Friday Facts #442 - Flip, Flow, and Fresh Paint

[Mecejide]

Please tell me there will be a shortcut enable/download all recommended dependencies of the selected mod!

[Necronium]

At that point you're overinflating the power of space casinos. Even if you build a space casino you'd still need to gather many other materials by repeating the conventional methods across multiple production chains, so why is a bit more variety that big of a problem? I don't think people would rely nearly as much on space casinos if the other approaches were less tedious - and I don't mean challenging, just tedious - but I don't think it's reasonable to expect many changes to quality, so why bother removing a different approach from a system that already lacks variety?

It is more that you people try to downplay space casinos being OP. We wouldn't have that much meltdown by their removal if it weren't OP. And as Ive said if someone wants to have that not OP solution they can mod it in back. But at that point why not just mod in legendary everything rather than pretneding that you engage with mechanic.

Only that Space Casinos are not OP at all, if you consider that only the first 2 science packs are done with it on a larger scale base, the other science packs lack a method that can be scaled to the same level as common science, without the need to buy a new CPU and RAM.
Its true it makes bootstrapping a majority of base quality buildables trivial but so does ore washing with the only diffrence beeing you need 100x more buildings, but you need so little buildables anyways even for the largest bases that you can use ore washing on nauvis and never depleate a single patch to make all buildables for a 10M eSPM base. If we dont get any changes its clear to me devs dont play with quality and they didnt playtest there own game in this aspect of the game at all.

The only thing that is impacted by removing the casinos is quality science, not the buildables at all, since only need a very little trace amount of quality compared to science. Getting Legendary Quality buildables is as easy as it is before, only more boring, less engaging and with less of a logistical challenge than before, but with a longer grind tied to it
However Quality science is removed from the game, and that is the saddest thing about this.

It is funny when you people mention that they are not OP and then say that they are OP yourself. Quality science is not removed. The boring and OP solution to it is removed. Your entire argument comes as most of those comes that we should simplyfy game even more casue then it will be more fun. So assembler should just get ore in and throw a lot of science cause it will be even more UPS and who needs balance in single player game /s. As Ive said you can mod that in and pretend that you put effort in in getting something.

Quality science is gone. Not a single science can be made now with quality, efficiently. it will require 100x the assembling machines and inserter to produce a fraction of the eSPM compared to common. The system needs a overhaul, not remove the only way to make it borderline viable.
If we would consequently follow the arguments of people who think quality casinos are too op and want it removed, we have to remove inverters, chests, belts and miner because they remove the fun in handmineing, handcrafting, handfeeding, and inventory management. Casinos make quality borderline viable, not more or less. For buildables it does not matter in the first place since you can buffer them, and there are even easier ways than casinos to get more than enough you need for all buildables.

Im not gonna argue with emotional argument cause there is not logic there. Just mod it back in and pretend that you have skill, but at that point I would mod more fun in with everything legendary mod and stop pretending that it isnt OP

[emty]

Trains are unfortunately (almost) useless and quality won't save them.
Bigger acceleration and top speed is nice, but they are still inferior to belts.
Belts effectively teleport items, with a small downside of buffer capacity.
But they do not require all the setup with the stations and their management, schedules, intrusive tracks in the middle of the base, refuelling, congestion, and so on.
It is too much work for no reward.

Then only way to make trains competitive is to nerf belts very hard by limiting their max span or max length, like the pipes function today. Probably nerf pipes as well.


Asteroid reprocessing quality nerf is fine, locking asteroids route behind productivity upgrades like other upcycling loops sounds fair.


What I would really like to see is the following addressed:
* UPS improvement for everything space based (asteroids, collectors, turrets, thrusters, etc) or making promethium science pack require less asteroids chunks
* UPS improvement for cargo drops especially, since just dropping hundreds at once slows down the normal 8ms updates to 16+.
* Some improvement with save sizes (An optional incremental saves maybe? 100MB saves are too much.)
* Maybe tuning down the pollution on Gleba so it does not reveal map x20 of your base area, contributing to problem #2.
* Better alert system, so the game does not scream at you all the time when some mines die. And when the whole station dies, the alert is not persistent and just quickly goes away, as if it is not a big deal .
* Some improvements to circuit networks, since they are too tricky even with all the 2.0 improvements. Could you maybe implement something like fCPU mod natively in C++?

[LizardOfOz]

Suggestion for mirroring train station: make the train stops and signals rotate when mirroring a blueprint to preserve handedness. If the rotated train stop wouldn't fit, then block the procedure with a special message. This will teach the players to design their train stops and intersections with that in mind.

Alternatively, although it's a bit ugly in terms of backwards compatibility and is more restricting in general: make train stops and signals occupy space on both sides of the track, with the opposite side being "empty", unless you fill it with the second train signal.

[Leex2k]

Trains are unfortunately (almost) useless and quality won't save them.
Bigger acceleration and top speed is nice, but they are still inferior to belts.
Belts effectively teleport items, with a small downside of buffer capacity.
But they do not require all the setup with the stations and their management, schedules, intrusive tracks in the middle of the base, refuelling, congestion, and so on.
It is too much work for no reward.

Well, yes, everything would be useless compared to belts for the purpose of transportation, if one granted the premise that belts effectively teleport items.

Of course that would be a silly thing to do, since belts do not, in fact, teleport anything. Things on a belt move along that belt at a very steady and limited velocity. Very much a concept that is widely understood to definitively not be teleportation.


I do agree that some UPS improvements to space harvesting and shipping could be welcome, but it's probably a stretch to expect since it would require some serious compromising of the whole asteroid dynamic.
A more adjustable alert system is also a great idea. I am pretty much conditioned to ignore them as well, because of AoE rockets and mines and such. And the constant blinking of construction bots needing this and that.

[bobucles]

Does train quality improve their braking force? I notice that braking force is not actually a stat on the info sheet (aquillo thawing energy is also missing), but it's also a key stat for train performance. Side note, it'd be nice for tanks to benefit from braking force research, they slide around like they're on ice after a few speed boosts.

[emty]

Of course that would be a silly thing to do, since belts do not, in fact, teleport anything. Things on a belt move along that belt at a very steady and limited velocity. Very much a concept that is widely understood to definitively not be teleportation.

Ok, then what happens if that limited velocity (180/240 items per second) is big enough for feeding a machine through the inserters?
If inserters consume items slower that that, and the production on the other side is adequate, then I don't see much difference with actual item teleportation.

If you do train to machines insertion, to keep the flow steady you need a buffer chest, which is an extra inserter in the chain.
Without any buffer to mimic the belts, the trains themselves must rotate instantly ("teleport", yes), which they don't do, because of acceleration/length limits.

If a machine needs > 200 items per second (ore smelting, landfill), so that even a single belt cannot keep up, then a direct mining is probably needed anyway.

[andystrangelove]

Also sad to see the space casino nerf. I've read the arguments for it and I'm not convinced, and agree it's plain weird that the LDS shuffle remains but casinos are gone. Strange to see a dev say "we don't want to be fun killers"... then don't be? The LDS shuffle feels much more like an exploit, whereas casinos are still intensive to build and really only apply at endgame and megabasing.

The nerf is particularly annoying for quality science, which is a genuinely interesting mechanic for increasing SPM. But once casinos are gone, quality science requires so much waste that it's not worth it compared to just making more standard packs.

The deeper issue is that quality lacks polish as a mechanic in the first place. It feels like an early-access idea bolted onto a mature game, and given its design, unexpected interactions were inevitable. Look at the workarounds: recycling completed products for better odds, recycling chests instead of steel because it's faster. Even conceptually it makes no sense — making high quality equipment ought to require advanced production chains and a refinement stage, not turning everything into scrap on the offchance your garbage disposal magically turns out a better item than went in.

I wish they'd just given us new tiers of existing equipment — fusion bots, advanced roboports, tier 4/5 assemblers, etc. Everything in Space Age that's genuinely cool is the new recipes and machines; quality is a hacky shortcut to higher-tier equipment instead of designing and balancing specific new things.

Mostly I'm just disappointed. I'd hoped 2.1 would bring serious changes to make quality more polished, balanced, and practical, and while I get the logic behind the casino nerf, it's the absence of quality improvements elsewhere that makes it really sting. I also wish they'd engaged with the community more before going ahead, to get a read on how people actually feel about it.

And honestly, making Space Age playable without quality is a tacit acknowledgement that it's poorly integrated. I'd rather they reconsidered that than made it optional and nerfed casinos — give the mechanic the polish it deserves.

[SuperHyperMegaTurbo]

Might you consider adding a way to supply general quality filters by wire?
The same way Inserters or Splitters can use a quality filter with no item when set manually, but do that by circuit control.

Basically the "generic quality-only filters" part of this mod (though I guess the specific-item filters could be handy, too):
https://mods.factorio.com/mod/circuit-c ... ty-filters

[bobucles]

Buffs are always good game design, and nerfs are always bad game design. It's the classic game dev conundrum.

Ultimately they'll have to do the math and decide if yes it's fine, no it needs tuning, or hell no shut it off entirely. I do think there are a few viable tuning levers that won't have much collateral damage, but that's just an opinion.

[Gemma]

What a beautiful game. Thank you for these wonderful changes!

I would have preferred if asteroid reprocessing simply reduced quality, making legendary reprocessing less fruitful rather than banning it outright. I put a lot of work and fun into building and designing mine and I find them very satisfying, and I also use other methods to produce quality because I also enjoy those. So I find this change to be disappointing, although I understand the reasoning and agree that reprocessing trivializes quality in an unhealthy way. A big, high-productivity recycling plant in Aquilo deserves to be the lynchpin of a legendary factory, rather than the sidekick. It would be nice then if asteroid reprocessing platforms were able to be sidekicks, rather than entirely obsolete.

[Concordian]

First, I want to say that most of these changes are excellent and highly appreciated! There is a lot here that will improve the game, and I really look forward to testing them - especially the new space logistics!

On the asteroid reprocessing quality change specifically, Devs, please reconsider. In this topic alone, despite some vocal advocates, the count is roughly 3 to 1 against removing it. Factorio has always been about progressing from less efficient methods to more efficient methods: stone furnaces to electric furnaces to foundries, hand crafting to assemblers, basic production to advanced production chains. Asteroid reprocessing fits that pattern as a late-game way to enhance some parts of quality, while still leaving other resources, such as holmium and carbon fiber, to be handled through other methods. It does still require design, planning, and logistics, but it allows the few materials which can be produced in space to be produced at scale. Leaving asteroid reprocessing in the game does not harm players who do not want to use it; no one is forced to build that system. And the “just use a mod” argument cuts both ways: if someone wants asteroid quality removed, a mod could remove it just as easily as another mod could restore it.

The larger issue is that quality needs a better late-game path. A 10% maximum chance at each step is fine for early game, but there should be research or another progression system that increases the possibility of getting quality products, or recycler research which increases the percentage of returned items. I'd love to build a system to use quality science, but I was disappointed to learn it is much easier and more effective to make 100 normal science than 50 uncommon science. Legendary science is just pointless as it takes far more materials for only a 6x buff. It's better just to build more labs and stock them with normal quality. The massive amount of upcycling for legendary products is killing my UPS, and it isn't "harder" than asteroid reprocessing - it's just more boring. Oh look, more blueprints of recycler rows chewing through insane amounts of raw material. Great. We must be having fun now.

Bottom line is this is an incredible game which has consumed 5,500 hours of my life, and I'll keep playing it either way... but please don't fix what isn't broken.

[krozu]

After having played around with the quality system extensively, I do like the idea behind it. But...
A big issue for me is miners seemingly _creating_ raw materials with quality. One would assume raw materials have an innate quality, regardless of having it researched. You could argue that miners acquire a built-in sorter the moment you stick in quality modules, but even then that wouldn't work. They should still be outputting quality materials one way or the other. To me this is one big sticking point with quality. I don't have a problem with the pseudo random nature of it, but it magically churning out quality materials the moment you stick modules in while completely ignoring what was mined previously irks me.

While I know they are never doing an overhaul of the entire system at this point, because holy hell that would take some work, I still want to give my opinion. After all, what else is a forum for.

1. The first change is to fix what my initial rant was about. Have quality on raw materials from the very start, capped at normal. Make it so any machine gathering raw materials can no longer take quality modules. The research essentially means 2 things:
   1. For raw materials it means we can detect their quality up to the research.
   2. On top of that, we learn to craft materials up to that level of quality
2. Instead of machines having a chance to produce items with higher quality, do the opposite. Have machines create a product of a certain quality, with a chance of it being lesser quality. With this chance lowering the further down the quality tier goes. This does require the recycler to be unlocked the moment you research quality. Quality and recycling are so intrinsically linked I'm surprised this wasn't already the case.
3. Quality modules no longer increase the chance of getting better quality. Have it reduce the chance of it being less quality to make it work with the previous point.
4. Have recipes target a quality as they do now, but let them take whichever quality as input, as long as it's at least as good as the quality it wants to make.
5. With upcycling gone, quality modules in recyclers should do the same as other machines. Though I would argue that scrapping should have a far less likely chance of it getting reduced. But that's just me.

No gambling to get something better. Don't like a product is of lower quality? Recycle it to its base materials and use it in science, or scrap it into oblivion. I can't speak for everyone else, but I don't interact with quality at all until I unlock legendary (except on fulgora). When it comes to science not until way, waaaaay late. On top of that, when you do the initial research it won't brick your entire factory.

Before anyone starts asking "How am I ever going to get all my legendary machines?!"; Obviously the numbers would need tweaking so raw materials return higher quality way more often. Sounds like a math problem to me, not a design flaw. Though I suspect there plenty of those. I have no idea if any of this is technically feasible, but thankfully that's not my concern.

[GregoriusT]

Legendary science is just pointless as it takes far more materials for only a 6x buff. It's better just to build more labs and stock them with normal quality.

Actually in a lot of cases Uncommon Quality is better for Science Pack UPS, and Rare Quality for the White Space Science Pack.
See abucnastys youtube channel for how that works out, if you have a few days to burn.

The primary reason higher Quality Science Packs are pointless in vanilla, is that Quality Modules are less good for Science Production than Productivity Modules on a per resource basis. And since resources cost UPS too, this ends up in Quality Science Packs being only useful at lower levels.

Also please do not forget that the Asteroids floating around Space Casinos, as well as the Turrets constantly shooting them, are causing UPS Lag as well, which kindof makes your whole point about them moot, sadly.

[Hurkyl]

The nerf is particularly annoying for quality science, which is a genuinely interesting mechanic for increasing SPM. But once casinos are gone, quality science requires so much waste that it's not worth it compared to just making more standard packs.

Explain what you mean by "waste". Because the most obvious meaning would be the proportion of destroyed materials, and space casinos are not especially good on this metric: one legendary asteroid per 47.7 common input asteroids is a mediocre result for quality legendary tier 3 modules. And IIRC they are fairly subpar when using lower quality%.

Sometimes people mention "20% loss in reprocessing" versus "75% loss in recycling", but that's just making an oversight, since that numbers are only part of the equation and accounting for the different quality%, getting two quality rolls per cycle, and productivity bonuses on craft steps swings things back the other way.

I've seen one or two people ultimately just talking framing -- e.g. losing materials to recycling bad because it's "waste" but losing materials to reprocessing is good because it's "refining".

So I don't see how talk about "waste" supports the conclusion you're trying to make. Did you actually mean something else?

[bobucles]

Sometimes people mention "20% loss in reprocessing" versus "75% loss in recycling", but that's just making an oversight, since that numbers are only part of the equation and accounting for the different quality%, getting two quality rolls per cycle, and productivity bonuses on craft steps swings things back the other way.

I think that 20% decay is more significant than you think. Obviously, 0% loss means that 1 normal input will inevitably evolve into 1 legendary output, there are no losses so there are infinite chances to succeed. That's the power of lvl 30 LDS and blue chips, after all.

A 20% loss means that 1 input can expect 5 chances to upgrade before it's spent. A 75% loss means that 1 input loses nearly everything and can't retry, in this case getting 1.33 chances before the inputs are exhausted. If I'm not dumb at math it's just an infinite sum of the %_resources_returned to the nth_power ([E] %^n), obviously giving at least a 1st chance for trying. Any change to the recursive decay rate would be quite significant on how powerful the setup is.

The formula for how much "value" a quality module performs is pretty much the same as with productivity modules. The more resources that flow out of a quality module, the better. Always. High prod bonus, high crafting speed, and fast hungry recipes all give more output from quality modules.