Friday Facts #275 - 0.17 Science changes

[rolfl]

Two comments on this FFF:

1. Thanks to all the devs at Wube - Merry belated XMas and a happy (early) new year. Your efforts spent improving the game are valued by many, and I presume most people, and I am a very happy customer of yours, and wish you all well, and lot of success with 0.17 and hopefully 1.0 in 2019.
2. The discussions about prod 1 modules, that fell back on a discussion about beacons, is one that's critical for me in my mega-base games. Undoubtedly, in my experience, the biggest value of productivity 1, 2, & 3 modules is the reduced demand for raw resources, but in a mega base, they absolutely have to be combined with speed modules in beacons - but not for the reasons you may expect. It's required to team them with speed modules in beacons to improve game performance (UPS). Assemblers & inserters impact the frame rate in a linear fashion, but beacons do not... i.e. adding beacons improves the mega-base output without slowing the game.

Not many discussions about beacons/modules consider the UPS impact (entity update time), but I would love to find/build a spreadsheet that optimizes the number of inserters/assemblers/plants with productivity modules and speed beacons required to produce X number of resources per minute.

In a mega base I built not long ago, I converted a 2 rocket-per-minute base using no beacons to a 2 rocket-per-minute using speed modules in beacons and productivity 3 where possible in assemblers, and the UPS went from about 25 to 50.

[Serenity]

There is already a simple application for productivity modules right in these science builds:

Take blue science with 24 AM2. These need 1.8 chemical plants making solid fuel. But if you slow each one down with a PM2 you need exactly 2
When you upgrade to 24 AM3 you can use 4 chemical plants with 2 PM1 each to get an almost correct ratio.

All in all I don't really like the module + beacon meta the game pushes on you. The builds all look the same with it. But even on somewhat smaller scales it's absolutely required. Take a calculator and look at the numbers for low density structures for example. Without beacons it's absolutely insane.
Plastic and oil in general is another thing that benefits immensely from modules. Not just in resources consumed, but just the plants you need.

[featherwinglove]

I wonder, do my 2101 played hours make me a 'core player'? Or am I just one of those 'There might be some core players who agree (...)'??? Or maybe I'm just a filthy casual scrub, dunno? Tell me what do you think of me, pretty please.

I don't really know. I don't know how many hours I've played, it probably wouldn't be that many if I found every save file and added up the times; it would probably be more if Steam were reckoning it. I've made three gigantic mod packs, in versions 0.12.35, 0.14.21, and just now in 0.16.51 (I think it's finally done; 135 mods) and I still play the older two. There is something that I do know for sure...

I, a 2101-hour-player (and counting) say that pickaxe removal was good...

No one has explained to my satisfaction why those who don't want it removed shouldn't be able to mod it back in. I just had a dev PM me saying (inter alia) that it can be modded back in (he didn't say how.) I told him to make an announcement if that's really the case! Hopefully, it'll appear soon.

Something else, it's been easy to cheese in an invincible hand drill for a very long time (I have a 0.12 version), yet it seems not very many people have. Have you?

Edit (forgot something):

I say it as someone who poured some emo-shit over FFF in the past too. I matured since then, just saying (and hinting). Devs here are amazing, even when I feel like *I* know better...

Probably back when they took players seriously and didn't tell them to their faces that they were overreacting and being emotional, and instead explained their decisions better, or changed their minds. The problem now isn't so much the content of the FFF#266 update, it is how the devs have responded to the backlash.

Actually they have job offers open atm I think? Feel free to apply if any of you think you can do better, ok?

I hate MacOS, and Apple hates their customers even more, so I am definitely not getting that one. I've preferred PC (which has OS choices) over Apple since at least 1991, and my school was stuffed with LCs I worked with until 1997 when I could get away from them. I have never owned a Mac.

[malventano]

...
Are you referring to the catastrophe following FFF#266? I don't like anything in that one, but IMHO, the removal of AM ingredient limits is by far the least offensive of the announced changes. Wube's behaviour following it actually has me warning my friends off this game (that included not updating the roadmap thread for three months and then deleting my post when I did. And that's not the only part of it that is Blizzcon 2018 level bullshit.)

I'm sorry but that's a massive overreactiion.

Regardless of what you think, responding like this is an indicator of the problem. Even now, you are brushing off players' concerns in exactly the way we're complaining about. That is not going to help matters at all, it only adds irony to the fire.

You guys have to realize that Blizzcon 2018 was not the moment Blizzard lost touch with its audience, it was the moment that let the world know that it already had. Here, we are not observing Wube actually being so disconnected with its audience, we are observing the process that leads to it: a sickness that I would like to see treated and cured before it is too late.

I went back and skimmed through that prior FFF thread, and there did appear to be a decent amount of outcry. As someone who had missed that titbit of info in FFF#266, it does seem a bit silly to remove the AM input limits. I always thought that was a novel tiering to the assemblers. What was the reasoning for removing that limit, anyway? To me, it makes assemblers too generic/simple. Is it not possible for the devs to take a step back and look at that earlier feedback with less defense / an open mind?

[Unknow0059]

Why make the requirement be the Empty Rocket instead of letting it continue being the Satellite? Just make it clearer that a satellite is needed.

[Shakes]

Overall the cost streamlining is a good thing, I agree that in the current game grey and yellow science are very raw resource thirsty compared to the prior science types, and these changes should smooth that somewhat.

I agree with the requests to make uranium useful for something. I was thinking with the satellite no longer required to win the game, why not make the satellite take uranium fuel cell(s)? That way uranium becomes required for white science while the randomness of uranium processing doesn't become a factor for people trying to speed run the game.

I'm not sure the names for the packs really make sense - they're clearly still tiered since you use eg red science packs in *everything* and not just automation research. Maybe if you're renaming all the packs to give them focus then some thought should be given to having research in the tech tree not automatically take all the science packs up to X? Why can't some later research only take purple or yellow for example? Why not have different branches of infinite research that takes only X and white? That would actually make for some interesting choices where you build one part of your science facility or the other out depending on what you want to focus on, rather than automatically having to have balanced capacity in red/green/blue no matter what you do.

[featherwinglove]

As someone who had missed that titbit of info in FFF#266, it does seem a bit silly to remove the AM input limits.

I agree, but that can be modded back in. Officially, the axe is never coming back even with mods. Mining hardness is a mod-soluble problem, but it mostly affects modders and mod players only, and the former have a workload to fix. Top modder put it

That's a lot to dump on me from just one FFF... It's probably more than everything else previously in 0.17.

Anyway, let's keep discussion regarding that FFF under that FFF, thanks.

I agree with the requests to make uranium useful for something. I was thinking with the satellite no longer required to win the game, why not make the satellite take uranium fuel cell(s)? That way uranium becomes required for white science while the randomness of uranium processing doesn't become a factor for people trying to speed run the game.

Not "the satellite" exactly. Real life space probes getting too far from the sun (including the surface of Mars in three cases, but excluding the orbit of Jupiter in one) tend to use 238-Plutonium radioisotope thermoelectric generators or RTGs, and actually nuclear powered radar satellites launched by the Soviet Union, i.e. c/w with critical nuclear reactors - one of them crashed in Canada and caused an international incident.

I'm not sure the names for the packs really make sense

Neither am I, but I don't understand it well enough to really articulate the problem.

[Sarkazeoh]

Reusable Rockets

I feel like this is a missing tool in the game.

https://www.reddit.com/r/factorio/comme ... it-android

[featherwinglove]

Reusable Rockets

I feel like this is a missing tool in the game.

https://www.reddit.com/r/factorio/comme ... it-android

Which is that mod I found? (flip flip) Ah, there it is. I don't think it's necessary to have reusable launch vehicles in vanilla, but I won't object if they put them in.

[AndrewIRL]

Bold changes, looks impressive!

[LoSboccacc]

So people will win accidentally if they ignore the existence/recipe for space science?

Doesn't seem a solid choice, at that point better not to have a victory screen at all

We need space research to unlock the component of an escape ship, that's what would close the game properly

[Engimage]

Overall awesome stuff as always.
I will point out several concerns I got about it all.

• Blue science pack recipe still takes unnecessarily too long compared to all other packs. This does not impact resource consumption but just results in huge lines
• Space science pack still requires military in research tree. Not sure if this fits your strategy with peaceful path
• Damage refactoring requires rebalancing of namely gun turrets as they were viable only cause of two stacking upgrades
• I do not think it’s a good idea to make all beams of the same color. I would prefer being able to differ them one from another
• why not to follow same strategy and introduce energy weapon fire speed which would effect both drones and turrets and balance accordingly? I would go for it
• Should definitely introduce personal equipment connector to base power to make early to mid game much more viable as well as personal accumulators. PSPs are just too bad

[Ferlonas]

The role of the rulemakers in an open forum is, mostly, to say "no, this is not up for discussion at this moment" or "yeah, that's worth thinking about". That doesn't make it less frustrating when the answer is no, but I'm happier if they spend their more of time making the game better and less of it sugar-coating their nos.

That's exactly the point.
Wube does a great job presenting their changes, ideas and thoughts. They haven't to do it, but they do because they want to discuss with the community.

And that is exactly my point. They do a great job presenting it, but then at the first attempt to discuss something that felt wrong to me, I get shot down with one sentence and ignored. How is that for discussion in your eye? This is not a case of forum moderation saying "This is not up for discussion", because if it wasn't up for discussion, then we shouldn't have this thread at all. "Discussion" doesn't mean they have to change it. I know I get absolutely no say in that, but at least I want confirmation that they have read and tried to understand what I'm saying, instead of just saying "No, it's fine that way, here, have a proposal that will make what's bothering you even worse".

That goes for other people as well by the way, because this discussion (and I presume I'm speaking for featherwinglove here as well - sorry if I don't) is not about getting the devs to do what we want, but to understand why they chose to make a specific change, and to see if there are alternatives to be found that might fit the situation even better. Newsflash: software developers are not all-knowing (unfortunately, as I'm one myself). Assuming that just because a developer thinks about a problem means he finds the best solution for it is a mistake, because he will most likely just find the solution he personally thinks is best. That might be the most appropriate one, but it's also possible that he can use another angle. (Sometimes a rubber duck will suffice, sometimes it won't: https://rubberduckdebugging.com/)
So don't shut down someone trying to understand it by saying "Let the devs do their job" or something along those lines. If that's all you have to say, you will not further a productive discussion but instead a heated debate on everything but the original topic.

And this is seriously off topic, so let's please get back to a productive discussion about the FFF and not about our interpretations of how software development should work - open or closed, beta or released.

Regards
Ferlonas

[bman212121]

All in all I don't really like the module + beacon meta the game pushes on you. The builds all look the same with it. But even on somewhat smaller scales it's absolutely required. Take a calculator and look at the numbers for low density structures for example. Without beacons it's absolutely insane.
Plastic and oil in general is another thing that benefits immensely from modules. Not just in resources consumed, but just the plants you need.

No, not it's not. I've heard this point made many times now that the game forces you to do something. The game will force you to make prod 1's now that they are a science requirement. The game has absolutely no requirement for you to ever use beacons. I don't use beacons, and I don't think they are all that valuable. They eat a ton of power and give little bonuses. The only way people have figured out how to make the most of them is to make those grid patterns that everyone complains about because they are boring. Outside of the "mega base" builds, you don't ever need to use beacons. The only reason why you have to use beacons in those cases is because your computer isn't fast enough to handle the game. As we get faster memory you'll be able to build those same "mega bases" without the use of beacons.

There is actually kind of a oxymoron going on in the thought process. You want to build a "mega base" that consumes massive amount of resources, but you're also worried about all of the resources you need to be able to consume and the number of plants. You can take a step back and realize that the reason why you're being forced down one path is because of the constraints you put on yourself. If you are choosing to use the least number of factories and squeak the most amount of efficiency out of them, then there are already hundreds of people who have optimized that and you can just copy / paste their blueprints. If you want a real challenge then come up with new criteria to meet, and you'll find that things work out differently. Factorio is a game, and a game is meant to be fun. I think it might be more fun to see how poorly I can make use of resources, and make the biggest cloud of pollution possible. If my goal is to turn resources into pollution now all of the sudden I'll need completely different designs.

[bman212121]

The FFF has a ton of info on it. For the most part I don't think that I'm compelled one way or another about the changes. There is so much to digest at this point I'd certainty like to see it in action. The only parts I can comment on is that I don't think we need a pipe requirement for Logistic science pack. The science pack is already showing you how to automate things, and simply adding one more item to automate doesn't really add value. I think it's more of a moment for the player to realize they need to automate pipes because setting up oil refineries certainly needs to have that to make the game go smoother. Players won't realize this at first, but it's that process of keep building pipes that will irritate you enough to where the light bulb goes off. That said, pipes build so fast that you kind of ignore it until you need to build underground pipes. Those parts are the ones you'll get fed up with enough that you'll go why don't I build these in factories? I think it's better that other game mechanisms can also trigger production chains, otherwise you're basically having to setup every production chain in the game just to get research done.

I'd agree with a few other people that the rocket should simply not launch without something in it. Unless I'm mistaken, if you launch a rocket without a payload, the only thing that occurs is the counter goes up by 1. There's no real benefit to having this occur, so it does make more sense that you can't launch the rocket at all without something in it.

One of my biggest gripes with previous changes is easily the logistics system placement. I'm very curious to see how this change plays out now. You need green science to get the red circuits for the chests and the robots, so it's still a bit deep in the system. But if you don't have any prerequisites that require production packs at all, then it should be a bit better than before. If it's one of the first things you can craft for yellow science, then you might be able to still focus directly on getting that requirement and not have to setup 10 production chains just to get it. It seems like this might be a decent compromise now to keep it buried a bit, but still allow people to reach it if they really need it. Once you have it then you can at least use it to scale production if that's the route you want to take. Much better than forcing you to build everything to scale production before you get the item you needed to scale production.

Overall I think I'm going to enjoy these changes, and it should be fun to make some recipe changes to accommodate the new science. I like the replacement of gun turrets with walls instead, as it helps cut down some of the iron requirements. I use walls so it's something that flows well. Obviously there will be other people who don't use walls, so it's just going to be one of those things you'll just need to make regardless. But right now that already occurs as I'm sure a lot of people don't use gun turrets, so it's just changing one item for another. But I definitely like the larger inclusion of stone.

[featherwinglove]

That goes for other people as well by the way, because this discussion (and I presume I'm speaking for featherwinglove here as well - sorry if I don't)

Aside from the fact that I wasn't previously aware of the rubber duck debugging site (how old is the concept? It occurs to me that it may have inspired the pizdek *ahem* psyduck pokemon in about 1998), we are in perfect agreement on the matter.

It took some dragging, kicking, and screaming, but there is now a productive thread about modding axes back into 0.17.

All in all I don't really like the module + beacon meta the game pushes on you. The builds all look the same with it. But even on somewhat smaller scales it's absolutely required. Take a calculator and look at the numbers for low density structures for example. Without beacons it's absolutely insane.
Plastic and oil in general is another thing that benefits immensely from modules. Not just in resources consumed, but just the plants you need.

No, not it's not. I've heard this point made many times now that the game forces you to do something.

"forcing" and "pushing" are not synonymous IMHO. When designing my modpacks, I find ways to "push" practices I want to use, gameplay mechanical motivations which are optional (and when it comes to Factorio, very little is truly forced - there are kid mode mods, and I'm guessing that a mod which turns 0.17 science back to 0.16 recipes is probably going to be one of the first 0.17 mods to come out a handful of hours after the first experimental drops.) The beacon push is pretty mild, especially if you have lots of real estate, modest megabase ambitions, and plentiful resources: speed runners beacon only the rocket silo.

[Serenity]

I'm not obsessed with UPS like so many people. And I don't care for theoretical nonsense like "the map is infinite". Just an example, to make one yellow belt of low density structures you need 320 AM3. With beacons and modules you cut that down to just 52 machines. That's an enormous saving. Sure you have the space in theory to build as big as you want, but you then need to actually build all that stuff. A lay down belts or trains. It looks cool, but even with robots it's just a lot more work.
And don't say you don't need really a red belt of low densities or whatever. Other products show similar results where the beaconed build is a lot more manageable

It's not a question of the game literally forcing you to use beacons. Of course it doesn't. If you have tons of time you can build a huge base without them. Some people have just for the sake of it. But using them is way more attractive than the alternative. And not just for the usual overly simplistic "It's all about UPS"

The game is certainly designed to get you to use beacons, even if you don't. Modules are also meant to be a resource sink.

[Matthias_Wlkp]

I'm cautiously optimistic about all these changes.

I add to the voices that Chemical Science (formerly known as blue) takes too much time. Compared to others tier above, it gives one less beaker for roughly the same time. This means I will need 50% more assemblers for this particular science pack type compared to higher tiers. I guess that can be interesting, but it adds to the progress challenge. This science pack still adds to the very sharp steepness of the learning curve. It's very easy to automate green and red science compared to the blue. I'm a casual player and this hits me hard every time I play. I don't have any remedy for it, maybe it's the way it has to be. For sure removing the mining drill from the recipe helps. My overall impression is, that as soon as you automate blue science, it's downhill from there. If I can force myself through that, everything else becomes easy.

As for military rebalance, I always used the military techs as a "pause" for meaningful research. I typically have 4 research centers and I'm so slow, that need to put the research somewhere, before I automate the f*** blue science... I only hope I will still have something to burn red/green on, before I finally get the blue going.

I would also welcome some sort of upgrade for the gut turrets, but I realized that uranium ammo should help with that (I actually never tried it). If it's not good enough, that means uranium ammo needs a buff.

[bman212121]

I'm not obsessed with UPS like so many people. And I don't care for theoretical nonsense like "the map is infinite". Just an example, to make one yellow belt of low density structures you need 320 AM3. With beacons and modules you cut that down to just 52 machines. That's an enormous saving. Sure you have the space in theory to build as big as you want, but you then need to actually build all that stuff. A lay down belts or trains. It looks cool, but even with robots it's just a lot more work.
And don't say you don't need really a red belt of low densities or whatever. Other products show similar results where the beaconed build is a lot more manageable

It's not a question of the game literally forcing you to use beacons. Of course it doesn't. If you have tons of time you can build a huge base without them. Some people have just for the sake of it. But using them is way more attractive than the alternative. And not just for the usual overly simplistic "It's all about UPS"

The game is certainly designed to get you to use beacons, even if you don't. Modules are also meant to be a resource sink.

Yes beacons do make the number of plants required smaller, at the expense of power. So yes beacons are a key feature if you are trying to reduce the number of buildings, or space consumed. If those are your constraints you want to work in then I completely agree that beacons are a key part of that. I would imagine that you'd be using nuclear power rather than solar, otherwise the amount of space you saved by using beacons goes right out the window by all of the extra space required for solar. There's absolutely nothing wrong with that, and if that's how you like to play then by all means do so. For me I actually like seeing massive rows of factories, and just brute forcing the problem. So it doesn't bother me one bit that I need to plop down 320 AM3s if my goal was to saturate a yellow belt. That simply comes with the territory of what I'm trying to do. The cross roads to that thought is where the question lies. Why stop at 1 yellow belt of low density structures? Is that your criteria for "it's fast enough"? If you could make the game more manageable by only using 52 AM3s, then wouldn't the game become even more manageable at only 26 AM3s? Do you really need to supply the rocket silo with that many resources, or can you do what you're trying to do with less? I think we used like 30 or so AM3s with prod 3s in them to get the "there is no spoon" achievement, and that didn't require any use of beacons at all.

[Hiladdar]

Overall I am impressed with the amount of effort put into the redesign of sciences! I am looking at it's implementation. I do have a few concerns and recommendations.

My first concern is that some non-intermediate products are now used as intermediate products used to create the final science packs. My recommendation is to consider tagging those items as intermediate products.

My second concern, is when I look at the cost in raw resources and assembly time in the FF 275 spreadsheet: https://docs.google.com/spreadsheets/d/ ... 2039786206, as it is listed on 30 Dec 18.

The cost for raw iron goes as follows from blue - purple - yellow is 24 to 157 to 99. I am OK with the other raw resources for copper, stone, coal. Consider the raw iron requirement to be in the 40 to 100 to 140 range as it progresses from blue to purple to yellow I would also consider lowering the time required to produce a purple science from 160 to about 110 - 125 range.

Since this Friday Fact covers research including post satellite launch research, consider adding introducing several other lines of research, such as pipe and pump pressurization allowing for greater liquid throughput through pipes. Also, belt, inserter, splitter speed, for greater throughput of dry materials. Finally, consider introducing an increase in engine optimization for all vehicles within the game increasing the top speed, acceleration, and turning radius of vehicles, for car, tank, and locomotive, but also vehicles from mods. The vehicle mods can be handled similar to how train breaking is handled in .16. Also, consider breaking research to also affect cars, tanks and other vehicles introduced via mods. These new lines of research do make sense given that this is a game where the focus is not on a bug hunts, but construction and logistics.

One raw resource which is not used for any science is uranium. The one item where it does make sense to use it in assembling a satellite, by changing the recipe for the satellite to include either a nuclear fuel cell, or nuclear fuel.

Hiladdar