Friday Facts #376 - Research and Technology

[Xorimuth]

The whole point of manufacturing is to produce an item of exacting, consistent, and reproducible specifications with as little waste as humanely possible.

Real life quality mechanic: https://arstechnica.com/science/2022/01 ... -lifetime/

[draslin]

The whole point of manufacturing is to produce an item of exacting, consistent, and reproducible specifications with as little waste as humanely possible.

Real life quality mechanic: https://arstechnica.com/science/2022/01 ... -lifetime/

Not sure I follow? Is it supposed to argue for or against the way wube wants to do quality? I don't think it provides support for either side of the argument, the way NASA build's rockets is different from the way one would build something for mass manufacturing like we do in Factorio.

[Tooster]

Awesome! I liked the quality idea in the previous FFF, but now you even addressed the thing I was almost sure won't ever be addressed — the "raw" feeling of completing researches in bulk. The triggers have huge potential for other mods and they seem like very interesting addition! I imagine the ruins mod adding things like "find ancient technology in ruins". I was also always taken aback by the sheer amount of unlocked items and how I "missed" some researches because they were queued in the background. Oil processing, barelling, uranium were always quite overwhelming. The addition of triggers seems like a great way to add some more awareness and attention to unlocked technologies with a stronger feeling of being a part of the progress. Often before the researches were just queued and completed automatically in the background, we will see how it goes along now!

[FunMaker]

I am okay with quality. I like the loopbacks it creates(like some mod recipes). I am a little bit afraid of inventory clutter it may create.

But I am worried about the triggered technologies. I know, it is a way to reduce technology creep with the expansion. Researching to a point where i am good to go to build a complete complex feels good to me, because i have a greater plan in mind that i can lay down. But putting research triggers everywhere feels like a bad gating for advanced players. I am okay if the triggers are put at choke points (i would see 1-3 of them in a current 1.1 vanilla game). But the shown examples look too much for me (like some have written previously).

One overdone bad example in my mind:

• Research Robots: unlocks Robotport, Construction Robot and Storage Chest, Blueprints
• Next deconstruction you make will unlock: Deconstruction Plan
• Building something with a construction robot unlocks: Logistic Robot, Personal Request Slot
• Transporting something with a logistic robot to the personal Request Slot unlocks: Passive Provider, Active Provider, Personal Roboport
• Laying down a Blueprint unlocks: Level 2 Construction Robots
• Building something with a Level 2 Construction Robot unlocks: Level 2 Logistic Robot

I know it is a nice way to make transitions easier for new players and i like the idea, that the recipes at the start are condensed, but please consider players that "know what they do".

And i would remove such triggers like "Build 50 Iron Plates". These are dull like hell.

And: Consider Speedrunning this.

[Drizznarte]

Great to get some feedback about last weeks FF .I am still worried about quality , you mention how the law of large numbers will just transform probabilities into ratios. You don't however justify the random approach rather than quality modules simply increasing the ratio of input items .The recycler is the only difference in this case. It does however add some mechanics that arn't currently in vanilla , can they load into a chest or directly into a assembler ? > can inserters pick up from them or can they only output on to a belt ?
Also in the example you give for assembling machine 3 , you dont use quality intermediate products . I am not sure quality should be used on items in the intermediate tab , this would also solve the need for quality versions of ore and such that just make no sence to me.

Research queue always on ! Fantastic !!!

Can you please also have a new music soundtrack made for each different planet and space , great for the atmosphere of the game .

[serhatozgel]

Do we have enough clarity on if we will need or want high quality items in science production, or if we will only use them in the ‘mall’ factory?

Or to put differently, will there be something like a higher quality science pack or a higher quality science pack ingredient (or ingredient of an ingredient) that makes a difference?

The idea of improved productivity in some intermediate products, therefore perfect ratios of the factory changing over time is very exciting. There won’t be a single perfect ratio blueprint for a factory anymore and people will potentially find new ways to optimise for months (years?) to come.

Speed runs may potentially take a lot longer to optimise. Do you upgrade this productivity or that productivity first to go faster?

I’m also hoping that some of these late game researches will require only red, red and green, red, green and blue science pack etc. and not all sciences. That way, once our initial factory of a certain spm is complete, we can continue to slowly grow it (e.g. increase red science to 2X now) instead of trying to upgrade the entire factory to a much higher X spm at once.

[Tooster]

One thing I am worried about with the quality of items is how they behave in GUIs and item stacks. If I put a quality 5 (supreme?) cog in the assembler and want to add another cog of other quality 1 (subpar?) — are the qualities summed/averaged/combined somehow or is the slot "locked" and can only accept other cogs of exactly the same quality? I would certainly much prefer the first option, so that many items with different qualities can be used at once and the final quality is determined based on some math formula for that.

[Sopenas]

Again about quality. Instead of this "salvaging setup" IMHO would be much better to have "integrated salvagers" in assembly machines.
So you would just setup on assembly machine what minumum or exact quality you want and it would try to produce stuff using same algorithm you have implemented now. Just that it if does not produce required quality it will auto salvage it and put stuff into machines input slots.

Regarding technologies i think this sounds interesting and natural, so it is good.

[mmmPI]

Not sure I follow? Is it supposed to argue for or against the way wube wants to do quality? I don't think it provides support for either side of the argument, the way NASA build's rockets is different from the way one would build something for mass manufacturing like we do in Factorio.

From my reading, it's not about NASA but the European Space Agency. They also send rocket to space, like we do in Factorio. And it's written that to launch the James Webb telescope they didn't use a low quality engine in the rocket, not even uncommon or a rare, but at least an epic or legendary, probably because it is the James Webb telescope, and not just another satelites that would cost 1/10 or 1/100 of the price. That's probably not what one would expect for gears or copper wire, but it seem that for rocket engine they generally make many prototypes, and discard most of them to keep maybe the 1% or the 10% that performed the best which sounds pretty similar to how quality was explained.

Maybe the engine was made using quality modules that is not specified.

[D0rus]

One thing I am worried about with the quality of items is how they behave in GUIs and item stacks. If I put a quality 5 (supreme?) cog in the assembler and want to add another cog of other quality 1 (subpar?) — are the qualities summed/averaged/combined somehow or is the slot "locked" and can only accept other cogs of exactly the same quality? I would certainly much prefer the first option, so that many items with different qualities can be used at once and the final quality is determined based on some math formula for that.

The devs said a few times that when you construct from multiple quality sources, the end result will be the lower quality source item (and then quality modules could roll the dice on a quality upgrade, but their odds aren't affected by anything other than the lowest quality input item).

That said, it's unclear if they where talking about multi-ingredient inputs or that you can also mix quality on a single stack. Since in your inventory/chests etc every quality item will create its owns stack. I hope it's the first thing where you can mix quality stacks in assemblers, otherwise you might get a hard puzzle where you want to upgrade your items 1 time, but random luck sometimes gives you +2 quality levels and these items make your machines get stuck. (The reverse puzzle where you have to filter out the quality items you care about and ignore those you dont care about seems more reasonable).

[adam_bise]

I love quality but on the fence about trigger techs, as I stated in previous posts. But I want the devs to know this: I love this game so much better than other games that I will likely gift copies of the expo to my friends. I just wanted to say that even if some of these discussions might be critical, it doesn't change the fact that overall your game is a treasure and a great gift to me. Thank you!

[Cakoluchiam]

Maybe it is a bit late to offer suggestions on the Research Queue, but after reading the original post about removing it, I realized that even with the notifications, I do still have the problem of not mentally keeping up with my research—as a mitigation feature, since you already can't add to the queue without looking at it, I think it would be helpful to keep recently-completed technologies visible in the queue (but turned green, like completed technologies in the bigger list) until you either clear them or queue more tech than the queue can hold.

[Saphira123456]

Hi, casual player here. As someone who plays this game, whether with mods or without, I have to say I LIKED being able to put down conveyor belts, mining drills, crates, steam engines and other basic pieces of equipment right off the bat. Knowing how to build and deploy these things from the start is a pretty basic tenet of the game.

And we need those steam engines up and running fast to start making stuff for our defenses. The alternative to this is giving us renewable energy sources to put down, like water, wind and solar along with power storage, so we have access to reliable power in that early stage of the game. Having to research and craft these extremely basic things before being allowed to start the game, pretty much, is a pain in the neck!

I also agree with the idea that exploring for oil is going to result in a lot of unnecessary trips back and forth to plunk down pumpjacks, refineries and chemical plants, etcetera. Super-duper-uber frustrating.

These things are so bad that if you proceed with this, I'm sure that this will cause a mod to spawn that will disable this part and set it back to how it was before. Or it might even cause people to leave the game. Hopefully the former, most likely to be the latter.

[Locane]

Thank You, and...

Kovarex - your clarifications were welcome and it's clear that you understand the concerns, thank you.

"The path to winning needs to be fun" is true and something said by Mark Rosewater as well! You're not wrong.

Consider that a lot of people are telling you that the Quality system you've designed would not be fun, in whole or in part.

Another trueism in game design philosophy is "kill your darlings"... Being partner to a game design industry veteran myself has given me insight to what happens at the top companies in the design process.

Fit

There are a lot of diverse drivers that make games fun for players, and you've hit most recently on the Gambling itch - winning at a random chance event, and better, being able to affect the chances of the event, feels good!

But being stuck forever in RNG is not Factorio, to us.

A big part of why I personally like Factorio is the fantasy of engineering my way out of a survival situation with a fancy manufacturing space suit. That requires there to be some level of realism to the mechanics in the game, some justification - otherwise why doesn't my suit just fly me off the surface as a personal skin-tight space ship? Why do I die to biters at all?

That suspension of disbelief is thrown way off when something doesn't feel like it fits - a manufacturing machine randomly producing a somehow better and more efficient cog, or modular armor, a better circuit doesn't feel like it fits. Even if it did happen, it should be reliably replicatable with the right level of precision in a machine in a game about automation, and if it's replicatable it probably has its own process, and if it has its own process it probably needs some extra ingredients, and if it truly is better than its predecessors it probably looks a little different, and here we've arrived at a new recipe.

Consider Random Recipes for Better Fit Instead

Quality modules instead are Research Modules. They make machines produce some scrap, but allow discovery of new randomly generated recipes for higher tier products that are permanent.

These new recipes have a random chance to produce scrap themselves, range from good to bad, from simple to complicated, but ultimately they produce a tier of the product also randomly chosen. Many of them will be less efficient, but might fit anyway if the player has more of one resource than another.

Once a player has the recipe, they can automate it forever - this is the part that makes all that RNG work feel like it fits and feels satisfying still.

As an example:
You have 12 machines with Research Modules all making iron gears.
They produce scrap instead of gears at a rate of 20% per module per machine.
You have an automated system recycling the scrap, and harvesting the gears.
Your setup produces a Rare gear! You pick it up, and you're given the recipe:

Precise Iron Gear (+60%)
Requires Assembler Tier 3
1 Water
1 Light Oil
1 Iron plate
8% scrap chance

Now it is up to you whether you want to build out on that recipe, or wait for a better one. Maybe you'll get a "legendary" recipe that is even better and has 0 scrap chance! Or maybe it will be Epic and have 23% scrap chance and require sulfur, for some reason. Who knows, it's random, but - once you discover it, it's yours for life, and you can choose to automate with it or not.

This to me feels a lot more Factorio. I put up with some RNG, feeding resources in to my Discovery Factory, and I produce different items that give me recipes I can automate on in my "real" factory.

This leads me to situations where I am re-doing things or building them bigger and better (the main Factorio loop) because I got a MUCH better recipe, but it requires petroleum. Or I am hooking up my Discovery Factory to my Main Factory because I needed more space and had to move it.

I think this might scratch the itch you are looking for, but in a much more in-spirit way. It would require some UI work around recipe management, but it would be worth it.

[Ghulmeister]

I am super excited about the quality mechanic and i am definitely going to rush +1 medium electric Poles asap for that sweet both sides of an assembler-reach hehe.
it makes me think that maybe we will be getting "quality"-enemies aswell, depending on difficulty/evolution level etc. imagine a big-ass legendary endgame raid of the biggest aliens! maybe they also slightly scale with size aswell?

wish i could just freeze myself for 12 months and then wake up to binge-read all the fffs at once xD

i dont mind the research-triggers at all and imo its just people being unflexible that are complaining about it.
i totally agree with the point that every time you make a jump in tech (start automated production of the next science-pack) you just spamunlock a whole set of techs that you wont even be touching for hours. like, "oh i unlocked red chips" while i am still doing steel and ammo and havent even located the first oil-field.

[Tertius]

But being stuck forever in RNG is not Factorio, to us.

Please only speak for yourself and let Factorio staff collect feedback and weight the significance themselves.
Personally, I'm fine with the new functionality.

[blazespinnaker]

We're not talking about thousands of an inch variances here, we're talking about significant advantages over lower quality items. If you want a car to run at 100 mph and it can only run at 90mph because of substandard manufacturing, that is a failure. Particularly if what you paid for was a car that could do 100 mph.

By any measure, Quality is a step towards better simulation of IRL manufacturing. What you see is a very significant simplification of fault detection and containment - it is not an over complication.

Does it make for grindy game play? Well, hard to argue against that, as it is purely a matter of taste.

[blazespinnaker]

One rather large change, again Quality related, was the recyler and how it's self-direct-inserting.

There's nothing in Vanillla that does this for heterogeneous items (unless you count straddling resource patches).

This would be OK in the DLC if we got some bigger containers to put our inserters around.

Otherwise in order to have factory design flexibility we need to push onto belts and that feels like an unnecessary UPS killer.

Alternatively, the resources could build up inside of the recycler and we can pull them from there.

[daemonworks]

Research queue: FINALLY.

Quality: Meh? Hmm. Yes, definitely meh. I'm wanting interesting optimization problems to solve, and it doesn't seem right for that in it's current form. Maybe if there's more that can be done to influence quality level, so that optimal quality item production isn't just a simple loop that recycles anything that isn't of decent quality? I'd prefer to ditch the RNG entirely myself, but at least it doesn't seem as bad as uranium's RNG where it's more or less "now that i've unlocked uranium based power, i will do absolutely nothing with it until RNGesus decides to allow it", in that at least it doesn't seem to lock us out of anything.

[MrDayne]

Edit: I was pointed out that prod on labs don't count towards consumed science statistics, so I don't mind about this anymore. Maybe a hard cap on productivity for things other than mining wouldn't be a bad idea though.

I don't know if I like the change of endless productivity on labs. It would completely break how SPM work in the game. I know that SPM is very dependant on infinite technologies, and you need to play a world over hundreds of hours if you want to reach a point where you can get the most out of it, but now the more productivity you have, the more SPM you will get, forever. It will no longer be about how optimal a base is, but of how long you've been playing that world.
I like the extra productivity, but in my opinion it shouldn't be endless.