Friday Facts #376 - Research and Technology

[Vulkandrache]

My beginning of the end was when the mod forced me to set up rubber production, completely locking up the techtree otherwise.
Making the rubber isnt complicated but its multistep, almost all chemplants and fluids, which are zero fun to build with compared to belts.
It took about 40 minutes to plan and make.
The ~50 pieces of rubber then sat on the belt for another 3 hours until i actualy needed it.

Rubber is used in the next research pack, plus numerous factory building supplies. It's not just a random unnecessary intermediate, and you wouldn't have been able to procrastinate it for much longer no matter how the tech tree was organized. The tech tree is always frozen when you've researched everything else and don't have the materials for the next research pack. It happened in a slightly different order, but it was inevitably about to happen anyway. It sounds like your actual problem is that you don't like fluids, in which case that was never going to be the mod for you, because there's a ton of chemistry, and that's not going to stop. Nothing about the Space Age expansion has indicated that it will have as much fluid crafting as overhaul mods do. Your anger at trigger techs seem misplaced.

You are right that i dont like fluids. And i figured out that Nullius uses them excessively.
But no, the dissatisfaction is not misplaced either.
Ive had the same happening multiple times with different items where the game told to stop playing
and do exactly this right now otherwise i cant continue research. Regardless of wether i wanted or needed that at that time.

If i load the stashed save right now the next milestone is 1000 concrete. Which for some reason locks "Robot speed 1".
So rather than researching stuff, unlocking multiple things that cost concrete and thinking "hm, maybe i should set that up soon"
(or my personal preference "i should research concrete making") i feel like im being told to clean my room.
"Concrete, now!!".

[Tricorius]

This could undermine that whole sub culture and ruin a big part of the game. Yes, some people will enjoy racking up that infinite SPM I guess, but anyone could probably do it just by leaving their computer on overnight.

So, a core part of your argument for you not liking the system is “they got it and didn’t have to ‘work’ as hard as I did for it”?

Seems..silly. Why should that undermine the enjoyment a subgroup gets by doing something fun and challenging for them?

One thing I have a lot of fun with is building out designs and logistics supply processes to deliver ammo to outposts for physical military defense instead of just stamping repeating clusters of laser turrets down.

It’s fun to me, so should I come here and ask the devs to remove laser turrets because most of the people just stamp those things down and the entire biter aspect of the game is trivialized…?

[astral-entity]

To be honest, your reaction to the quality-related feedback is kind of interesting. Despite the fact that you obviously took some time to go through it (which I appreciate), it reminds me of the typical we-do-what-we-want-anyway-and-we-keep-explaining-it-to-you-until-you-shut-up approach which is usually used by politicians when doing something unpopular. I'm just saying

I understand that you really really don't want to add tons of recipes to the game. And I think that's a good goal, design-wise. However, what if the quality modules would instead add a multiplier to the recipe (materials and time) and energy consumption? Leaving the original recipe as it is, only scaling it up directly in the assembler. So yeah, it would still take like 10 minutes or something and mountains of energy to create a legendary tier 3 module, but it could be done more compact and more predicable, without any RNG mechanics. And if you don't like it, at least make it possible to allow such alternate mechanics for mods.

I was wondering, if there could also be a mechanic to upgrade existing items to higher quality items, instead of just recycling them and try again?

Simply put: consider to allow multiple ways to get there.

[Tricorius]

As near as I can discern, you don't even make high quality resources to build your high quality products, its just random. My approach has you produce high quality materials at the base level.

The system they proposed (and is already working) has the concept of feeding higher quality intermediates in to lock in at least a given quality output. The quality modules just give you a chance to boost higher than this periodically.

So, if you had stacks of legendary ingredients that you tossed into an assembler making Spidertrons, you’d be guaranteed to get out legendary Spidertrons.

And I’m sure they have at least a few other ways to influence this and make it more deterministic. Or at least to push the probabilistic model higher.

As I’ve mentioned countless times, whining about this mechanism now, when we still only have the barest information about the expansion is kinda silly. I’m expecting that WUBE isn’t too worried as they ALREADY have a bunch of mechanisms in the code that address a lot of the common complaints.

They aren’t stupid people. They are way smarter than me, for sure. And if I can see some possible shortcomings, I’d assume their thousands of hours of play testing have uncovered some things they have figured out solutions to.

The way I see this playing out is that you basically have two factories (or in my case I’d solve it with two separate busses).

- productivity bus: this is basically my current play style which drives everything toward efficient science as consistently as possible
- quality bus: I’ll be building a looping bus that aims for the highest common denominator that I can currently make across all raw (plates, bricks, etc) and intermediates (circuits, engines, whatever)

The productivity bus is built on consistency, productivity, and speed. It’s meant to drive science as quickly as possible so I’m not tech bottlenecked.

The quality bus is there to supply me first, and the factory second. It will only build stuff where quality matters. And as I tech up, the highest common denominator will slowly morph a bit and raise. But it is there to make me stuff like legendary armors, personal roboports, construction bots, ammo, etc. It’s also there to build the highest level of stuff I can in my HUB. I want it building the highest quality miners, assemblers, etc that it can.

I also expect to have a few mini-modules that I want to internally improve themselves. I look forward, for instance, to building a module that is optimized to build and increase quality solar panels and accumulators for use in my solar builder train (or sent on the legendary Spidertrons meant for building which my HUB is producing).

I already like having a construction train that contains solar panels, accumulators, and a few other things to build my solar module blueprints. It seems pretty cool to me that over time that train will be loaded with higher and higher quality buildings. Which means over time I can stamp down fewer, more powerful, dense solar fields, and slowly pack up and recycle older low quality solar fields.

Will there be an achievement for pumping out a gigantic amount of energy completely from legendary solar panels and stored solely in legendary accumulators. Man, I HOPE so! That would, truely, be a legendary sort of achievement, dontchya think?

I think that is *pretty damn cool* and a fun upgrade, efficiency, and optimization puzzle that has its core roots in pure automation. I honestly can’t think of anything more Factorio than that.

I think a lot of people are caught up so much in the “loot box” nonsense that they aren’t considering the possibilities quality is going to bring to their creativity and optimization.

[Tooster]

The amount of "I AM NOT A SPEEDRUNNER BUT SPEEDRUNNERS WILL HATE THIS!" is just comical XD

Also the amount of "we, the people"... Please, all, speak for yourself, don't project your opinions on others...

I, for example, like those changes and have no say in what others will find tedious.

[mmmPI]

I would like to leave some short feedback on the quality topic.

In general it makes sense to implement crafting better items by some possibility. But what I find really awkward is that during deconstruction of Tier 1 item you can get higher tier components. If you watch latest animation with factory sample, you can get Tier 4 components from Tier 1 item. This doesn't seem right to me.
From my point of view deconstruction should always return some part of components either same tier or lower tier in comparison to initial item. Probably if randomness is introduced - there should be chance either get each component during decomposition, or get 'damaged' ie - downgraded version, or completely destroyed part, when you get nothing nothing.
I understand that it's a game, but still it somewhat always represented realistic model of assembly lines on factories, and I doubt it's believable or realistic that by tearing apart some machinery you could extract better components than it was initially built from.

The Tier 4 components from Tier 1 item should occur very rarely ! But the point you mention seem to come out regularly, and as you said i think is more a gameplay decision rather than modeled from the reality and made to be easily pictured as being analog to it that the recycling process can yield higher quality component than the source item. However when doing some research and from readings the discussions in the other 35 pages thread x) , i found some videos that hopefully illustrate a claim that it can happens :

Extracting gold from computer parts (Part 1) 14 mins 20 sec
that's some explanations and 2 sample prepared
Extracting gold from computer parts (Part 2) 17 mins 52 sec
That's a 3rd sample prepared and the more interesting part at the end where the gold is actually extracted, if you want to watch only 1 video keep only the part 2

From old computer parts like pieces of Ram to pure gold, which can then be used for any newer generation of processors in theory. They say in the videos that obviously the prices at which you get the scrap computers parts in real life and the chemicals plays a big role in the spread of the process, the price of the metal too. In real life Iron is cheaper than copper and both cheaper than gold, and even gold is not something financially opportunistic so it is not a widespread thing to recycle those, but on a far away planet, where there is no money maybe it make sense

Maybe the electric pole has a thin copper cable, making it an overall low quality one, but if you dissamble several of them, you could make a taller one, with thicker cable and it would be overall of better quality, maybe the car has a good engine, but the rest of the component not so much, which gives it an overall rating lower than what is possible to get from the engine reused with better car parts, maybe if you disassemble several inserters, you can make one that is faster, i imagine robotic arms capable of more movement could be more precise/faster but also require more motors / more ressources or more attempts at making it faster, during which some configuration of good component are tested, but do not yield the expected result, and the component are re-used for those which can so you somehow recycle "good" component from "not-good" inserter. I feel like it need a little abstraction from the player and i like it.

Sorry for the long ramble but i saw you are a long time lurker and it was your first comment and i thought it could be saddening if you broke the ice and felt it had gone unoticed

[Qon]

I read "Quality clarifications". It seems 35-page discussion was just wasting of time and NOTHING will change (except the naming which is not a "strategic" issue).

It was a waste of time (for improving the system), because:

• The 35-page discussion was very low quality, mostly talking about unimportant things like the name and useless critique based on misunderstanding of the quality system. And it was mostly redundant. It wasn't really 35 pages of discussion, it was 35 pages of the same post redundantly posted in copies over and over from people who hadn't even read the thread (or blog post properly) to see that their view had already been shared by others.
• Wube already knows how to design game systems better than more than 99.9999% of the player population (they made Factorio after all, the best game).

There are some good posts but those are <10%. They could all fit on the first page of the thread.

Kovarex just repeated his beliefs we did not agree. I feel disappointed.

I agree with Kovarex though. Because I thought about the quality system until I understood it well enough, making sure to not misunderstand it like everyone else.
I have seen a few rare quality posts with potential improvements though.

And I haven't read through all pages either, waste of time.

[Artorias]

I really like the trigger concept!!
I hope you also consider some changes with bots and their chests. I would love to enable blue chests by for example building x entities by bots or producing x amount of transport drones.

For my part I will just ignore the quality stuff.
I don't really like RNG on it's own, especially in factorio.
Also when building a mall I just want it to last until I finish my playthrough and not shifting into some RNG blueprint for every other item.

Excited for what's coming in the future!
Keep up the good work!

[Luaan]

Ah, nullius type rech unlocks. Very nice. That actually helped keep that game's massive number of tech tiers under control.

However, I still don't feel quality is right, and you still should have alternate recipes instead. But more of a substituting ingredients thing, using a red circuit instead of a green one, or steel instead of iron, but it is still the same recipe object.

But I am confident it will all work out pretty well, regardless, even if the transition period will suck.

I hope there's going to be a great modding API regarding quality. It'd be great if there was a more generalised way of treating these recipes and pseudo-recipes. There's load of mods that are already experimenting a lot with more freeform recipe-style (Galdoc's Manufacturing comes to mind; having multiple quality-like "tags" attached to materials would be pretty awesome there).

Imagine something like good old Angelbob's. You have raw ore that you process. Some of the ore comes in higher quality, which in turn gives you a chance of producing higher quality components. But you can also reach the higher quality with just better stuff. When you want to make something like a tier-3 ore crusher, you can use the usual tier-3 components... but you could also make them with higher-quality tier-2 components - they just happen to be strong enough to allow e.g. the higher production speed.

If there's freedom to develop mods like that, I'm really excited for what people like Angel, Pyanodon or Anachrony could do with this

[coppercoil]

The 35-page discussion was very low quality, mostly talking about unimportant things like the name and useless critique based on misunderstanding of the quality system. And it was mostly redundant. It wasn't really 35 pages of discussion, it was 35 pages of the same post redundantly posted in copies over and over from people who hadn't even read the thread (or blog post properly) to see that their view had already been shared by others.

You are right, but actually discussion quality is not important if you can find 3-5 posts which sound reasonable. I'm not sure if kovarex has found these posts, because reading entire discussion is quite boring, that's true.

[jodokus31]

The 35-page discussion was very low quality, mostly talking about unimportant things like the name and useless critique based on misunderstanding of the quality system. And it was mostly redundant. It wasn't really 35 pages of discussion, it was 35 pages of the same post redundantly posted in copies over and over from people who hadn't even read the thread (or blog post properly) to see that their view had already been shared by others.

You are right, but actually discussion quality is not important if you can find 3-5 posts which sound reasonable. I'm not sure if kovarex has found these posts, because reading entire discussion is quite boring, that's true.

One improvement for me would be, that no probability is used (which has the random character), but instead a bar (like the productivity bar) gets filled, which guarantees an item of next higher quality.
On a more general note: I really would like those type of recipes in the API, where you accumulate fractions of a certain product per cycle.

Which others did you find?

[gGeorg]

The quality feature intentionally designed as "separate grind loops" is sick. It feels like masturbation, again and again but result is non.
One of the keys of Factorio success is understanding of importance continually rising complexity.

Grind_loops do not add any complexity. They are just relatively small blue-print need to be stamped to grind out the goal.
However,
If you change the design,
allow recycle only the final products(defensive towers, armour, weapons, ..) not mid products (green chips) then quality need to integrated into the main factory, which rises up complexity. Rising up complexity in this case, is Not only design of production lines, but also logistics and sorting things. True explosion of complexity. The optimal blue prints which are used now, are no longer optimal. So people should think differently, build different factories.

I would really ask you reconsider quality feature away from separate_grind_loops into whole factory integration.
Keeep it optional is good, so player could create a base Fab(as now), or add an extra complexity by quality feature. This way players get new collection of optional challenges. That is the goal of the Expansion, i guess, new chalenges.

[KuuLightwing]

The quality feature intentionally designed as "separate grind loops" is sick. It feels like masturbation, again and again but result is non.
One of the keys of Factorio success is understanding of importance continually rising complexity.

Grind_loops do not add any complexity. They are just relatively small blue-print need to be stamped to grind the goal.
However,
If you change the design,
allow recycle only the final products(defensive towers, armour, weapons, ..) not mid products (green chips) then quality need to integrated into the main factory, which rises up complexity. Rising up complexity in this case, is Not only design of production lines, but also logistics and sorting things. True explosion of complexity. The optimal blue prints which are used now, are no longer optimal. So people should think differently, build different factories.

I would really ask you reconsider quality feature away from separate_grind_loops into whole factory integration.
Keeep it optional is good, so player could create a base Fab(as now), or add an extra complexity by quality feature. This way players get new collection of optional challenges. That is the goal of the Expansion, i guess, new chalenges.

Yea, this will end up with everyone just using bots for everything, because sorting five different qualities via belts at all intermediate steps is just going to be too much of a headache. People already gravitate to bot malls, this would be even worse.

[coppercoil]

Which others did you find?

I think there are several important points. I should refresh my memory, but unfortunately, I lost motivation of spending time on re-reading entire discussion again. After the last FFF, I have feeling that kovarex is not going to change something, there's not a single sign saying he found something important for him, nor a support of other players. I give up.

[SuicideJunkie]

if you want to build refinery before organizing oil outpost, no, must build outpost first

You just need to touch grass oil. I picture an engineer arriving at unknown world, having no clue what's needed and what's not. Stumbles upon oil — my goodness, time to remember how to build that oil processing stuff they taught us in the academy! Seems pretty natural to me.

I would like to proffer the suggestion that oil and perhaps some other unlocks could/should be the *discovery* of oil by revealing a chunk which has oil sources in it, rather than actually pumping some.
As a bonus, when your radar station detects some oil, you'll get a Ba-ta-da! sound effect and know to check the map

On the other end of the scale, poison capsules are a thing that ought to need some biter corpses to be produced before you can unlock and research it.

I feel like the key difference may be "things that Everybody Knows" and you only have to realize you need it before you can direct your research to all the starter stuff, contrasted with things you'll definitely need samples of because its an unexplored alien world.

[gGeorg]

The quality feature intentionally designed as "separate grind loops" is sick.

allow recycle only the final products(defensive towers, armour, weapons, ..) not mid products (green chips)

I would really ask you reconsider quality feature away from separate_grind_loops into whole factory integration.
Keeep it optional is good, so player could create a base Fab(as now), or add an extra complexity by quality feature. This way players get new collection of optional challenges.

Yea, this will end up with everyone just using bots for everything, because sorting five different qualities via belts at all intermediate steps is just going to be too much of a headache. People already gravitate to bot malls, this would be even worse.

No. You need just one splitter to peel off standard production of the higher quality production. Then direct the higher quality belt into the High-quality mall. And of course you need divert standard quality parts from High quality mall recycler back into mass pro. Which makes nice spaggeti challenge, with possible sushi on top. :=)

[Tricorius]

Yea, this will end up with everyone just using bots for everything, because sorting five different qualities via belts at all intermediate steps is just going to be too much of a headache. People already gravitate to bot malls, this would be even worse.

I feel like you haven’t looked at any of the examples.

To me, it looks easier to drive it with belts. But I will leave that as an exercise for anyone that cares to put a bit more thought into it.

[KuuLightwing]

No. You need just one splitter to peel off standard production of the higher quality production. Then direct the higher quality belt into the High-quality mall. And of course you need divert standard quality parts from High quality mall recycler back into mass pro. Which makes nice spaggeti challenge, with possible sushi on top. :=)

No, it's not. There's five levels of quality, and combining quality in a recipe essentially voids the higher quality item. Therefore in any recipe chain, products of each recipe should be sorted by quality and combined with other ingredients of the same quality. It's not something solvable by "a single splitter".

If you introduce quality in a production chain anywhere aside from the final product itself, you have to use it everywhere in that production chain. There's also a pitfall of mismatching quality proportions between production chains of different lengths, so naive approach would also fail. That's probably why devs imply that everyone just does recycle loops for the final product and not implements it everywhere.

[KuuLightwing]

I feel like you haven’t looked at any of the examples.

To me, it looks easier to drive it with belts. But I will leave that as an exercise for anyone that cares to put a bit more thought into it.

I feel like you haven't read the posts you've replied to. Both examples we had were just a recycle loop on a single product. Concern is about implementing incremental quality increase through production chain utilizing quality at intermediate steps, as this is likely to be more resource-efficient than recycle looping the final product.

[mmmPI]

No, it's not. There's five levels of quality, and combining quality in a recipe essentially voids the higher quality item. Therefore in any recipe chain, products of each recipe should be sorted by quality and combined with other ingredients of the same quality. It's not something solvable by "a single splitter".

Concern is about implementing incremental quality increase through production chain utilizing quality at intermediate steps, as this is likely to be more resource-efficient than recycle looping the final product.

There are designs that works for modded mall with 50 or 100 different intermediates, and that can work with only 1 or even 0 splitter. Yes it's "complex", you need combinators, wires and filter inserters, but there is nothing forcing you to do a single belted mall for all qualities if that sounds too complicated, you can use bots that's fine too, or make a huge spagetthi, or several branch from a bus, or separate dedicated factory, or even use trains dedicated to specific qualities, there are many different ways of doing that i haven't even tried but reading such comments is triggering me, because the expansion is not there yet, so i can't make one such mall to show you but please let me have this puzzle instead of saying it sounds hard to make.

If the "most-obvious-simplest-optimal-way-that-is-better-than-all-the-other " already existed it would remove part of the excitment of the new challenge no ?

I think the aim of the feature is to add this kind of complexity in the game, but not forcing it onto players.

That's probably why devs imply that everyone just does recycle loops for the final product and not implements it everywhere.

The first design shown was making green circuits , and recycling them, that's not a final product. I don't think the devs implied everyone should play the same way. Although from math it sounds logical to expect that upgrading the quality early in the production chain, would waste less ressources and allow using productivity modules on the chain. But it is also possible to brute-force recycle 1 finished product, in order to deal with the quality as few as possible and get only 1 item that you feel is worth getting in high quality. ( the second design)