Friday Facts #376 - Research and Technology

[mmmPI]

Moreover, nothing they've written has suggested that there are any guarantee's about the quality that comes out of an assembler on any given product. This is so jarring to me, logically, its like Optimus Prime showing up in Lord of the Rings.

Same here, this is wrong, you meant "nothing that you read/understood" , because it was written this in the FFF 375 :

but an important part of the mechanic is, that the quality of the ingredients is the base for the quality of the product, so getting higher quality intermediates is valuable too.

Ok, point 1, I'm not getting it.

I just wanted to let you know that you posted things that are wrong. You have not fully read the FFF and complained about things here without reasons. You don't even feel embarassed and push again your proposition that is imo not good at all. It has less depth in terms of gameplay, i already played mods with scraps. I'd rather try something new.

If you don't understand and it still sounds illogical to you, it's ok, after all it's a game where you can put a 1000°C reactor in your inventory.

But once you start making components out of those materials, things get almost completely deterministic.

No this is not true at all in the semi conductor manufacturing industry.

[FuryoftheStars]

V6 and V8 are different products. Like assembling machine 1 and 2
A regular V6 would have less horsepower than one made from better-balanced cams. A regular assembling machine 1 is slower than an assembling machine 1 made from more precise, or "rare" gears.

You can't get 250% performance from a V6 by just balancing cams better

I suppose an argument of "gameplay > realism" could be made here, but I do agree with you. They went so heavy-handed with the bonuses that it just feels like another/cheap tier system on top of the original tier system. That may be part of what's contributing to my struggles with acceptance of Quality, aside from the RPG feel that the names and RNG gives it.

[Khagan]

You design an exact chip, with a specific number of transistors, for a specific purpose. During production you do not magically get a more useful/powerful chip because additional transistors appeared in the final product. Nor do you set out to make a V6 engine and occasionally get a V8.

It's the other way round. You set out to make a V8, and sometimes get a V6 because one pair of corescylinders isn't working properly. No, this doesn't happen with car engines, but it absolutely does with chips. Concrete example: Apple's M1 Pro processors mostly have 8 fast and 2 efficient cores, but the cheapest version has only 6+2; the cheap ones are not a different design, but QA rejects from the same wafers as the main design.

[Tricorius]

You design an exact chip, with a specific number of transistors, for a specific purpose. During production you do not magically get a more useful/powerful chip because additional transistors appeared in the final product. Nor do you set out to make a V6 engine and occasionally get a V8.

It's the other way round. You set out to make a V8, and sometimes get a V6 because one pair of corescylinders isn't working properly. No, this doesn't happen with car engines, but it absolutely does with chips. Concrete example: Apple's M1 Pro processors mostly have 8 fast and 2 efficient cores, but the cheapest version has only 6+2; the cheap ones are not a different design, but QA rejects from the same wafers as the main design.

Yes, but semiconductor manufacturers are clever capitalists. So when that happens, they “bin” it (disabling some features) and sell it for cheaper.

But imagine how annoyed everyone would be if THIS were the way quality were implemented in Factorio.

Yes, it’s more “realistic” to get worse parts when building better ones. But it doesn’t FEEL good when playing a game.

I like quality giving periodic bonuses that you can then try to make even better. I would be less happy with having to “recycle failures” even though it’s more realistic. Realistic if isn’t always fun. In fact, realistic is hardly ever fun.

[Tricorius]

Moreover, nothing they've written has suggested that there are any guarantee's about the quality that comes out of an assembler on any given product. This is so jarring to me, logically, its like Optimus Prime showing up in Lord of the Rings.

You keep saying this. But they have said it’s based off the quality of what you put in. The lowest common denominator of all inputs is the base, then quality modules provide a chance of it being better from that base quality.

Also, just to repeat, you don’t even have to start to consider quality variations unless you choose to put a quality module in the line. So don’t put them in your main production lines, just in a few specialized areas where you want to start improving quality.

You then get to choose whether you want to “go big” and gamble on the end component’s quality loop OR slowly ramp up the intermediate quality (starting with plates) for a more probabilistic route by feeding the machines higher quality inputs. Or I guess you could do a combination (which would probably be the best bang for the buck and maybe the most interesting logistics builds).

This was all about increasing quality, but an important part of the mechanic is, that the quality of the ingredients is the base for the quality of the product, so getting higher quality intermediates is valuable too.

[FuryoftheStars]

I like quality giving periodic bonuses that you can then try to make even better. I would be less happy with having to “recycle failures” even though it’s more realistic. Realistic if isn’t always fun. In fact, realistic is hardly ever fun.

I'd like it better if the outputs were guaranteed, just required more supporting (for lack of a better word) infrastructure to make.

[Tricorius]

I like quality giving periodic bonuses that you can then try to make even better. I would be less happy with having to “recycle failures” even though it’s more realistic. Realistic if isn’t always fun. In fact, realistic is hardly ever fun.

I'd like it better if the outputs were guaranteed, just required more supporting (for lack of a better word) infrastructure to make.

Technically. It does. You could start by putting all of your quality into improved intermediates. Then feed those into assemblers with a guaranteed quality output. But it’s going to be more expensive and take longer because you have to build all of your intermediates up. The base quality of the outputs is the lowest quality of all the inputs.

You can get a legendary iron gear by feeding it legendary iron plates, but if you toss legendary plates and a basic copper wire into a green circuit assembler you wasted your legendary iron plate. If you can feed your copper wire assemblers legendary copper plate, then you can get legendary copper wire and therefore legendary green circuits out.

The problem is that the infrastructure in front of your green circuits to only feed them legendary plates is going to have some complexity. And some luck, but I’m guessing by the time you’re to that point there will be a few ways of making it less random and more probabilistic.

I think it’s going to be pretty fun figuring out a way to “rush” building out a module to supply quality upgrades for personal gear.

Another fun one will be an early quality build for vehicle fuel. I’m curious what boosts fuel quality gets. Maybe it’s worth it to boost up fuel for sending down your power lines or smelting lines. I like sending solid fuel down my smelting lines (I don’t use electric furnaces until very late game, usually well after I’ve launched my first few rockets). So, is it worth tossing some quality modules into solid fuel and sending the better quality fuel down the smelting lines? Is it better to send the basic stuff through smelting and reserve quality for my car or trains? Who knows.

[aka13]

An interesting concern is also bad items, which get viable after a certain quality level.
In SE, there are oil cracking recipes, which are only worth it after you get a certain productivity level, and a certain speed module level.

I'd hope that the "useless" vanilla items, like the combat bots, or the mk1 power armor, or the solar panels, get a rebalance, which make the base version useful, and the "quality" versions great, not that bad items remain bad, but with a certain quality they can be useful.

[Tricorius]

I'd hope that the "useless" vanilla items, like the combat bots, or the mk1 power armor, or the solar panels, get a rebalance, which make the base version useful, and the "quality" versions great, not that bad items remain bad, but with a certain quality they can be useful.

I was thinking of that as well. I actually use all versions of the armor, including modular armor. They are still useful, just not for everything. For instance, you can use modular armor with batteries, solar panels, and bots for the stuff that is more annoying to place, like inserters or whatever. I run the belts and other easy stuff myself then kick the bots in.

You can also use a second set of armor with military parts. A shield or two can turn the tide in a close battle to your favor.

But I’m not sure how fast you get quality. It sounds like you probably don’t get it soon enough to be able to bump up modular armor and equipment that soon.

[Panzerknacker]

Reading all the critique in those threads I really hope the devs stick to their OWN vision and not cater to 'player feedback' because that is how so many potentially good games got ruined. Have faith in yourself, you know the best. You make what YOU want to make, wether we like it or not. We can choose to buy the game or not buy the game (I am sure more will buy the game eventually if you do it like that).

Great stuff so far imo, this is gonna be one sick expansion.

[FuryoftheStars]

I like quality giving periodic bonuses that you can then try to make even better. I would be less happy with having to “recycle failures” even though it’s more realistic. Realistic if isn’t always fun. In fact, realistic is hardly ever fun.

I'd like it better if the outputs were guaranteed, just required more supporting (for lack of a better word) infrastructure to make.

Technically. It does.

I mean right from the beginning, using a process other than slapping some quality modules in and filtering out the undesirables to a recycling loop.

The problem is that the infrastructure in front of your green circuits to only feed them legendary plates is going to have some complexity.

To maximize your chances, possibly, but slapping quality modules into everything in a multi tier recycling loop isn't exactly what I had in mind as fun. I guess ymmv.

---------------------------

Reading all the critique in those threads I really hope the devs stick to their OWN vision and not cater to 'player feedback' because that is how so many potentially good games got ruined. Have faith in yourself, you know the best. You make what YOU want to make, wether we like it or not.

Forgive me, but you do realize that the game we have today is because of all the feedback they got while developing the main game, right?

[Panzerknacker]

Reading all the critique in those threads I really hope the devs stick to their OWN vision and not cater to 'player feedback' because that is how so many potentially good games got ruined. Have faith in yourself, you know the best. You make what YOU want to make, wether we like it or not.

Forgive me, but you do realize that the game we have today is because of all the feedback they got while developing the main game, right?

Those were different times when the game was unknown to the bigger crowd, now there is toxicity everywhere fueled by the social media which are controlled by the same people that own Wube's competitors (gaming industry).

Also, they have a way bigger development team now.

Stick to your plan, and quality is a great addition to the game.

[KuuLightwing]

Forgive me, but you do realize that the game we have today is because of all the feedback they got while developing the main game, right?

Even though I don't like quality (as presented) myself, I think there's merit to what they are saying. That is, until the conspiracy stuff lol.

Those were different times when the game was unknown to the bigger crowd, now there is toxicity everywhere fueled by the social media which are controlled by the same people that own Wube's competitors (gaming industry).

Also, they have a way bigger development team now.

Stick to your plan, and quality is a great addition to the game.

If you want to imply that criticism is anything but people not liking the mechanic as presented, you are being extremely uncharitable.

Also I don't know what part of "gaming industry" is the competitors of Wube, I don't see many major players in factory building genre. What, Satisfactory developers are doing that?

[draslin]

Moreover, nothing they've written has suggested that there are any guarantee's about the quality that comes out of an assembler on any given product. This is so jarring to me, logically, its like Optimus Prime showing up in Lord of the Rings.

Same here, this is wrong, you meant "nothing that you read/understood" , because it was written this in the FFF 375 :

but an important part of the mechanic is, that the quality of the ingredients is the base for the quality of the product, so getting higher quality intermediates is valuable too.

Ok, point 1, I'm not getting it.

I just wanted to let you know that you posted things that are wrong. You have not fully read the FFF and complained about things here without reasons. You don't even feel embarassed and push again your proposition that is imo not good at all. It has less depth in terms of gameplay, i already played mods with scraps. I'd rather try something new.

If you don't understand and it still sounds illogical to you, it's ok, after all it's a game where you can put a 1000°C reactor in your inventory.

But once you start making components out of those materials, things get almost completely deterministic.

No this is not true at all in the semi conductor manufacturing industry.

And without explaining, how I don't understand, I'm simply supposed to accept your opinion that I don't understand. Nope. Explain what I'm missing, and I'll concede.

Meanwhile, semiconductors. If you say so. But I'd like to see an example of this supposed process. Show me a microchip factory, designed to make, lets say a pentium II processor, but every once in a while a pentium III popped out. Show me that, and I'll concede this point too.

[draslin]

You design an exact chip, with a specific number of transistors, for a specific purpose. During production you do not magically get a more useful/powerful chip because additional transistors appeared in the final product. Nor do you set out to make a V6 engine and occasionally get a V8.

It's the other way round. You set out to make a V8, and sometimes get a V6 because one pair of corescylinders isn't working properly. No, this doesn't happen with car engines, but it absolutely does with chips. Concrete example: Apple's M1 Pro processors mostly have 8 fast and 2 efficient cores, but the cheapest version has only 6+2; the cheap ones are not a different design, but QA rejects from the same wafers as the main design.

And that situation, in the context of chips, is fine. Its still low enough in the production chain that its not that jarring.

But what doesn't happen, is a factory producing V6 engines occasionally popping out a V8. Moreover, even if you were building V8s and occasionally only got a V8 with 6 working cylinders, you'd fix the issue. You would never be able to sell a V8 with only 6 cylinders in a brand new car. And more importantly, it would be the exception, not the norm.

There is an important difference here between making the best of a bad situation where the product you produce was inferior to the standard but still useful verses building to a basic standard which magically produces a far superior product every once in a while.

[draslin]

Moreover, nothing they've written has suggested that there are any guarantee's about the quality that comes out of an assembler on any given product. This is so jarring to me, logically, its like Optimus Prime showing up in Lord of the Rings.

You keep saying this. But they have said it’s based off the quality of what you put in. The lowest common denominator of all inputs is the base, then quality modules provide a chance of it being better from that base quality.

Also, just to repeat, you don’t even have to start to consider quality variations unless you choose to put a quality module in the line. So don’t put them in your main production lines, just in a few specialized areas where you want to start improving quality.

You then get to choose whether you want to “go big” and gamble on the end component’s quality loop OR slowly ramp up the intermediate quality (starting with plates) for a more probabilistic route by feeding the machines higher quality inputs. Or I guess you could do a combination (which would probably be the best bang for the buck and maybe the most interesting logistics builds).

This was all about increasing quality, but an important part of the mechanic is, that the quality of the ingredients is the base for the quality of the product, so getting higher quality intermediates is valuable too.

Huh. Well there you go. They are already essentially doing what I thought would make sense, at least up to the final product then.

I kept going back over the article but missed the sentence about the lowest common denominator AND misread the table of probabilities. I thought, at most, you had a 10% chance of higher quality due to the table of module probabilities. But there is an implicit 100% for each blank space in that table. At least if I'm understanding it this time. In which case, it would have been a lot less confusing if they had explicitly called out each cell in the table as a 100%.

In any case, thanks for the call out. I still think its weird and inexplicable on higher order components, up to completely stupid for apex products (I'm the engineer but I just can't figure out how to reliably bolt an extra cannon on this tank, best I can do is roll for charisma), but at least its semi deterministic below legendary quality.

[mmmPI]

Maybe it's the picture with the % of probability that you didn't understand, i don't know.

I kept going back over the article but missed the sentence about the lowest common denominator AND misread the table of probabilities. I thought, at most, you had a 10% chance of higher quality due to the table of module probabilities. But there is an implicit 100% for each blank space in that table. At least if I'm understanding it this time. In which case, it would have been a lot less confusing if they had explicitly called out each cell in the table as a 100%.

In any case, thanks for the call out. I still think its weird and inexplicable on higher order components, up to completely stupid for apex products (I'm the engineer but I just can't figure out how to reliably bolt an extra cannon on this tank, best I can do is roll for charisma), but at least its semi deterministic below legendary quality.

I knew it !

You are welcome, Though there is still a mistake, the implicit 100% is for the whole lane, not for each blank space, because it would make no sense. so really it 0% on every blank space except the right most one of each lane, where it's 100% - the other % of the lane. Because probabilities always sum up to 100%, so if you already have 10% 1% and 0.1% and you only 1 other choice, it will be 100% - (10%+1%+0.1%).

And without explaining, how I don't understand, I'm simply supposed to accept your opinion that I don't understand. Nope. Explain what I'm missing, and I'll concede.

I guess you understood part of it ... for the rest in case there are still things unclear : it was discussed several times already. Basically with the scrap mechanism you use RNG to punish the player with "bad product" sometimes, instead of the proposed system where it's only "good product sometimes", which is better for gameplay i found. given that many player seem unable to access what's necessary to avoid feeling punished by RNG.

Also otherwise it makes no sense to suddenly having your factory produce scrap in the mid game when you want better item ? and if you have scrap since the start you make it no-fun for many players as dealing with the quality as proposed by the dev is optionnal for those who wants to go to the simpler solution.

As i said scraps are fine for mods but leads a niche gameplay. Your propostion i would have ignored since i have nothing nice to say about it but you pushed it again and again as if i'm not understanding it, so i said what i wouldn't have otherwise, that i don't find it interesting at all for the vanilla game compared to the quality feature explained in the FFF.

what you describe is already existing as a mod : https://mods.factorio.com/mod/ProductionScrap2

Meanwhile, semiconductors. If you say so. But I'd like to see an example of this supposed process. Show me a microchip factory, designed to make, lets say a pentium II processor, but every once in a while a pentium III popped out. Show me that, and I'll concede this point too.

I feel what you are asking me shows your ignorance and willingness to ignore further evidence because it's not too hard to document and other people tried to explain it to you, but if you want to learn you can read this ,it's described with not too many words:
https://www.hardwaretimes.com/amd-ryzen ... -future/2/

[mmmPI]

I still think its weird and inexplicable on higher order components, up to completely stupid for apex products (I'm the engineer but I just can't figure out how to reliably bolt an extra cannon on this tank, best I can do is roll for charisma), but at least its semi deterministic below legendary quality.

Ok a tank is made of may parts, like a plane, so when they are assembled they are also tested and they can say meet the spec when the spec are at 100, but some of them would also meet higher spec like 105, not all though, as the are not made for it, but nothing prevent you from testing a few over limit to see where the failure occurs, and maybe during those test you will find out that 20% of them can actually wisthand a higher spec, say 150 which is great because 150 is also the spec used for the the demonstration models, or those to be upgraded to receive a new system somewhere. So yeah it's not magic, it's just that you label them of higher quality.

How come that is possible ? well because when the tank gun weight a metric ton, you don't look at 1 nano gam, nor at 1 milligram, if there is a little variation it will still meet the spec, otherwise it would be too costly, but if you make 10 000 guns, amongst those, it may happens than some of them are close to the nano, micro, milli order of magnitude to the spec. You could always decide to rescale your specifications to increase the value of the pieces.

Which can happens if a new model of tank just happens to require similar guns as before but with a higher precision to the spec, qualifying only 10% of the guns one production lane was making whereas 100% was passing the previous spec. So that the factory will start suddenly producing 10% of "higher class" guns due to them beeing looked after and identified.

Now if you imagine the engineer is making all the factory starting with his bare hand, you can easily imagine that he started using rough iron plate, i mean their weight and size are not super standardized when it's used to make burner drill, though , you can picture that if he makes millions of those rough plates, with say a weight of 100, then there may be a variation of 10%, but some plates will be exactly 100,0 as weigh, and some maybe even 100,00 or 100,000 it could become interesting at some point to identify those to use for a place where precision is important.

That could be the weight but that could also be the precision of the shape for a gear, the vibration at 200 km/h for a car, the decay for a tank , the composition of iron and coal for steel things , the conductivity/resistance for wires, things that are based on so many factors or that can have so tiny variations that still matters that it's tested after it's made.

And the quality module would be like the tool that is used to measure the pieces to certify quality level.

[dead-duck]

Grade 1
Grade 2
Grade 3
Grade 4
Grade 5

- Kovarex said the criteria is clarity and feels good

- Self evident versus the lack of clarity where exactly some terms like "improved" and "enhanced" would fall.

- It's all impersonal function that feels in-universe with Factorio.

hit the bulls eye

Disagree, immersion plays a role for me, also in Factorio. Amongst many things, the graphics contribute to that, so do unique names for items and quality levels.

That's interesting because its the marketing names like legendary and epic that break the immersion for me. The character that builds these ugly industrial machines would not name the items "epic"

[Tricorius]

That's interesting because its the marketing names like legendary and epic that break the immersion for me. The character that builds these ugly industrial machines would not name the items "epic"

I would. I’m stuck on an alien planet with bugs trying to eat my face. I’m going to take opportunities to amuse myself.