Q3 gives you a 8x12 grid, which allows you to put in more (quality) PFiRs. Sure, Q5 is an 11x15 grid, and if PFuRs are bigger than PFiRs, that could be quite helpful. But at the same time, high quality exoskeletons and PFRs are going to be more important than a higher quality Spidertron.

And way easier to get.

Q5 Spiders feel more like a cool thing to have rather than something you should "go for".

[dense kraken]

grid is 20x14

[kindred crater]

thats some serious griddage

[dense kraken]

wait am I right?

10×6
11x7
12x8
13x9
15x11

having odd amount of rows sounds cursed

[warped obsidian]

I think it's 14x10 at max level

[dense kraken]

nah, legendary is quality 6

unless you start at 0...

[kindred crater]

at least for the vertical row its a convenient place for batteries

not sure what to tell you about the horizontal one

[dense kraken]

yeah odd columns isn't that much of an issue (except who would like to multiply the cost of their armor by 60 just to get a few more batteries) but the rows... 🤷

that's why I assumed "each direction" to be up, down, left and right

but I guess it's just x and y

[kindred crater]

I guess put some solar panels in there for 2% more power generation lol

[dense kraken]

wait, direction or dimension?

can't be more clear than that

[merry spindle]

Who in the world starts with 1?
Pascal-man?

[dense kraken]

LUA

also regular human beings

[grizzled lagoon]

Excel, matlab... all things to make quick transitions between Python and other often used programs as confusing as possible

[kindred crater]

i wonder how viable shipping on platforms might be

you tend to have an excess of the stuff on nauvis

very viable I'm sure

[obsidian crescent]

Shipping it for what purpose? Making quality plastic?

[half raptor]

Rockets will cost 20 times less, so not very cumbersome

[obsidian crescent]

It might be useful for Vulcanus, where quality LDS can be made on the cheap from just quality plastic.

[kindred crater]

vulcanus already has coal I thought

[half raptor]

Vulcanus has coal

[kindred crater]

coal, calcite, tungsten, (and sulfuric acid)

[obsidian crescent]

You need that coal for oil products. Whereas Nauvis has a lot of coal but no real use for it outside of making plastic and explosives.

[half raptor]

It's your rockets

Would be nice if gave more cargo space

[daring siren]

What even is ?

How would you make one

Also, cargo space is not the limiting factor

[obsidian crescent]

Yes; the main reason not to ship coal (quality or otherwise) is platform storage space. So if your goal is to get quality plastic somewhere, it's better to ship the plastic rather than the coal.

Also, Nauvis has plenty of cheap petrol to make plastic from. So you may as well ship plastic.

[keen igloo]

ship barrels of petroleum to save on water on vulcanus

[obsidian crescent]

That's even less stack-dense than coal. One stack of petrol barrels only uses half a stack of coal. Unless you have more than 100% productivity for plastic (if there's a plastic prod research, a super-chem plant with 50% prod, or both, that's possible), it's better to just ship the plastic directly.

[kindred crater]

electricity

is more wh per rocket

[obsidian crescent]

I don't know what "wh" means. As far as we know, quality of fuel doesn't effect the energy stored in the fuel. It can affect train acceleration and top-speed, but that's it.

[burnt reef]

I think a wh is 3.6 kJ

[kindred crater]

watthour and quality affects item durability so would last longer in a boiler

[burnt reef]

also even if quality does effect fuel value, is still less energy dense than (assuming you can fit the same amount of each in a rocket)

[obsidian crescent]

Fuels don't have "item durability". They don't get burned the way science packs do.

[kindred crater]

oh

[obsidian crescent]

Well, I guess nutrients have freshness, which is a form of item durability. And those are biochamber fuel, so...

[kindred crater]

then i guess my really will just be for and, of course,

(throw a single one of those in your base on accident!)

this is why i will put them in my hotbar right next to power poles and belts

unfortunately according to my calculations it wouldn't deal enough damage to oneshot any particularly expensive buildings

except possibly things caught by 2 or more of the cluster explosions

[dense kraken]

if it's only for electricity then why not go for

[fallen urchin]

lvl 13 prod research is not cheap, and probably not reachable before SA endgame

[fallen urchin]

Why not use fusion instead

will probably be useful in trains

I wonder whether will have a fusion version

[fallen urchin]

IIRC damage upgrades also apply to friendly fire

[dense kraken]

idk really 😅

[dense kraken]

inb4 hypersonic trains

[half raptor]

Either you are one of the privileged few to beta test SA or you're driving the cost based on an educated guess. We don't know if the infinite scale is going to be the same in SA or even what the starting cost for each prod research will be. I imagine higher grade intermediates will have more expensive research costs and earlier tier stuff will cap at 300 first if we follow the "upgrade the cheapest" rule. I base my arguments on 1) confirmed information either by FFF or stated by a dev we can quote and 2) late game arguments because if 1M SPM becomes normal, those costs aren't going to be insurmountable.

Saying something is too expensive to worry about is missing the fact that it's an eventual goal we can build toward and that even as we approach the limits we will see some major improvements. I personally guess we can research it so high we won't need prod modules after a couple thousand hours of researching.

[daring siren]

Didn't a dev say level 30 is like 200 billion? or was it million

Ah 200M yes

#friday-facts message

And someone had an educated guess on the curve

[half raptor]

Even if most players only achieve 200% prod it's still a good return on investment.

But I think thats referring to researching a natural 300 without use of modules or special machine

[warped obsidian]

Oh that's actually not unreasonable at all

Given 1M spm is possible

[daring siren]

It goes out to be around 1000 * 1.5 ^ N

Which is cheaper than all infinites other than mining prod and followers

for 4x science (and other prods)

[half raptor]

I just hope research prod is one we can research, not just research speed

[warped obsidian]

I think they said it is

[half raptor]

Nice

[daring siren]

It's definitely prod research. It's in FFF-376

[half raptor]

That's how you creep toward 1M, prod everywhere you can

[daring siren]

That's a given

1MSPM is likely at max prod

[half raptor]

Seems like science prod can be an infinite tech with no cap

[daring siren]

It is

fr fr no cap

does mining prod affect asteroids?

[warped obsidian]

Probably not

It would be weird if you could get unlimited rock per rock

[half raptor]

Well the logic behind prod bonus should be we are wasting less of the raw material and we can produce extra items

You might be able to put modules into the collectors themselves (though might only be eff or speed)

Can't imagine quality asteroids

[obsidian crescent]

Considering that the Ag tower doesn't allow modules, and asteroid collectors are a similar kind of resource generator, it's likely that collectors don't require modules either.

[cold wadi]

Do modules have quality?

[obsidian crescent]

Yes, but this was clearly stated in the quality FFF.

[daring siren]

Of course. Every item has quality.

Even fish.

[gloomy torrent]

do you get that by launching high quality science pack into space?

[obsidian crescent]

... that's an interesting question. In SA, launching science packs doesn't do anything special. But presumably the behavior hasn't changed in a vanilla run.

A better question is this: if you launch a high-quality satellite, do you get an equivalent quality of space science back?

[daring siren]

High quality satellite does give high quality science. That was confirmed iirc

I wonder if consuming is available in base 2.0 with quality enabled

i.e. is it a spage achievement or a quality achievement

[frank warren]

It's kinda a pain in the ass to make high quality u235 right?

[daring siren]

Yeah you need to recycle it

Once you have 40 you can with quality 238 though, which you can get from fuel reprocessing

If you don't prod it

It's probably easier to just recycle loop it though

As makes u235 "free"

[frank warren]

Makes it no freer than iron plate

You'd have to loop either the u238 or u235 though, so the u235 is easier (denser)

[warped obsidian]

Does the rocket also need to be quality or just the launchee

[daring siren]

There's no quality rocket. Just the payload.

[solar osprey]

It’s been pretty much confirmed that quality fish can only be got by recycling spidertrons.

[obsidian crescent]

In SA at least. Unless they get rid of the "launch space science to get fish" thing in 2.0, then you could make quality fish by making quality space science.

[frank warren]

they did

[fallen urchin]

Does the BP item have quality?

I think this could be a cursed mod idea though

[daring siren]

I don't see why not

Also decon planner 🙂

[fallen urchin]

Also upgrade planner and BP books

The problem is, they are not crafted in any way, so it's not possible to obtain quality variants of them

Maybe if you put a BP into a recycler, you get 25% BP back, so qual cycling works on it?

[winter saffron]

We still don't know how to get fish in SA, do we?

[fallen urchin]

If there's a way to automate fish in SA, then it should be on Gleba

[half raptor]

feed fish your

[quick rampart]

launch space science pack to get

[jaunty citrus]

Doesn't work in Space Age

[daring siren]

Was that established?

I don't remember where a dev confirmed it

Well except JG confirming and uncofirming it

But I don't consider uncofirming as a confirmation of the negative

[jaunty citrus]

At the very least they removed the space science from launching fish. It only makes sense that they would remove the reverse as well.
That and JG's "confirmation" #friday-facts message

[daring siren]

#friday-facts message

He uncofirms the uncofirmation

[jaunty citrus]

Hmmm, Maybe I'm imagining the space science from fish

I could've sworn I've tried it before though

[daring siren]

It's fish from science, but a more realistic yield

[jaunty citrus]

Good question

[daring siren]

In vanilla it's a reference to Hitchhiker's Guide, similar to the "Thanks for all the fish" achievement. It supposes fish live in space, so when you give them science they come.

[solar osprey]

Actually. Wouldn’t you be trading with dolphins? 1 satellite = 1000 science, 1 science = 1 fish.

[half raptor]

Dolphins are the previous civilization on Fulgora.... confirmed

[kindred crater]

what if engineer is actually a member of the fulgoran race who lived long enough away from the dead planet to forget it

[wary beacon]

i wonder if the 'space station starter pack' they mentioned has a quality bonus besides hp

[kindred crater]

I wonder what that could do- send stuff back to planets faster?

[heavy ember]

More cargo space?

[wary beacon]

or if/when they show us the method of transferring items from orbit to the landing pad has quality, quality drop pods?

i figure it's probably not worth, just more hp, but it's neat to think about

[kindred crater]

definitely. it would be cool if there was some benefit/use case that justified getting it

[obsidian crescent]

If you're referring to the platform hub, probably not. It's mostly just a container.

[grand igloo]

yes. they scale up to +150% (or 2.5x) positive effects @ legendary

[cold wadi]

🤯

[jaunty citrus]

Every item has quality

[final nimbus]

We knew about modules with quality before we knew about quality.
It was part of a joke post for plausible deniability.

[half raptor]

That was in an old April Fools joke about "quality modules"

The real joke was that all of those blueprints in that image were referencing real things coming in SA

[lyric yacht]

anyway try the janky quality mod and have any functioning designs to make legendary stuff?

[half raptor]

That mod is good for simulating intermediate production

[keen igloo]

I've played with it a lot, but my designs are utterly unusable spaghettie at this stage, as I was mostly interested in what quality gear early would be a good thing to work on.

[half raptor]

The train interrupt system for sorting materials through stations will make spaghetti disappear

[kindred crater]

the best lesson I learned from ~100 hrs of janky quality is theres a time for quality, and theres a time for horizontal expansion. hours I spent messing around with it were hours I could've done more things to progress the game (more science, more outposts, more production, etc). excluding ammo and a few other things earlier on, unless I'm playing with multiple people in which one can work on it, I won't be going deep into quality until lategame

[keen igloo]

it basically confirmed my ideas on what was good(modules, labs, assembly, solar etc) and it was good, the surprises were some weird things, like gun turrets/laser turrets. That extra range is amazing for expansion/defense.

And beacons, beacons are great and the are rediculous even at uncommon/rare

[kindred crater]

putting in will actually be pretty neat

i hope foundries and emp are really expensive to make to incentivize moduling

[dense kraken]

I've been trying to figure out ways to handle varying ratios of higher quality over lower ones without resorting to recyclers
here you'll find two solutions to the study case of electronic circuits assuming T3Q5 modules and only and

I have yet to compare them but I need sleep (unlike sheldon)

white boxes are using quality modules, red boxes are using productivity modules

[winter saffron]

assuming T3Q5 modules
feels weird using ^ w/o recyclers

[dense kraken]

why so?

[half raptor]

Also assuming you're not researching any tiers of intermediate prod bonus

[winter saffron]

Your research is more interesting early in the game where you don't have access to higher tiers imo.

[half raptor]

The fact is that the math is going to be janky the entire way through and ever changing. Setting up recycling for imbalances is going to be common

Easy mode. Go for the entire way and quality loop at the end to make machines

Hard mode, go for and balance as you go

[dense kraken]

the first solution wouldn't suffer from that

the idea is as easy as firing a prod-only factory if low quality items are missing

because having too much high quality items only means you could produce more high quality of the other, meaning you'd have too much low quality of the other meaning you don't have enough low quality of the former 🙂

[half raptor]

I think bot malls are going to be great for handling quality machines. Logistics of various quality materials can be done via bots and request chests

[dense kraken]

wait I put the same factory twice on the second solution

[dense kraken]

what if I'm mad enough to not go the easy way tho?

[winter saffron]

I'm going to setup a train mall using lowest priority (excess) to cycle train -> recycle -> recycle -> .... into lowest intermediates of set Qual X and feed these into a mall.

[dense kraken]

shit my values are wrong

I guess I'll have to crunch the numbers again tomorow

[pale wraith]

Okay, I have a really dumb question that I've yet to have explained. How on earth does Quality not cause desyncs?

[rotund egret]

how would it?

whether an item gets an upgrade in quality is rolled on craft start

[pale wraith]

The randomness of production has to be synced such that when I get a rare circuit, you also see a rare circuit produced.

One of the first bits of advice when modding for avoiding desyncs is "Factorio is deterministic. Be very careful with randomness"

[frank warren]

google: PRNG

but I'll ask you a question: How could a computer generate a random number in the first place?

[solar osprey]

By generating a random number on 2 computers with the same seed, you’ll get the same result.

[pale wraith]

I'm aware of Pseudo-Random-Number-Generators. But I wasn't aware that Factorio initially had a global random seed, nor how you would "catch-up" a random number generator when someone joins the game.

[jaunty citrus]

Yeah it's used for uranium processing as well

[solar osprey]

And terrain generation.

[pale wraith]

Ah, of course. The tech was already built in for global randomness.

[frank warren]

A PRNG gives the same sequence of values every time. Because the 'catching up' player calls it in the same order, the same thing happens.

the PRNG is determinstic code like the rest of the game - nothing actually special about it

It's a totally not random number generator that just seems random and is distributed

[pale wraith]

Wouldn't that require calling the number generator an almost arbitrary number of times on load? I started playing at 10am and had something random going for 4 hours. Then my friend joins, do they just churn through 4 hours of generation?

[frank warren]

the state of the random number generator is sent when they initially join

think of an RNG like this

def get_next_output(value):
  

The seed is just the first value we pass in

for every subsequent output, we pass the previous output back in as input

[pale wraith]

Except of course we don't, we modify the state in some way with an expectation of avoiding loops. Often performing the operations on that state to get the output in whatver manner we've decided our interface outputs at.

I'm familiar with PRNGs, I'd just not encountered one which was syncable. Though it does make sense that there's no reason you couldn't serialise its state, they are fully deterministic after all. Interestingly I can't find much on syncable PRNGs, most game dev stuff focuses on the host/server informing the clients as to the result rather than syncing up their PRNGs.
I wonder if Wube made their own that was a little more amenable to serialisation than your standard implementation

[frank warren]

This is how I have always implemented a PRNG, as a function of (state) -> (next_state)

You can convert any mutable RNG to be written this way

instead of modifying the state, simply return a new one

[pale wraith]

It would be the functional way of doing it, serialise that and send it across. Why not

[frank warren]

Of course you don't pass in an input value / output value directly, but a state which has more information - which can be turned into an output, or derived from a seed

but that was implementation detail as far as I was concerned

but that is why you couldn't find anything - if you convert any of these state transition algorithms to be immutable they are also (trivially) syncable

[pale wraith]

It's probably because I've never bothered to actually implement one from scratch.

The Modding API does provide a random generator but it interestingly pushes you away from using it in it's own description. "If you aren't sure if you need to use this over calling math.random(), then you probably don't need to use this" which at least implies it has notable overhead costs to use. Probably because it immediately forces the other clients to be aware of it or something.

[frank warren]

not really how it works

all mod code is ran by all clients

idk why they suggest not to use it, maybe just complexity... but there is no 'syncing' or 'awareness'

even when you open a gui - every other player has you open that gui

[jaunty citrus]

Factorio's multiplayer straddles the line between stupid and genius in my mind

[frank warren]

I kinda agree

[jaunty citrus]

"just make the game fully deterministic and have everyone simulate it" sure is a solution

[young bison]

The reason why it's discouraged is because it's actually really easy to accidentally make a non-random number generator like this. For example, if you create a new random number generator every tick, you're starting with the same seed every time. This means that it'll always give the same series of outputs.

[pale wraith]

Do you happen to know if the modding engine sets a synchronised lua random seed when it loads? math.randomseed

[young bison]

If you create a new one with no parameter, it uses the starting map seed as the random seed

https://lua-api.factorio.com/latest/classes/LuaGameScript.html#create_random_generator

[pale wraith]

I meant the default lua one that it encourages you to use instead

[young bison]

https://lua-api.factorio.com/latest/auxiliary/libraries.html#•-math.random()

Here, it also starts with the map seed (in the control phase)

However, the random number generator is shared between all mods

and also the core game

math.randomseed() does nothing in Factorio

[dense kraken]

the quickest answer would be: like uranium randomness

[pale wraith]

Bonus unrelated question: We're not expecting different tiers of the same item to stack in a slot are we? So if you have Uncommon and normal circuits on a belt. They do need to be sorted else you could get blocked assemblers as it holds 1 uncommon and has nothing but normal to scoop

[dense kraken]

afaik yes

[rotund egret]

that is correct

[pale wraith]

Additionally, extending the above scenario, if the assembler those uncommon circuits are being fed into have no quality modules, we're expecting that you will always get normal products out and that the quality of the input was effectively "wasted"

[rotund egret]

output quality cannot go below the smallest input quality

basically if all your ingredients are uncommon, the result will be 100% uncommon

no quality modules required

[pale wraith]

Interesting, do we know if it's guaranteed to be the smallest input quality?

[daring siren]

You choose the specific quality you want

And only that quality input items are accepted

[pale wraith]

Ooh, that sounds like it could do sushi belt shenanigans.

Mixed quality input line on a sushi, Assemblers pulling from with quality set

[daring siren]

Whole belt reading helps this too

[pale wraith]

Ooh, true and cursed. Setting the recipe for the assemblers based on what's available

[iron root]

I think I have some ideas why they did this
First and foremost, it removes most forms of latency from multiplayer, which just overall makes it a much better experience. As to why they didn't use a predictive networking scheme to deal with this instead, well...
it also greatly reduces the amount of information that has to be sent back and forth. You're no longer communicating the entire game state every tick. You're just communicating player inputs. Given just how large the game state can get (all the maps with factories on them!) and the fact that it has to simulate all the time, as you cant really section off and unload far away parts of the map when they contain active machinery, the devs might have just found out that the network requirements for sending the game state are enough to consider going deterministic

and you cant just send the nearby chunks of a state to a player, cause that wouldn't work well with remote view, which is going to become even more important

[dense kraken]

now I did the math properly

the results are unexpected

the quality ratio is the same for both

they make 11/9 or about 1.2 times more than

that's super convenient because I'll thus be able to compare the ore per circuit regardless of the circuit quality

solution 1 requires 0.8 iron ore + 0.37 copper ore per circuit
solution 2 requires 0.8 iron ore + 0.3 copper ore per circuit

in short, I haven't found a better way to match quality ratio than to apply the same amount of quality steps to each intermediate

[iron root]

yeah, i dont think theres any way around that one

[kindred crater]

i wonder how manipulatable factorio rng is

the tas might actually get crazy speed boosts from quality

[lapis leaf]

been trying to get the maths sorted for each quality level and different properties, but i'm having trouble getting the negative amounts, like power consumption decrease or the big drill's resource depletion rate, I can get to the 17% but only with weird numbers

[obsidian crescent]

Well, we don't know what those are, so there's not much you can do about it. We're unlikely to have all of the numbers for these things before release.

[lapis leaf]

i'm mainly trying to intuit a sensible range of numbers based on the 50% to 17% we have

[obsidian crescent]

It's 6.6 percentage points per quality level.

[lapis leaf]

yeah i calculated that, but it just seemed off and i thought there was a more elegant formula to have a more normal number

[obsidian crescent]

17 isn't a normal number, so I don't know why you would expect the mechanic to yield a "normal number".

[iron root]

17% corresponds roughly to 1/6

and it might even be a written approximation of an actual in game chance of 1/6

so like, if modeled as a rational equation, it could be modeled as the divisor increasing by 0.4 per quality level (1/1 to 1/3, and then multiplied by the 1/2 for BMD)

[obsidian crescent]

It's 1/3rd of 50 though. So if we think of it as a multiplier, then each level gives 1.0-Q*(1/15) (with Q5 being a double). That suggests that regular mining drills have the following depletion scaling: 100%, 93.3%, 86.7%, 80%, and 66% (not that you'd really bother with Q5 mining drills, unless it's vanilla).

[iron root]

...No?
The order would go 1/1, 1/1.4, 1/1.8, 1/2.2, 1/3 (100%, 71.4%, 55.5%, 45.45%, 33.33%)

there's no addition/subtraction, only multiplication

and that would multiply with the BMD 1/2

[obsidian crescent]

Um, why not? Other bonuses add (wire reach, armor grid increases, etc).

[iron root]

i mean
technically you're adding something to a divisor

there's something to be said that all the other quality bonuses do produce a linear relationship with something that appears on the tooltip, and this relation wouldnt
but it does give sensible numbers mathematically

[obsidian crescent]

I was going to say that it'd be ridiculous for a Q3 regular mining drill to have almost the same depletion bonus as a Q1 BMD... but then I remember that this is basically true for a lot of upgrade items. A Q3 assembler 2 is almost as fast as a Q1 assembler 3 (for way less power). Q3 module 1s are usually close to as good as Q1 module 2s (and for efficiencies, they're better).

[iron root]

also if we decide to dig into higher (modded) qualities...
this would avoid singularities

[obsidian crescent]

The important question for depletion bonuses is this: how long do the resources last. At 50%, the patch lasts twice as long. 33% is 3x as long, and 25% is 4x as long.

Your upgrade progression is non-linear in percentages, but linear in patch duration extension: 1x, 1.4x, 1.8x, 2.2x, and 3.0x.

[iron root]

yep, that's what I'm expecting and is what is shaping my theory of its equation

[obsidian crescent]

Quality mining drills are super cheap though. I'm not sure they're going to let you get nearly a 2x ore patch extension with those.

Admittedly, running out of resources while off planet and without Spidertrons to expand isn't ideal, so being able to curtail this even before leaving Nauvis isn't a bad thing. And Vulcanus is where you get the BMD. And Gleba's patches are already super-rich. So really, the only place where this would really matter is Fulgora with its less rich larger island and rich smaller ones. And then, only if you didn't go to Vulcanus before or just afford the power cost.

[daring siren]

There's divisors and additions

Like how innate prod is +50% and other prod adds to that

What I can say for sure, is even if the base has innate reduction of 99% down to 1%, legendary won't bring it below 0%

Legendary regular drill would be 33% in my speculations

Also, remember that go 1.5-2.5, which is "weird"

[frank warren]

Don't we already have numbers for quality drills?

At least the legendary big one?

[iron root]

i think we only have the 17% number for legendary bmd

[frank warren]

Hmm, I see - and what to interpret from that is the debate?

[iron root]

more or less
is it the actual % number that goes down linearly, with a cap to ensure it doesnt go below some threshold
or does it go down with some nonlinear relationship that makes the patch life linear

[daring siren]

We know %50 on BMD and 17% on BMD

[grand igloo]

it also serves as a fascinatingly effective anticheat

because if someone cheats, right

they break determinism with the host and the other clients

and then they desync and get kicked out

[grand igloo]

haven't we calculated the prod scaling for everything

and it's like 1/2/3/4/6?

for each respective tier?

with modules following different patterns

I remember packs are that

[grand igloo]

if this is the case, then the formula is probably 1/(1/2/3/4/6) * base drill consumption?

granted this is all speculation

nah

[daring siren]

We don't have enough points to make it make sense

We know 0,1,2,3,5 is the multipliers, but we don't know how they apply to ore consumption reduction

It starts with 50% for BMD, and gets down to 17% i.e. 1/6

[grand igloo]

gotta be a different scale

hmm

[daring siren]

We need at least one more point to figure it out

Either in or BMD in e.g.

For some reason, I think the actual values are inverted i.e. normal is 0 or 100, while 50% is actually "100" or "200"

And then quality adds to this

[grand igloo]

maybe

did they specify it was additive or multiplicative w/ quality?

[daring siren]

Like "how many to mine before something is taken"

They have not

[grand igloo]

if this were the case, then I'd expect it to cap out at 14%, not 16%

hmm

[daring siren]

100 + 500

[grand igloo]

Unless there's an inherent halving of the bonus here

[daring siren]

Yeah

[grand igloo]

so it's 0/.5/1/1.5/2.5

add 1 to it, then multiply by base bonus...

but even that doesn't line up

[dense kraken]

I'm gonna bet on 50%, 43%, 37%, 30%, 17%

[grand igloo]

what formula for that are you using?

[dense kraken]

it's just a linear regression 🤷

50% is 15/30
17% is about 5/30

just subtract 15 with twice the quality bonus

[merry spindle]

• 5/30

[jaunty citrus]

Whatever the "actual" numbers are depends on how they implemented the consumption modifier.
I would have implemented it as something akin to "depletion efficiency", where 100% efficiency is 1 item from 1 ore, 200% is 2 from 1, etc.
So then the bonus would be as simple as 200% base efficiency on the BMD, then an additional 80% per quality level.

[dense kraken]

good catch

[hoary current]

1/6 * 50% doesn’t give u 17% tho

[wraith moss]

Mining drill efficiency doesn't have to be +150% anything. We know legendary tesla turrets only get +50% range, while the rocket turret is said to get "+10% for each quality level", which is probably also +50% at legendary, ignoring the double-bonus from going from Epic to Legendary.

Mining efficiency could be something like 100%/ 83.3%/ 66.7%/ 50%/ 33.3% ; with every number halved for big drills.

[obsidian crescent]

That's just a 1 - (Q/6) progression. That would give 8.3% for Q5 because it is a double-bonus; you can't just ignore that.

[half raptor]

Tell me if I'm off my rocker but I had an idea of how to deal with excess intermediates in a quality modules only base. I plan to have two pickup stations per modular base, one for non upgraded items and one for upgrades. Using filter splitters I can filter legendary to one chest, epic to another etc. As a side thing I'm also taking averages of the priority scores for demand stations with is a circuit condition of how full on a scale of 0-100 how empty the station is with 0 being full. I was going to use it for an alert system elsewhere but I realized I can use it for holding certain qualities back from loading into output trains. If my demand priority score for > 0 then let the inserter put from the chest to load and go deliver to the legendary station. This way it will not be in the trains trying to deliver if destinations are full.

Alternatively for inserters loading normal materials, a combinator would check if priority > 0 or if > 0 or any grades higher is in demand, then go ahead and load. If destination for normal is full and 5 sec elapsed it would then go to the recycle station.

The idea is that if the legendary, epic, rare, uncommon, AND normal stations are all full with a priority of 0, then there's no need to load up, even for recycling purposes since there's no need. It's self regulating and only recycles if there's demand for higher quality versions and will not load even for recycling unless there's demand for that grade or higher

It's a bit vague but I'd want to recycle any time my destination for delivery is full and to stop loading the material if I'm full on that quality and higher

[iron root]

Yeah this is sort of what I expect
But then I also got thinking…
What do you want to be linear with quality? Patch life, or how often the player has to build new patches?

Cause if the % to deplete is linear with quality, that’d make the rate of mining expansion linear as well

[jaunty citrus]

As with other turrets, higher quality tiers give the turret more range. The base range is 30, and the range at legendary is 45. The Tesla turret is special in that higher quality also increases the chance for the secondary chain lightning to fork and hit additional targets. The base fork chance per jump is 5% and increases to 30% if legendary.
The forking chance doubling at just seems really good

Making me think if you're going to use the tesla turrets you'll want to use quality ones

[open halo]

It's actually 5, 6, 7, 8, 30

[grand igloo]

nah it'll be 5/10/15/20/30

[kindred crater]

braxbro i think xe knew that

[kindred crater]

re-reading the recent fusion power FFF, is it reasonable to assume quality reactors are more efficient for fuel, while quality generators are more efficient for coolant?

[obsidian crescent]

Reactors are more fuel efficient, but generators just increase the power generation cap. They consume coolant based on the heat and how much power they generate. So a higher power generation capacity means using more coolant. They're purely space compression; they don't offer a productivity bonus on top of the reactor.

[kindred crater]

I see. so coolant consumption is directly coupled to reactor power output with no way of changing that if thats the case. The last line of the FFF is what triggered me to ask: In many cases if you maximise the efficiency of the fusion system then you end up with way more power than you need, and that's part of the fun. so thats referring to productivity in the processes to create the fuel and coolant itself, as well as higher quality reactors.

[half raptor]

assuming it's unlocked on the final planet, by that time quality should be relatively easy to achieve and we can reliably get some quality reactors and when legendary unlocks, upgrade to that

[kindred crater]

thats why I'm thinking first to start making higher quality quality modules ASAP so I can hit the ground running with quality once its the right time

[obsidian crescent]

By that point, getting quality raw materials for most intermediates will be pretty easy or cheap. But for the fusion reactor/generators, I assume those things are going to require a ton of the planetary intermediates from every planet. And making those in quality is going to be harder. So unless you find a way to overcome this hurdle, your only choice will be to ship massive volumes of the things and quality cycle reactors and generators on Aquilo itself.

[kindred crater]

theres a decent likelihood each new planet's main crafting resource will have productivity research (carbon fiber, tungsten, holmium)

[obsidian crescent]

True, but unless you choose to delay your Aquilo visit for a while, how many levels of that do you expect to get? Maybe 2-3 each? There's going to be a lot of competition when it comes to infinite research, and it's going to be way easier to research the Nauvis ones.

[kindred crater]

yeah. these wouldn't be researches to go deep on, but a little bit would go a long way especially if you're shooting for epic+ fusion items

I'm looking forward to being overwhelmed by infinite research choices

[cold wadi]

Now I want a quality of life mod, which will upgrade buildings to higher quality automatically too

[daring siren]

There was that idea of progressive upgrade planner. Where you can set a list of compatible items, and it'll try to get the highest available item of that list for your buildings, automatically upgrading when something new is available

[chrome mauve]

that would be very nice one, I didnt need to wait for items to be available in network, just set and forget

[kindred crater]

this is an interesting twist when quality cycling to get better biochambers lol

[final nimbus]

Apparently you can automate egg collection in some manner.
Raises the question of if there are other recipes that require alien eggs.

Actually, i'm reminded of the space platform we saw with turrets in odd places.
And the fact that other planets have recipes that take Gleba spoilables.
Not that pentapods are at all likely to survive in space, but it's an interesting thought.

[obsidian crescent]

Not egg "collection"; my understanding from the text is that you'll be able to create eggs from other eggs with a Biochamber recipe.

But that means if all your eggs "spoil", then you have to either recycle some Biochambers or find more.

[rotund egret]

since there likely has to be an expansion mechanic to pentapods, the solution might just be automatic farming with turrets

I'm curious what would be used for collecting the eggs. robots are an easy way to do it but the pick-up task can't be automated... or could it? 🤔

[obsidian crescent]

Inserters. You pull eggs out of the Biochamber from the Biochamber recipe that turns eggs into more eggs. That's almost certainly going to be the process. I highly doubt the devs are going to make you build some kind of pen for natural egg production.

[rotund egret]

the issue with this is that it has to be manually restarted if it fails

eh, I guess you could do it with Spidertrons killing nests and picking up eggs

[obsidian crescent]

Sure. And there are two options: go get some more eggs, or recycle some Biochambers to get some more eggs.

[rotund egret]

yeah. what I'm saying is that pentapod farming might be a way to get the eggs without venturing out into the wilds

with emphasis on the "might"

[tall sandal]

true chads coat their egg processing facilities in turrets and just let everything spoil

i love the idea that things can spoil into monsters (or maybe even buildings?) so cool

[frank warren]

It's going to be a fun circuit challenge to make it work.

[obsidian crescent]

One interesting point is that the Biochamber is a non-spoilable. So when you recycle it, you get completely fresh eggs back. So at least you have the maximum possible time to use them.

[stray marsh]

Hm, is there on_item_spoiled event?

[tall sandal]

i wish i knew!

[obsidian crescent]

Everything entity has a unique name, enemies included, right? I assume you would just set the spoil target to be an enemy, and the game would detect this and respond accordingly.

[stray marsh]

I think they'll sorta spawn capsules

[rotund egret]

yes, but the pentapods expand their bases somehow

[stray marsh]

Noone said that

[obsidian crescent]

If they don't, then managing them is going to be really easy: just push their nests out of your cloud.

[chrome mauve]

entities cant use spoil mechanic, only items

[young bison]

Unlikely, but there's probably an item spoiling trigger (which is how I suspect they made the eggs spoil into spawning wrigglers)

[half raptor]

Everyone having a Monday?

[daring siren]

We will on Oct 21st 😉

[lyric yacht]

is oct 21st a monday

huh, it is

[half raptor]

Yep, makes sense to release at the start of the work week. I'm sure wube is going to spend the week fixing bugs

[pure spire]

gotta fix that nauvis to aquillo speedrun skip

[half raptor]

I'm doing the anti speed run option. Every time I unlock something, I'm not researching the next thing until I've automated the last thing. Exceptions will be stuff like bullet damage, research speed, prod bonuses

I'm taking my time

[half raptor]

Machine unlocks

There's so many times I won't even make all the power armor stuff until I've built my rockets because I'm just chasing the science

[hoary current]

trigger techs probably do the thing for u already

where it stops u from researching too much before u get handle of certain things

[wheat sluice]

Do we have an Idea of whether a quality module inserted mid craft would still yield quality products? My though for a tedious exploit would be to swap prods for qual, right before the craft finishes.

Leaving one in the machine if it has no other sources of prod

[obsidian crescent]

From what's been said, it appears that at a low level, the quality of a recipe cycle's outputs is determined when the cycle starts. If there is a quality roll, that's when it happens. Inserting modules after this point won't help.

Now, you could do the opposite by removing the qual modules right after the craft starts. But there may be things that prevent that too.

[half raptor]

In 1.1.110 they made it so the prod bonus bar disappears if you force remove ingredients and I think replacing your modules will do the same. You might as well keep your or your for the whole craft and don't try to exploit the system

[kindred crater]

i will exploit the system

you could even put your assembler on a separate electric grid powered by a single with a bunch of constantly running to starve the assembler of power so it runs slower so you can modify the modules at more precise times during the craft

[obsidian crescent]

OK: how would that exploit anything? You put prods in, you wait for the bar to mostly fill up, then you change the recipe... and the prod bar gets reset.

You'd have to get over 100% prod to be able to exploit it without halting and restarting the recipe.

[kindred crater]

remove qual modules right after the craft starts

[half raptor]

so you lose the chance to quality as well as lose the small gap of prod bonus you could have had if you had hte prod modules in from the start

the quality chance rolls at the end of the craft, not the beginning

[obsidian crescent]

I was fairly sure the roll happened at the beginning. That allows any productivity products that happen along the way to get the same quality as the output.

[half raptor]

I have zero proof, but I seem to remember someone else had the same idea and devs said "no, that's not how that works"

if i'm wrong, expect that particular exploit to be patched out day 1 once someone posts it on Reddit

[jaunty citrus]

The roll is determined when the craft starts, all prod outputs afterwards use that roll. I wouldn't be surprised if removing all the quality modules mid craft stopped this, but even then you could just leave one in and stick prods in the rest

Hmm, this could be a way of using speed beacons with quality modules, turn off the beacons until the craft starts

[half raptor]

Well if that's true, expect that to get patched out as a bug

[frank warren]

'turn off' beacons is not so easy

they only do electricity checks every so-many ticks

[wheat sluice]

maybe they get circuit connections now? as many things do

[frank warren]

the only reason to allow circuit connections for beacons are exploits

[jaunty citrus]

You can do a batching process to ensure all the beacons turn off before craft. It would definitely still save time on recipes with long craft times like modules.

[frank warren]

Are you expecting to run quality on modules specifically? Wouldn’t it be much better to do their intermediates?

[jaunty citrus]

I'm expecting to do both so I can better understand each process

[obsidian crescent]

This kind of exploit doesn't work if you use it on fast recipes. It needs to be slow enough that you can turn on and off the beacons while running.

[jaunty citrus]

But I have a feeling that recycling T3 modules will be far easier than recycling circuits due to throughput requirements

Processing units are a 10s recipe, while T3 modules are a 60 second recipe, but need 30 processing units. So you'd need ~5x the quality cycling machines to match the same throughput by recycling the intermediaries.
This doesn't count productivity bonuses lowering the multiplier or any other intermediaries increasing the multiplier so I'm just assuming it's roughly 5x.

The other side of the coin: recycling the intermediaries gives you some base set of quality ingredients that can be used to make other quality items. So it may end up being easier in the long run

The only thing I know for sure is that in all the screenshots we've seen of quality setups from the devs none of them are recycling intermediaries.

[frank warren]

The devs have shown us very little of substance overall anyhow.

Most builds are to show off/test certain mechanics, not be optimal or metagamed in any way

[obsidian crescent]

True, but I also don't think we've seen trains with more than one cargo wagon either. That's not going to stop me from using 1-4 trains or even 2-8 trains if it is prudent.

[jaunty citrus]

Nah we've seen 1-4-1 trains in the reveal screenshots on steam

https://shared.akamai.steamstatic.com/store_item_assets/steam/apps/645390/ss_6b17be1ec4fa50721a6688ded0329ee84006faa9.1920x1080.jpg?t=1720177231

[lyric yacht]

well its not like you can productivity them so its better to do quality on them than to not

[obsidian crescent]

Also, module 3s are likely to take a special, planet-specific resource. Because this resource isn't used in nearly as many recipes as others, making quality versions of them are going to be somewhat difficult. So quality cycling the module 3s themselves makes more sense.

[half raptor]

I do agree that each module 3 will probably take a special resource, and with 4 types of modules and 4 outer planets I imagine speed for Gleba, quality for Fulgora, productivity for Vulcanus, and Efficiency for Aquilo

[obsidian crescent]

I do want to see each planet get a productivity research for one of their planet-specific intermediates though. It'd be a bit strange if the only prod researches are for the generic intermediates.

[half raptor]

I wish there was researches on every intermediate but I think Klonan said that would be too OP and they weren't doing that

Still 7 weeks 6 days out so anything is subject to change

[obsidian crescent]

I think there's something to be said for having a well-chosen corpus of researches compared to a shot-gun approach.

That being said, it might be interesting if there were a small number of productivity researches, but you choose which item from that group to make more productive. But each step is still exponentially harder. So you might have to choose between plastic and steel or gears and green circuits. Or blue circuits and LDS. There could be more choices.

[half raptor]

The problem with choice is that it'll only take 2 hours for the community to have mathed out the optimal choice

Big reason why some game developers only give choices for ways to play rather than how the game plays.

I'm just glad that this game has many different ways of solving the same problems and as long as fits the definition of "good" then it works

[kindred crater]

It doesn't seem like it now because of its known uses, but I wonder if calcite will have any kind of use

[obsidian crescent]

Well, it could be used to create quality stone from lava processing. But I don't think there are any other calcite-consuming processes that produce a non-fluid.

[kindred crater]

hmm. concrete has uses outside of flooring that require pretty large amounts of it (silos and reactors, and to a lesser degree centrifuges). you could always just recycle loop stone bricks to use as a component for that concrete, but if you're producing a lot of the stuff I listed, looks like quality modules in miners could be a nice boost to increasing the output of them.

shipping it will be costly I'm sure, but might be justified since its being used for permanent buildings, and not being consumed endlessly

[obsidian crescent]

Concrete is rather more widely used in SA. The Foundry recipe even uses refined concrete (I think the EMP does too), and I'm fairly sure the recycler uses regular concrete. I don't think the Biochamber will though (already requiring two seeds and eggs, they'll probably make the rest of the inputs fairly simple to get).

[half raptor]

we won't know til we know all the possible uses of calcite, but considering the only known uses so far is making fluids, I doubt quality calcite will be useful

[obsidian crescent]

Given how much stone lava foundries on Vulcanus will be spitting out, there's a good chance that you can make quality bricks for concrete just by cycling the stone directly as a way of getting rid of it.

[kindred crater]

at least with what I know so far, quality modules in calcite miners will depend on how much quality stone bricks I'll produce vs my demand for stuff I'm building with it

if I'm producing vast amounts of stone just from my planned production on vulcanus, that might be all I need without going any further

[half raptor]

The longer I've been theorycrafting my endgame quality base, the more I realize that 1) It's so much easier to just do productivity instead and 2) I'm doing quality based production on each step for the meme and solving the inevitable problems as I go will be the fun part.

[kindred crater]

I'm not going to act end product recycling loops won't be a part of my strategy- they certainly will. Perhaps stone on Vulcanus isn't the best example because its a relatively easy item to source elsewhere for this point. But if theres an opportunity to source quality intermediates from other planets that are inexpensive, I'll be fitting that into my quality strategy. Or heck, even processes where you can exploit the recycled products from items with productivity research. Like you said, figuring all of this out and making some autonomous system to produce all of this if one of my primary goals. I'm not trying to just beat SA and be done with it. I want to go as big as my computer can handle, and enjoy the journey along the way. 👍

[jaunty citrus]

It's important to note that recycling an item into itself is an order of magnitude worse than crafting it into something and recycling that. To get quality stone bricks you're far better off recycling walls than just recycling the bricks/stone.

Assuming all legendary modules - recycling something into itself has a ratio of 2727 : 1
crafting it into something using 4 quality modules (assuming it can't be prodded) has a ratio of 154 : 1

So you'd end up with almost 20x less by removing the crafting step

[obsidian crescent]

So maybe craft and recycle stone furnaces using the stone from your Foundries.

Also, does that ratio take into account that you don't actually want the item being crafted, so they all get recycled even if they're Q5?

[jaunty citrus]

No, it does not. But the efficiency difference would be the same

[obsidian crescent]

Looking at the possibilities, landfill would be a better quality cycling product.

[jaunty citrus]

Definitely, it would be the highest throughput recipe for sure

[cold wadi]

I just settled for 1:1 prod:quality in all my buildings with empty slots occupied by speed/productivity

[hoary current]

how about assembler mk3

[jaunty citrus]

Not sure what you mean

[hoary current]

does it has higher throughput than landfill

it takes two modules in 0.5s

[jaunty citrus]

but it doesn't take stone

so you couldn't use it to get quality stone

[hoary current]

oh I didn’t realize the context was for just stone

[vernal bay]

Is it more efficient to do multiple production steps with quality modules instead of recycling one step over and over?
So instead of building an assembler 1 with quality modules and recycling any that are bellow the quality you want you'd put quality in very step from ore, plates, gears, copper wires, circuits, sort after every step to use the highest possible quality recepie and then only recycle what's left over at the end.

[obsidian crescent]

That is potentially more resource efficient. But it's way more logistically complex. Beyond that, there's the fact that you're still losing resources through opportunity costs. Every intermediate you make along the way using quality modules is an intermediate that could have used prod modules. And thus, you could have had more for the same resources.

[vernal bay]

Except that that breaks all the ratios

[vernal bay]

Anyone done the actual math for that?

[kindred crater]

Not Alfonse, but yeah it has based on what was known at the time. A lot of convos took place 4-5+ months ago breaking down the numbers. Its also worth nothing there still may be major buildings/mechanics that can affect things. Such as this statement from #417: There is one more building typically combined with foundry and electromagnetic plants, which will be covered later.

[hoary current]

one thought, assuming that u want to get some quality , ignoring the possibility that u can get quality by recycling other things (such as ), is it more efficient to get quality instead of quality cycling , given the latter is a much slower recipe?

[solar osprey]

The issue is that you’d end up with a quantity mismatch between quality circuits, wire and plastic.

As circuits, from plate, have 2 crafting steps, wire has 1, and plastic has 1.

[hoary current]

only apply quality at the specified recipe, not to the basic productions

[solar osprey]

If you’re qualitying miners and furnaces, that goes to 4, 3, 2.

[solar osprey]

Maybe? But you’d probably have to run 3 quality, 1 productivity on your circuits and wires?

Or you’d have a mismatch again.

[hoary current]

just have the same modules setup for all the ingredients

or does it matter?

[solar osprey]

Depends on whether you want to backlog on quality ingredients.

[hoary current]

what I mean is u quality cycle the intermediates instead of the final product

if the final product is a much slower recipe

[solar osprey]

You cant quality cycle plastic though.

Efficiently.

[hoary current]

u can, in a wasteful manner

[kindred crater]

doesn't plastic have prod research?

[hoary current]

but I’m only comparing the local infrastructure cost here

[solar osprey]

Shrugs.

It’s also just made from coal and oil. So near useless and infinite.

[hoary current]

on u probably have plastic made from resources that never runs out

[obsidian crescent]

True, but it's not useful in the same way that blue circuit prod research can make quality blue circuits free.

[daring siren]

It's not like it grows on trees you know

[hoary current]

well my current speculation is that green cube is made from wiggler eggs

and u need green cube to make plastic

[kindred crater]

definitely, its a much more expensive item. this research would likely be most beneficial on vulcanus if you're manufacturing plastic there, since coal is used more heavily

[burnt reef]

quality arcospheres

[fallen urchin]

They are turned into normal arcospheres because the player accidentally mixed them up with low quality ingredients

[lyric yacht]

That's a good point though, if you keep cycling the same 238 or something with kovarex is that an easy way to get legendary 238?

and also legendary 235 cuz most of it is recycled?

[solar osprey]

You can’t put quality in kovarex.

[lyric yacht]

really? Did they say?

[fallen urchin]

If catalysts are not affected by productivity, I think they can let them not affected by quality

Though this is difficult to code, so might just as well ban qual there

[kindred crater]

I believe devs have acknowledged this in the last few months and indeed prevented quality from being used in it

[daring siren]

They have. After I mentioned it was tough to get working in Janky Quality

I still hope to see catalysts and non-catalysts split in outputs

[kindred crater]

say you are aiming to obtain legendary productivity module 3s, which of these is the superior option?

1. intermediates fed into EMPs to directly make prod 3, and then standard recycle loops with your prod 3's until you reach legendary
2. recycle loop until you reach legendary, and then use them for your legendary modules as well as recycling them as needed to obtain legendary / for the modules as well

wouldn't it be #2?

[obsidian crescent]

The problem with 2 is that prod 3s in SA likely require a planet-specific intermediate.

Which won't be able to be quality cycled in an EMP.

[kindred crater]

yeah thats at least a possibility based on what we know with fulgora

[fallen urchin]

What will quality nightvision goggles do?

[solar osprey]

Give you quality vision of course.

But unless they have a power use modifier unlike every other piece of quality equipment, I guess quality night vision will be about as useful as quality belts. Or chests.

[fallen urchin]

What is quality vision? Less green color distortion?

[solar osprey]

Makes it so you can’t see anything of lower quality than the the night vision goggles.

It’s too inferior to be visible to you.

[fallen urchin]

That's a big problem because I certainly won't be using quality belts and pipes

Not sure about power poles, I think they are not worth their cost until post endgame

[solar osprey]

I can see myself making quality large poles to take to fulgora. To see if I can bridge a gap between islands.

[fallen urchin]

Makes sense

[burnt reef]

what does more HP do for personal equipment

[solar osprey]

Nothing.

But then again, what does more HP do for belts and chests?

[burnt reef]

biters take longer to eat it

[solar osprey]

By the time you get to quality, especially belts and chests, if biters are chewing your base, you’ve already failed.

[fading fern]

What about pentapods ? They could sneak a hit here and there.

[lyric yacht]

that's what bots with repair packs are for

[fallen urchin]

Tank a few hits from accidental car collision?

This is not going to heppen after getting spiders, though

[solar osprey]

Actually, on this point, who remembers when armour, including modular and power armour had durability?

[burnt reef]

I

[daring siren]

In Janky Quality, having better NVG made it closer to daylight.

[solar osprey]

How upsetting would it be for your legendary power armour mk2 with legendary modules to just break and lose it all?

[daring siren]

In solo game? Not at all, up to 10 minutes of progress 😆

[lyric yacht]

And when the steel axe was an item

[solar osprey]

Remember when you had to craft iron axes?

[lyric yacht]

yes

I remember when concrete was added

[full hedge]

[grand igloo]

did they prevent qual from being used in it, or is it impractical because Kovarex enrichment is 40 235 + 5 238 -> 41 235 + 2 238

[daring siren]

It's impractical given a single output slot.

[grand igloo]

sure qual looping the 235 is easy enough, but you need to consistently have 5 Q5 238 to perform a Q5 process.

[daring siren]

Which I personally think should change.

[grand igloo]

wait so it's just recipes that output multiple products can't be qual'd?

[daring siren]

You can get quality 238 from reprocessing

No, it's specifically kovarex that can't be

[grand igloo]

That seems very odd

because getting a consistent stream of high-quality 238 to make high-quality 235 would be difficult

[daring siren]

Because catalysts are not handled in a way conducive to this

[grand igloo]

certainly far more complicated

so that means Kovarex always spits out Q1?

[jaunty citrus]

It always spits out the same quality as the inputs

[daring siren]

No. It means you can put Quality modules in it

[jaunty citrus]

You just can't use quality modules

[grand igloo]

...then how do you get quality 235...

not that I'm sure it's useful but

[jaunty citrus]

By crafting anything else out of 235 and recycling it, using quality modules in miners and uranium processing, starting with quality U235 and U238 for kovarex, etc.

[grand igloo]

is nuke fuel prod-enabled?

[daring siren]

Yes

[jaunty citrus]

Yes, I don't know if it's recyclable or not though

I would assume so

[grand igloo]

should be recyclable

it's an assembler recipe iirc?

oh no, centrifuge

maybe not then

[jaunty citrus]

No it's an assembler recipe

So I would expect it to be recyclable

[grand igloo]

wiki says it's a centrifuge recipe funny enough

[jaunty citrus]

Ah, wrong nuclear fuel

I'm thinking of nuclear fuel cells

Which would be better because you could use more modules

[fallen urchin]

Nuclear fuel is 90s, which is too slow for cycling

I'd want something faster

Either fuel cells or nukes are ok

[jaunty citrus]

Fuel cells are what you want

[grand igloo]

it's a shame, qual cycling kovarex via fuel cells would be so fun

[jaunty citrus]

It has very high throughput of U235 and U238 which you can then use for quality kovarex enrichment

For

[fallen urchin]

Also, if you're trying to use kovarex to turn quality 238 into quality 235, then you need 40 more U235 for each quality, not just qual 1

[grand igloo]

yep

[jaunty citrus]

You would just keep the highest quality

no need to keep 40 of each one

[grand igloo]

yeah you do

[jaunty citrus]

oh also normal

normal + legendary would be all you need for kovarex

[fallen urchin]

If you use 40legendary 235 with uncommon U238, then you'd get 41 uncommon U235 out of this

[jaunty citrus]

Why would you do that?

[fallen urchin]

Yeah it's avoidable

[jaunty citrus]

Use the legendary 238 you're getting from quality cycling fuel cells

[fallen urchin]

The medium quality U238 are just untouched and thrown back into the loop directly

No need to kovarex them

[jaunty citrus]

So you'd just build up a stockpile of U238?

I don't see the point

You only need kovarex for normal quality and the highest quality

[grand igloo]

the point, my friend

is

[jaunty citrus]

You don't use U238 for

You need the kovarex to turn it into U235 for

[grand igloo]

you use the 238 to make the 235 for nuke

[jaunty citrus]

Exactly

[grand igloo]

so you need leggy 238

[fallen urchin]

So the medium quality U238 doesn't need to be kovarexed, and U238 goes into the kovarex

[jaunty citrus]

Yeah, that was my point

Exactly

[jaunty citrus]

^

[fallen urchin]

There are 4 uses for quality uranium: Fuel cell; Train fuel; Rounds (use U238); Nukes; And they could potentially be involved in modules, so it could be 5 instead of 4

Fuel cell for fission plants certainly won't use quality

[jaunty citrus]

I expect to be using it for train fuel

[fallen urchin]

Yes, and if it doesn't show up in modules, then train fuel will probably be its only use

Legendary nukes are probably for memes only

[jaunty citrus]

more fuel per fuel isn't useful when you consume so much of it to make the quality varient. A nuke is a nuke, it doesn't really need to be better. (having a few would be fun though) and I don't expect to be using the machine gun for long

The train fuel getting bonuses to speed/acceleration for quality would be very funny to watch

Trains are already ridiculous as is

[fallen urchin]

Yes, unless they get a superior tier like the fission plant

[fallen urchin]

But they don't get anything from quality

So better fuel is the only thing they get

[jaunty citrus]

I get why they don't get quality, replacing trains would be a chore

[grand igloo]

what does fuel get from qual again?

[fallen urchin]

Vehicle accleration and probably max speed

Fuel value is probably untouched

[grand igloo]

i mean science packs get extra longevity

even if it's impractical

[jaunty citrus]

Science packs get a pretty significant bonus but it doesn't compare to productivity

[fallen urchin]

Science is where quantity beat quality

[grand igloo]

yeah too many steps

but fuel might get a fuel bonus w/ quality

[jaunty citrus]

It's not a matter of steps, prod will always be better than quality for science

[fallen urchin]

Green science could make use of quality, but the logistic will be too complex and probably not good for UPS either

Basically the unproddable steps

[jaunty citrus]

you get x2 from prod (+100% prod from 4x ) but less than 1.3x from quality (+100% per quality level, 24.8% chance to upgrade with 4x )

Using quality modules also forfeits the use of speed beacons, so it's necessarily lower throughput

[grand igloo]

yep

though I’ve seen speculation that qual might be worthwhile on the spoilable science(s)

[fallen urchin]

So only quality in the factory, not the products

[grand igloo]

since quality iirc also affects spoilage?

[fallen urchin]

I think it still won't be worthwhile

[jaunty citrus]

I expect quality on science ingredients to be useful early on, getting some , flying robot frames, etc would be quite nice

[grand igloo]

prolly depends on where you have your labs

[fallen urchin]

However if has something like green sci (2 parallel steps that both reject prod), then it could use qual

[jaunty citrus]

But later just mass producing them and cycling would be more compact and higher throughput when you have effectively unlimited resources

[fallen urchin]

And high productivity/qual modules

[grand igloo]

if prod means 10 packs that spoil in 10 minutes, while qual means 6 packs worth that spoils in an hour…

if your packs have to travel 5 minutes, prod packs lose half their value while qual packs only lose a twelfth

[fallen urchin]

Probably still not enough though

And I believe Gleba packs won't spoil that fast, because the devs said they want 3 planets to be equally viable for first visit

[grand igloo]

in that example qual is 5.5 packs and prod is 5

[fallen urchin]

1 Q5 pack could be even more expensive than 10 Q1 packs

Without prod research and maxed prod/qual modules, the price is about 13x(?)

[jaunty citrus]

There could easily be a situation where it's better to stick quality modules in gleba packs, but the issue is you control most of those situations variables.

You could build a faster processing chain, make a faster platform, or send all the other packs to gleba instead

[grand igloo]

oh yeah I don’t doubt that gleba will be optimal lab location

so you can prod things where possible

[wheat sluice]

wait, science yield scales with spoilage?

[solar osprey]

Gleba packs do.

[wheat sluice]

yeah, found it:

On the other hand, the new Agricultural science packs do spoil which reduces their value for research, so you will be incentivized to try to bring home the freshest science packs you can.

had missed this fact

even more reasons to do quality, but cycling spoilabeles will probably be a bad idea

[solar osprey]

Okay, biochamber can take modules.

I would say the one risk with qualitying biological items is that there a chance for your quality items to spoil while waiting for other quality items to arrive to be crafted with.

So the quality biologicals would be more spoilt and less valuable than the regular biologicals.

[vernal bay]

One thing I'm pretty sure of is that the first thing you want in higher quality is quality modules

[solar osprey]

Yup.

The first 4 Q1 modules I’ll make will go into the Q1 module assembler.

Possibly even hand fed, so that I can put the modules in as they’re being made.

[wheat sluice]

yep, quality is a ball that you have to get rolling

[obsidian crescent]

Two things:

1. Quality goods have longer spoil durations. So if you do have to wait for more quality goods, it's not as big of a deal.
2. The final goal in most processes is not to generate a spoilable. Eggs are used in Biochambers, but Biochambers aren't themselves spoilable. So their freshness doesn't matter; the only thing that matters is whether they're still eggs.

That being said, you probably wouldn't bother with making recycling most plant material. It'd probably be better to just quality cycle the non-spoilable end-product, like Biochambers or plastic. Plastic seems pretty inexpensive, so generating a ton of it and throwing most of it away isn't a huge deal.

[willow rover]

is it confirmed that quality goods have longer spoil times?

[obsidian crescent]

Yes. We see the quality diamond next to the spoil time of nutrients in a recent FFF, and this was mentioned by devs at several prior points.

[daring siren]

Which won't help for things other than science unless you have the rest of the items of the same quality

[grand igloo]

not if farming HQ crops is an option

[wheat sluice]

Ohhh, high qual seed, didn't think of those at all, if they worked as expected, then you'd just have to get them first and that's it, they don't even spoil

[obsidian crescent]

Ag towers can't use modules, so that's unlikely.

[wheat sluice]

assemblers can, it's just a matter of whether a quality seed will plant a quality tree and if it will then produce quality fruits

[solar osprey]

Maybe once you get legendary quality 3 modules, it’d work, but unless quality massively boosts the spoiling time, I think the time spent waiting for quality ingredients could result in more spoilt products.

[obsidian crescent]

It depends on what you're making. I mentioned quality cycling Biochambers. It uses a spoilable, but when you recycle a Biochamber, the eggs you get are at full freshness. I could imagine that the spoil time for quality item is a 100% bonus per level, so even a Q2 egg lasts for twice as long as a Q1 egg.

And eggs need to have a decently long spoil time anyway, since early on, you may have to run halfway across the map to scrape together the seeds needed to make a Biochamber with them.

[solar osprey]

If you can get a quality % and throughput to guarantee that another quality ingredient will arrive in good time, I can see it working.

But if you’re looking at low quality % and moderate throughput, I think it could result in the quality ingredients waiting longer.

[obsidian crescent]

As long as they don't completely spoil, it doesn't matter.

[solar osprey]

Well, no, as the result inherits the spoilage.

[obsidian crescent]

Only if it is itself spoilable.

If you're quality cycling a non-spoilable, then freshness of the ingredients doesn't matter.

[fallen urchin]

So the whole farming process is leg in leg out?

I think this is highly improbable

[vernal bay]

Will it even be worth it to use qulity outside of a few select things (like for spaceships or whatever) until the megabase stage?
In a megabase everything that makes sense will be legendary for UPS, but given the crazy complextiy involved I doubt it would really be worth it to ever have your whole factory be uncommon, rare or epic

[fallen urchin]

Some recipes reject prod, so you might just as well use qual on them

[obsidian crescent]

What does "worth it" mean? And given that SA is a bigger game than vanilla, the line between "normal progression" and "megabase stage" will be a lot more blurry.

It costs basically nothing to put quality modules into your assembler 2 and 3 makers. Since you make so many assemblers, then you just get a decent supply of higher-quality ones. Similarly, the ingredients of purple science can't be prodded, so you may as well use quality modules in them and siphon off the high quality furnaces and prod 1s for use in the base.

Also, the advantages of a small number of high quality beacons are sufficiently significant that I would say that finding a way to get quality intermediates early (even just a few) is "worth it".

On a different quality topic, there's an interesting point about Gleba. Apparently, you make carbon by burning spoilage in a furnace (not as fuel, but as an ingredient). Given that, you can make quality carbon from quality spoilage, which you can get by putting quality modules in your nutrient makers; siphon off any quality nutrients and letting them spoil.

[grand igloo]

yeah, just a matter of if you can get quality seeds

[obsidian crescent]

I doubt quality matters for seeds. Yes, they can have quality, but I don't think that planting a quality seed grows quality fruit. If it did, it'd be way too easy to get quality fruit products.

[fallen urchin]

Exotic Industries mod has a recipe: 5 crystal fluid + 1 energy crystal -> 2 energy crystal

If you put in crystal, then you produce 1 each cycle

If you put in crystal, then you produce crystal for no additional cost

It will be ok as this only affects a few products, but for Gleba plant, this will be almost everything

[wheat sluice]

just introducing a recipe that requires something not grown could throw you off

[fallen urchin]

Like ore?

But I think Gleba will have a large amount of recipes not using them

[wheat sluice]

sure, up untill that point you could easily do qual, ensuring freshness, but past it you'd have to decide whether you ditch qual or do way more of it

idk, anything, a plate, or a piece of coal

Has anyone done the math on exploiting this? #space-age message

[obsidian crescent]

We don't even know if that's still possible.

[wheat sluice]

I quickly threw together a special case for non-proddable recipes in EMP and Foundry, just wanted to check my numbers.

And how would you prevent it?

[obsidian crescent]

Removing on buildings. Which they apparently did.

Changing the input quality on a machine can only be done manually or via circuit networks. Both of which probably reset the prod bar.

[wheat sluice]

so you mean that a quality has to be set, to use quality ingredients?

or if it's unset, the it will treat all ingredients as normal?

[obsidian crescent]

Yes, and the machine will only take ingredients of that set quality.

The default is base quality.

[fallen urchin]

Higher quality ingredient can be used, but it will be treated as equal to the set quality

[wheat sluice]

I don't like this solution

[obsidian crescent]

It prevents the exploit, and it makes the system a bit simpler for users. They can't accidentally lose high quality materials by combining them with low quality inputs.

[wheat sluice]

besides, on EMP the cost multiplier went from 30 to 20

[obsidian crescent]

The only real downside is that you can't get rid of "excess" high quality stuff by deliberately combining them with low quality stuff.

[wheat sluice]

I would sometimes like to mix qualities just to void them

has this been discussed somewhere by devs?

[obsidian crescent]

No, but it was leaked by someone who was at the LAN party.

[wheat sluice]

credible information, thnx

[vernal bay]

wait does that mean that researching epic or legendary is now guaranteed to break everything build before unlocking them that doesn't account for epic/legendary?

[vernal bay]

but quality also breaks ratios, you can't just place higher quality assembler willy nilly you have to design your blueprint around a specific quality level (or spend lot of effort and making a staged BP) , and mixing different quality levels in one build just won't work.

[wheat sluice]

You definitely can filter for qualities not yet ulocked, just not directly I believe

[merry spindle]

TBH I don't see reason to make quality assemblers before you can make them an mass
Maybe to boost some bottlenecks but that's usually speed modules duty

[wheat sluice]

My stance on the this quality exploit is that from a balance perspective, it could remain as the gain is not that substantial and would require a relatively complex signal setup to work properly. One of the things that make it so much more tedious to implement is inconsistent productivity bonus, that is, in some cases when you should fill the prod bar to exactly 100% it will only go to smth like 99.999..% ofsetting your craft cycles, not only would you have to make it alternate the normal and higher qualities but also detect this occurrence, to correct for it.

But on the other hand, it doesn't make logical sense thay it should work this way.

[grand igloo]

Does the planting require other ingredients? If so, that might be the intended difficulty for getting qual in this way

[obsidian crescent]

The devs haven't said, and the only things on the belts we've ever seen feeding Ag towers are seed, so...

[grand igloo]

Could be then that the fruit products take some other thing that doesn't come from seed.

[obsidian crescent]

Again, we've seen setups doing mashing, and they only take fruit (or fruit + nutrients for the biochamber versions).

[fallen urchin]

This can be fixed by giving it 4% purple bar offset

Suppose you have 50% prod, normally you run a recipe, the prod bar goes 0%->50%, then run a recipe, the bar goes 50%->100%

If float point precision problem kicks in, it becomes 50%->99.999% and this breaks

With 4% offset, it will be 4%->54%->104%, which won't be affected by float imprecision anymore

[wheat sluice]

how could you introduce the offset for a recipe that does not take prod mods?

[terse lagoon]

Pretty sure boskid mentioned once that it's fixed when i was suggesting such a thing after quality FFF came out

Not sure in what way fixed tho

[wheat sluice]

I looked thoroughly, but didn't find anything on it in discord

at least till the start of the summer

[fallen urchin]

Oh that's not possible, it's only possible for those do take the module

[distant eagle]

Isn't "any quality" a thing? And then the assembler can use mixed quality ingredients?

[obsidian crescent]

It's been said that this was removed for a variety of reasons, including dealing with an exploit.

[distant eagle]

So the issue is building the productivity meter with low quality ingredients, in a machine set to any quality, then getting the bonus output with higher quality ingredients.
Why can a machine not accept quality above what it is set to, but still output based on the set quality?

With Janky Quality, even when matching quality steps in ingredients, trying to match ratios, everything deadlocks without delevelers.

[obsidian crescent]

Don't build quality cycler setups where a few unlucky rolls can lead to a deadlock.

[distant eagle]

That's not the issue. But the solution seems to be destroying the excess in a recycler instead of using it like it's lower quality

[chrome mauve]

They werent talking about lowering quality, it was about using it like it as lower quality

[gleaming quartz]

guys, hear me out. How much would quality break space expansion?

[fallen urchin]

There are many quality intermediates that can't be used directly

[grand igloo]

Really? It was removed? Source on that claim?

if so, then that’s a shame, because it makes quality a significantly less interesting mechanic

[chrome mauve]

example? of one

[daring siren]

Janky Quality has a deleveler for this job ^^;;;

[winter saffron]

Boskid didn't like it. #friday-facts message

[lyric steeple]

exploit was removed by requiring quality of ingredients to be selected when selecting recipe to craft. All ingredients must be of the same quality and when changing quality of recipe bonus progress is cleared.

aka "any quality recipe" option is gone

[winter saffron]

Forgot I summoned him. Do Splitters support the condition: Filter for AnyItem equal/at least/lower than quality Y

[chrome mauve]

wonder if it would be possible to change "any quality" recipe to have no prod bonus so there would be no exploit

[obsidian crescent]

That would be very difficult to communicate to the user what exactly is going on and why their modules don't work. And a lot of people would want to know why they can't prod it even though it's the same recipe.

[grand igloo]

this sounds like it makes machine config really annoying and flattens the mechanic. a shame

K2/Space Ex are already annoying enough w/ furnaces

[chrome mauve]

the hardest part on mys solution is for devs cause I dont know if it is possible to exclude just part of recipe (the one that have any quality and leave quality type) from that system and communicating why it doesnt work to player cant be easier with how factorio is made especially with tutorial it have ingame. Just put that info with pop up one player use first time any quality type

not sure where idea that it would be very hard to do comes from

[open halo]

yeah, it feels bad (though we haven't played it, so who knows). there might be a better way to do it, but I understand why it was done this way :<

[winter saffron]

Not terrible since Ctrl+ c exists.

You mayhaps have to engage in belt/train sanitation

Depending on your base setup

[weak vault]

good to know especially about the bonus progress, i'm sure we all remember the fresh bonus exploit that was used by Hendricks in 1.1

i like how some of us find ways to complain about a yet unreleased game

[grand igloo]

it was fun to theorycraft ;-;

[distant eagle]

why not ">=uncommon", which accepts anything other than common and outputs as if all inputs were uncommon. or is there another exploit?

[frank warren]

actually, there’s no exploit there… 🤔 can anyone think of one?

[open halo]

my guess: propping up production on common stuff, then putting in legendary

[frank warren]

doesn’t work here

This is a recipe that requires >= uncommon, but it acts as if all of the inputs are uncommon.

It doesn’t have the switching problem that any quality does.

[winter saffron]

Maybe performance reasons? Since each inserter would have to perform max 5 checks each tick

[merry spindle]

One >= check

[winter saffron]

Does that work with their quality implementation?

I thought it might be set as an Enum?

Didnt know intEnum exists. Neat

[iron root]

I would very much appreciate a way to have a recipe take in any quality and just always put out a baseline quality, even if all ingredients in that pass were better quality

It’d make sinking excess quality material into science much easier

So basically bring back but it always puts out instead of looking at the minimum of input qualities

[obsidian crescent]

I really feel like "excess quality material" is just never a problem you should have 😉

[iron root]

I suppose I just bite the bullet until recyclers

Because those are proper item voids

It probably won’t be a problem until then.
Probably.

[wheat sluice]

yeah, the only way to scale quality is trough recyclers, any excess before that would probly be easy to buffer

and you get them early on one of the easiest of planets

[iron root]

I plan on quality modding miners and using the ore to make a full blown quality mall

So I’ll get it going before fulgora

[obsidian crescent]

Early on, a good iron sink could be steam engines (potentially useful on Fulgora or Gleba, depending on how they work). For a copper sink, it'd just be circuits.

[wheat sluice]

if you rush fulgora, you'll get epic quality and T3 quality mods unlocked

[iron root]

I think I’m going to go vulcanus first just because the foundry will be so good to boost holmium output

And bmd to stretch out the convenient scrap

[obsidian crescent]

It's not quite as good as you think. The normal recipe is actually an assembler recipe (holmium plate is always made from holmium solution), so the Foundry would "only" have its 50% prod bonus.

[wheat sluice]

fulgora is easy to leave, so it's my first pick, plus quality takes time, so while im tinkering away elsewhere, it can try and print out some quality quality mods for me

[iron root]

If there winds up being a solid reverse relation (emp helps vulcanus science in some way) I’ll 100% go fulgora first

But 50% prod on a rare key ingredient is uhh
Kinda good

[obsidian crescent]

You need to make circuits on Vulcanus too, so the EMP is always helpful.

[iron root]

But just how important will they be
As of right now I get the impression I’m gonna need a lot of scrap mining for any reasonable pace of electromagnetic science, and I don’t see a similar gate for vulcanus

[wheat sluice]

and you can apply productivity to module manufacturing with an EMP

[iron root]

That’s the big thing I’m looking at for fulgora
But I can always do quick and dirty vulcanus -> fulgora and then go module crazy

[obsidian crescent]

That's kind of the question with smash-and-grab. The EMP itself is a trigger tech; you don't need to be making science packs to research it. But at the same time, will the EMP have the ability to make circuits and modules when you get it, or do you have to research that?

[wheat sluice]

recycler is a trigger tech too, no?

[obsidian crescent]

A very early one.

[iron root]

Of course
Assuming we can see what triggers each tech in the tree when we start the game
A lot of this is just speculation that might get upended the moment I open the 2.0 tech tree

[obsidian crescent]

So you could always drop down, trigger that, then call down a rocket silo and some basic infrastructure to leave. But if you want the EMP, you have to trigger more.

[wheat sluice]

recycler and EMP coulnd't be anything else but fulgora trigger tech, as you can't even progress without them there

[iron root]

For all we know agricultural science unlocks something hilarious enough that first becomes a serious choice

I’m 90% sure I’m going to finish a production line for the planets science pack before leaving though

[obsidian crescent]

One thing I was thinking of with Fulgora smash-and-grab was to just set up on a medium-sized island and basically mine it out (using higher-quality miners to prolong things). Even after getting 200 EMPs, just leave the factory running making more EMPs and go elsewhere. By the time you come back to complete Fulgora, you should have a good pile of resources on hand to work with.

[wheat sluice]

I'm sure they'll buff Gleba to make it more enticing, but I doubt that it'll be easy without imports from other planets, so it'll probably be best to leave it for last

[fallen urchin]

recipes could also interfere with quality added by mods

Suppose a mod adds a mechanism that allows assemblers to level up with crafting, and the level is implemented as an additional dimension in quality

So we can have Q5E5 assembler3 for example

The question is, which is higher, Q3E2 or Q2E3?

[jaunty citrus]

Depends on the multipliers

If they use the same values (30% speed per level) then it would be the same

It also depends on the implementation

[fallen urchin]

If they are same, then output should be which quality?

[jaunty citrus]

What? I'm not following

[fallen urchin]

Because output is based on highest input

[jaunty citrus]

There is only one "dimension" in quality, if a mod added a different type of quality it would not be capital Q Quality

[fallen urchin]

But Factorio allows mods to hae 255 qualities in total, and a single dimension of quality certainly doesn't need this many qualities

[jaunty citrus]

Yeah, that would be far too many

But they're using one byte to store the quality value, which gives you 255 different possible levels

[fallen urchin]

In my example, the quality are 2 dimensions, Quality, and Experience

[jaunty citrus]

Yeah so the output would be whatever quality the item is

[fallen urchin]

Experience is implemented using the quality system, but obtained in different ways

[jaunty citrus]

It wouldn't be implemented "using the quality system" because quality is one dimensional

[fallen urchin]

The crafting output should probably be decided by the mod

[jaunty citrus]

also isn't a think anymore

You have to choose the quality of inputs

[fallen urchin]

If a mod actually uses all the 255 qualities, how will the quality selection screen be?

[jaunty citrus]

A really really long drop down I guess

Maybe with a scroll bar

[fallen urchin]

I think this is actually a good cursed mod idea

Gonna post it there

[jaunty citrus]

I think a suggestion was made a while ago to make only quality level 255 do something useful

Anyway no 'serious' mod will use all 255 quality levels

5 is fine

I'm sure there will be mods that expand that to 10, 20, etc. but each would be less popular than the last

[daring siren]

Py devs want to use more qualities.... But I doubt it will come to fruition

[wheat sluice]

are you doubting the dedication of devs that created a 1000h long modpack?

[daring siren]

Melon is insane (in a good way)

He wrote a whole mod for Space Age before Space Age is even out.

[jaunty citrus]

I could see it being interesting if the quality chances were way higher but the bonuses were more incremental

[vernal bay]

Without , a mod with a quality level down machine is going to be essential without it you'd have to sort 5 ways after every single machine. Like if you design blueprints around you don't really want machines because they'd mess up ratios so you'd end up with a bunch of stuff you can't use effectually so it get's stuck in chests or you'll have to delete it by recycling. I mean is fine to use for specific cases but it dosn't really make sense to ever build you whole base out of , becouse they you end up with wasted (and higher once unlocked) stuff you can't use.

[small egret]

Ok, I have been reading parts of this and I want to see if I understand correctly. Are believing right now that they removed the ability to use quality items as if they had a lower quality (e.g. using a rare item in a uncommon recipe)? And it is this way because they needed to fix an exploit? Are we sure we this is not just a misunderstanding because I see no way Wube would ship quality in that state.

[daring siren]

Or you make enough that it doesn't really matter

[obsidian crescent]

Even with any-quality, you can't really "level down" everything. You can only "level down" something that you use as an ingredient in something else.

Let's say we're talking about medium poles. If you're making quality medium poles by putting quality modules in the medium pole assembler, if you get some Q3 medium poles, how would any-quality deal with them? Medium poles don't get used for anything else, so... what do you do?

[obsidian crescent]

Technically, "using a rare item in an uncommon recipe" was never an option. What they took out was using a rare item alongside an uncommon item to produce an uncommon output.

[daring siren]

There are other ways to fix quality unbalances e.g. using more or less quality modules...

[small egret]

Ok, thanks. That was exactly what I meant. For some reason I thought you used to be able set an assembler to only accept >= quality but I might be wrong. So we seem to be certain that that is how it works right now, but do we have any reason to believe it will stay like this for the SA launch? The wording "exploit was removed by requiring quality of ingredients to be selected" makes it sound like it was removed to quickly stop an exploit, but if that exploit is fixed in another way we should be able to have any quality be added back in.

[distant eagle]

doesnt need to be (which allows boosting the productivity bar with low quality and then getting the bonus item with high quality items) ">=" if it was added, would work as a sink for extra . you could mix and but the default output would be . currently you cant mix qualities, or lower an items quality

[obsidian crescent]

That would still allow the exploit to exist.

[distant eagle]

if the assembler allows >= , and you put in all (without quality modules) it outputs

[small egret]

That sounds good to me

[obsidian crescent]

That doesn't make any sense.

[distant eagle]

which is why you put in an assembler set to craft
but if your is missing because the is backed up at the assembler because the is blocked by
you can >= the and use the excess to fix the deadlock

[obsidian crescent]

Or you can recycle the Q4 plate to try to get Q5 plate.

[distant eagle]

and now youre missing Q4 plate for the Q4 green circuits

[obsidian crescent]

Are Q4 green circuits the goal for some reason?

[small egret]

But can you fix it without a recycler? I don’t like the idea of needing recyclers to do quality

[distant eagle]

Q4 green circuits go to the Q4 red circuit production, which has a chance of producing Q5 red circuits
but the Q5 green circuits produced by the Q4 green circuit assembler, go directly the the Q5 red circuit assembler

[half raptor]

This is why each quality will have their own stations and when destination is full and inactive for whatever timer, go to recycle station

That way if you wind up with a full buffer of it'll go on to try to make which can then feed any other demand for copper and it'll keep your line continuing to make more in the hope some will quality up to

I'm going to have my loading inserters only load if demand for the current quality is > 0 OR high quality demand > 0

If my legendary and epic stations are full, the epic cable can just sit in the chests and wait for a demand

[vernal bay]

Can't just work as a >= that treats outputs as . That's how I thouht worked anyways.

[half raptor]

Yes it can but it's a loss of quality. I'd rather my stations eventually go idle if I get too much quality of something and wait until I start getting quality of my other ingredients

[obsidian crescent]

Any quality meant that there was no quality filter on the assembler's inputs at all. Therefore, it's possible for inputs to have different quality and the minimum quality among inputs is used to determine the base quality of the output.

[half raptor]

Having too much legendary sounds like a wonderful problem to have

[vernal bay]

better that then having to recycle and delete stuff that's to high quality

[vernal bay]

If you build your quality base before unlocking legandary any legenary at all is to much legendary,

[half raptor]

Recycle loops is how you deal with overflows. Honest you kind of want to keep your line running even if you're recycling half because 1/4 of new products made are upgrades and so keeping that production line cooking is great for higher quality acquisition

[vernal bay]

that wastes way more resources then if you could use higher quality ingredients for lower quality recepies or had a qualty downgrader assming you want to target anything bellow legendary (or the higest unlocked quality)

[half raptor]

If you're going to downgrade one ingredient to match a lesser ingredient to keep that flow going, you might as well have put productivity modules on making that step which is by definition a potential loss of resources. If you're trading quality for productivity, downgrading is a shame and a waste

Also I plan to have circuits measuring my demands and reporting back. If I have a high demand for and I have low demand (near full buffers) for and then it'll alert me to build more production

[vernal bay]

The alternative is that you never use any quality besides the highest unlocked quality so then then (depnding on your tech level), with the way quality currently works that's the best solution becouse you can't save on resource or build complexity by going for a lower level, trying to use is completly pointless.

[half raptor]

Honestly the quick and easy solution to "not enough ingredient" for production to set up a block that uses prod modules, and circuit it to only turn on if demand is high

[vernal bay]

That's true for but not for (or anything else bellow the highest unlocked quality)

[half raptor]

I plan to go for quality on every step, the more steps a product needs, the more quality chances I get.

[obsidian crescent]

Nonsense. You just store or recycle it and use it somewhere else.

If you're trying to make Q2 power poles, you'll get some Q3-5s too. Your setup can treat those as side-effects and deal with them like side products in any other build setup with side products.

[half raptor]

So there is also a much easier solution for people who get frustrated with quality. Go for Productivity bonus for intermediates and then make a quality loop for the very last step. It's not a waste of resources if you're pumping out a huge prod bonus to counteract it

[vernal bay]

then you have to rebuild every step the moment you unlock and so you can add recepies for thos qualites everywhere

[obsidian crescent]

Would you have to rebuild it? Couldn't you just have the infrastructure in place and just set the filters when you research it?

[half raptor]

Well yes, but my inserter for checks for demand for epic before loading so it'll save the epics until I set up factories that demand then

[vernal bay]

That still won't help with the to high qualty problem, I can still get random rare/epic/legandrys even if I only want uncommons if my blueprints are build around uncommens, and the only thing I can do with the to high qualty ones is to recylce until nothing left.

[half raptor]

Make each build modular. One for normals, one for uncommon, etc. If the block stays inactive for a little while then that means you should set up more normal production, not only for the chance it'll spit out some uncommon upgrades but also if you overflow and recycle, you get the chance to get uncommon ingredients

[obsidian crescent]

Or you recycle them to their base materials and use them elsewhere. Again, treat them as side-products, a resource to be used, not a thing to throw away.

Also, just because you're using Q2 stuff now doesn't mean you want use Q4-5 stuff later.

[half raptor]

If you're looking for a 100% efficient every machine runs all the time factory, quality is not a fun thing to chase

You're going to have an overbuilt factory that won't necessarily run all the time until you increase production and manage to get your hands on legendary modules

[vernal bay]

I want my factory to be an one uniform level of qualtiy at any given stage of the game, I want a progression like this to be viable Everything in the factory (or everything that handles significan't production and isn't like an assembler in the mall) is => => => => but since you get from the start you end up with a bunch of stuff you can't use in the so you either recycle and above or skip the stage since an dosn't really make sense when you can't save on complexity by only producing (only having 2 version of every step instead of 5).

[half raptor]

It's going to be rough though one idea I've seen passed around is to recycle loop plate until it is the quality you want then bring it back to your factory

[vernal bay]

I'll definitly use a quality downgrader mod or even better a mod to set the maximum output quality (if that's possible)

[half raptor]

Mods certainly solve logistical problems and im sure there will be a day 1 mod that'll upgrade every craft automatically to legendary

Personally going to go in vanilla and solve through modular production blocks on a train grid. I have the idea mapped out but it'll alert me when to build more or any other intermediate that I have enough ingredients for, including for quality alerts if I have too many of a quality ingredient and a demand for the quality product

I just can't test any of it until we get the game

[vernal bay]

If you go trough the effort of all that modular production with 5 modules for every single production step in the game that can be made with quality, why would you ever use ther intermediary quality.
Like let's say you have a green circuit blueprint build to use if you randomly used and above in that build becouse that'd be pointless the ratios won't work anymore or inputs/outputs won't be enough anymore so it don't make sense to use assemblers for that build. and if your build is designed for you don't want or machines in that build. Having 5 blueprints and picking depending on what quality you have in storage also seems like way to much work for the benefit. The only reasonalble option is to only use the highest availible quality of buildings when designing blueprints which means you nver really use the intermidate qualitys.
I'm not taling about what quality recepies are used (for maximum efficienty you should probably use all 5 at every step and use quality modules at every step) I'm talking about what qualites of builings/modules are actually worth using/desinging blueprints around.

[obsidian crescent]

I'm confused by this whole "the ratios won't work if I use the wrong quality" thing. A green circuit setup is built out of two things: copper cable makers and green circuit makers.

If I design the logistics between the machines such that the ratios exactly work when used with Q2 assemblers, then they ought to exactly work with machines that are faster than that. The only way this would not be the case is due to some insufficiency in logistics transport between the machines or into/outof them (inserters too slow, belts too slow, etc).

And even then, even if it doesn't output much faster with higher quality... it still works. It works just as well as it did. It won't produce less than it did before; it just won't produce more. It'll even be more power-efficient because the machines will spend more time idling.

Even just upgrading some machines and not others won't fundamentally break most setups. I know many setups are highly tuned to specific ratios, but even if you untune them, they still generally produce and consume what they used to.

[small egret]

My problem with no ”any quality” is that ”any quality” was the ONE thing that seperates it from normal factorio logistics. Without it different quality items are effectively just different items with similar icons and cool guis. I know the logistical challenge will still be fun, I will just be a little disappointed that’s all :c Especially when I would personally rather have the exploit in the game than losing the any quality

[rotund egret]

setting things up without "any" quality is 👺

[half raptor]

I don't think they understood what I was getting at, but quality is either a love it or hate it feature.

The only point I was making was if you are downgrading quality items then there is the waste, considering you at that point should have used prod bonus instead of quality bonus. Better to either let it buffer and wait for the other ingredients of the same quality or recycle the excess so start filling up higher and higher quality buffers and only when they're all full do you tell your inserters to sit on the buffer until there's a demand again

[small egret]

I think the ”use prod module instead of downgrade” is a really good point that I didn’t consider before. But I still feel like downgrading has a serious use in an optimal factory. Besides that, the downgrading solves the excess high quality items problem in a much simpler way than switching to prod modules which I think is a reasonable tradeoff for many people.

[half raptor]

My circuit logic is basically going to work like this. Each of my production blocks that goes for quality is going to have one station for expected qualitiy and a 2nd station for upgrades, that way I don't have to hit every quality station just to drop off a train load of products. In my upgrade station, there will be 1 chest for legendary, 1 for epic, rare, and then 3 for uncommons. Each one will be wired to a combinator that can read global demands for items (for my own alerts, but I can piggy back off of it here) Basically if the demand for legendary is 0, either because it is full or because I haven't built it yet, the legendary chest's inserter will not be enabled and will save those items for another day. My epic chest is also checking for epic demand, but it's also looking for legendary demand since I can always tell my train to go recycle if destination is full trying to deliver to an epic demand station. This logic goes all the way down to rare, uncommon and finally my normal. If I have demand for legendary for that product, all the inserters will be enabled for loading trains.

Eventually yes I will overfill one of my ingredients with legendary first, then the epic will fill up then the rare and eventually even the uncommons will be full and will not load onto a train because there's no need for upgrades beyond that point. This may junk up my normal production cell which may have too many upgrades jamming the belts, so I need a way to keep the flow of normal items moving. In the far distant future where this becomes a problem, I would build a cell that creates normal product, but with productivity modules instead, just to keep that flowing and it will only activate if the demand for every upgrade quality is 0 and the demand for normal is > 0. This solves my backing up problem, though getting full buffers of legendary and epic seems like a far away problem.

[grand igloo]

yeah. Also the exploit wasn't even trivial to exploit, and honestly was fun to theorycraft and probably would be fun to implement independently. Though I guess they probably fixed it because it would just get blueprinted and handed out like candy anyway.

[half raptor]

To be honest, the theory crafting will die when we get the first "quality base in a book" blueprints

[daring siren]

I'm sure Nilaus is working on it 🙂

[half raptor]

You're right and I hate it