Friday Facts #275 - 0.17 Science changes

[lacika2000]

...
The biggest problem other than the theme problem (I'm pretty sure new players are going to see this and just go "What the hell is this?" like I did; it's immersion breaking), is the fact that so many rails are needed that there is no way to share it with a mall, even for new players with small factories. The point of nudging players to automate what they need in the phase of the game that they are in is completely destroyed by this, and I thought this was the whole point, the theme be damned to hell.

It isn't as big a deal as the axe, but the logic here is still completely lost on me. Rythe might say "Rails tick the right boxes per them," but I really don't see how. The numbers and direct handoff strategy made necessary by them really seems to contradict the concept of encouraging the automation of items that need to be automated anyway - I'm pretty sure there's a quote to that effect, possibly on this thread, but I'm having trouble finding it.

I am with you on this 100%.

[V453000]

...
A lot of the complaints I see about the amount of rails is generally about:
-A: I don't like to belt this much.
...
-reA: As said many times earlier, belts are completely fine enough so it's entirely viable to do that, in fact a single yellow belt can get you very far, and doubling it is very easy as Logistics 2 are mandatory pre-requisite for rails so you sure have them available.
...

Thanks for answering, appreciate your time!

Let me try then from a different angle: having 30 rails as an input to this recipe is the same deviation as having 30 copper wire in the current yellow science recipe. It breaks away the typical logic of many basic ingredients to few intermediates to an even fewer final product(s). This is why some of us has proposed using concrete or refined concrete instead. It achieves the resource pull, but does not require the high number of ingredient input to a science pack, so it can be scaled better.

But hey, you are a dev and we are only users: so it is your call in the end.

I understand this reasoning, I'm not sure if I missed it or didn't understand.

At some point I was considering one of the versions of concrete, but the main problem is that concrete isn't really useful for anything. Sure, there are a few specific entities that you can build from it, but it's really just the rocket silo with nuclear being completely optional. Flooring is cute and definitely great to have for the movement bonus, but not something I would consider structural. The only link I could see to Production for concrete would be that it goes into centrifuges (and uranium processing is one of the topics of Production science).

Obviously the other option would be to start putting concrete in many more high tier recipes like assembling machine 3, beacons, tier 3 belts, that production science pack has something in common with.

To your original point, concrete is also a final product.

[Zavian]

Let me try then from a different angle: having 30 rails as an input to this recipe is the same deviation as having 30 copper wire in the current yellow science recipe. It breaks away the typical logic of many basic ingredients to few intermediates to an even fewer final product(s). This is why some of us has proposed using concrete or refined concrete instead. It achieves the resource pull, but does not require the high number of ingredient input to a science pack, so it can be scaled better.

1 blue belt of rails will let you build 240 science per minute. (336 with max prod modules in the science assemblers, something you should consider if you really want that many flask per minute). If you want to scale higher then surely you are choosing to accept the challenges that come with that choice?. Indeed isn't the only reason to scale higher because you want to demonstrate that you can overcome those challenges? But if you want to scale higher, there are plenty of ways to run multiple belts to a set of assemblers, or you could use multiple sets of assemblers. No recipe change needed to simplify the challenge.

[V453000]

...
The biggest problem other than the theme problem (I'm pretty sure new players are going to see this and just go "What the hell is this?" like I did; it's immersion breaking), is the fact that so many rails are needed that there is no way to share it with a mall, even for new players with small factories.

I am with you on this 100%.

The original assumption that "there is no way to share it with a mall" is utterly exaggerated. A single yellow belt of rails will support a lot of assembling machines running at full speed that even the more megalomaniac new players are never going to have.
Again, having 2 belts or upgrading the belt does not sound like "there is no way".

In my factories I had rails built in a mall part of the factory and just belted all the way to the science build. On one belt. It was enough.

[Schallfalke]

...
A lot of the complaints I see about the amount of rails is generally about:
-A: I don't like to belt this much.
...
-reA: As said many times earlier, belts are completely fine enough so it's entirely viable to do that, in fact a single yellow belt can get you very far, and doubling it is very easy as Logistics 2 are mandatory pre-requisite for rails so you sure have them available.
...

Thanks for answering, appreciate your time!

Let me try then from a different angle: having 30 rails as an input to this recipe is the same deviation as having 30 copper wire in the current yellow science recipe. It breaks away the typical logic of many basic ingredients to few intermediates to an even fewer final product(s). This is why some of us has proposed using concrete or refined concrete instead. It achieves the resource pull, but does not require the high number of ingredient input to a science pack, so it can be scaled better.

But hey, you are a dev and we are only users: so it is your call in the end.

...
At some point I was considering one of the versions of concrete, but the main problem is that concrete isn't really useful for anything. Sure, there are a few specific entities that you can build from it, but it's really just the rocket silo with nuclear being completely optional. Flooring is cute and definitely great to have for the movement bonus, but not something I would consider structural. The only link I could see to Production for concrete would be that it goes into centrifuges (and uranium processing is one of the topics of Production science).

Obviously the other option would be to start putting concrete in many more high tier recipes like assembling machine 3, beacons, tier 3 belts, that production science pack has something in common with.

To your original point, concrete is also a final product.

How about if I suggest a new intermediate product, which uses enough stone/stone brick/concrete with other stuff as ingredient?
Currently in 0.16, the rocket control unit is such an intermediate product where the only use is for the rocket part. Maybe we can do the similar thing here?
This will solve many things at the same time. Making enough stone "consumption" while also making the recipe looks more beautiful.

EDIT: About the name of such intermediate product, I can think of the term "prefabrication sections". There is a Wiki entry on "prefabrication" so I am not explaining too much here. In short, prefabrication is a common practice in construction of buildings, ships, aircrafts, machines, etc. Pre-assembled parts prepared from factories, so not everything needs to be done at the site, saving a lot of time and avoiding many issues.
In terms of buildings and civil engineering, prefabricated concrete and prefabricated steel sections are widely used. This suits the theme of "production" very well.
PS: Prefabrication is NOT limited to concrete, but can be based on other materials as well. But they are not related to the discussion here.

[Zavian]

How about if I suggest a new intermediate product, which uses enough stone/stone brick/concrete with other stuff as ingredient?
Currently in 0.16, the rocket control unit is such an intermediate product where the only use is for the rocket part. Maybe we can do the similar thing here?
This will solve many things at the same time. Making enough stone "consumption" while also making the recipe looks more beautiful.

But that destroys the variety of having recipes that need relatively large amounts of an ingredient per product. Seriously why is there such a fuss about this? It is no worse than green circuits for processing units. In fact it isn't as bad as green circuits for processing units. Processing units need 20 green circuits and have a 10 second craft time. That is 1.5 per second for AM2. (2.5 for AM3). Production science needs 30 rails and has a 21 second craft time. That is 1.07 per second for AM2. (1.78 per second for AM3). Also note that for high tech science packs you need 2 processing units per cycle, so you can effectively double those green circuit numbers.

[Schallfalke]

How about if I suggest a new intermediate product, which uses enough stone/stone brick/concrete with other stuff as ingredient?
Currently in 0.16, the rocket control unit is such an intermediate product where the only use is for the rocket part. Maybe we can do the similar thing here?
This will solve many things at the same time. Making enough stone "consumption" while also making the recipe looks more beautiful.

But that destroys the variety of having recipes that need relatively large amounts of an ingredient per product. Seriously why is there such a fuss about this? It is no worse than green circuits for processing units. In fact it isn't as bad as green circuits for processing units. Processing units need 20 green circuits and have a 10 second craft time. That is 1.5 per second for AM2. (2.5 for AM3). Production science needs 30 rails and has a 21 second craft time. That is 1.07 per second for AM2. (1.78 per second for AM3). Also note that for high tech science packs you need 2 processing units per cycle, so you can effectively double those green circuit numbers.

What does it "destroy"?
So did the intermediate product rocket control unit "destroy" the recipe of rocket part? Be reminded that it is a very simple recipe of just 1 processing unit + 1 speed module giving 1 rocket control unit. So why didn't the devs just cancelled this recipe, but just had processing unit and speed module as DIRECT ingredient at the rocket silo?

The point of these intermediate products, is to add reasonable steps between raw ingredients and very complex final products, right? Processing unit or speed module look very "basic" components to a huge and complex rocket, so the devs added the rocket control unit.
Same thing here, rails look very basic components, compared to a late-game production science packs, and I think that's what many forum users tried to point out here.
BTW, if you need such large quantity (30) compared to other ingredients, it is a bulk ingredient but not an "advanced" enough ingredient to qualify for a late-game science pack, right? (IMO, it is no much better than the 30 copper cable for the hi-tech SP recipe anyway.)

EDIT: A minor point why rails do not suit here. Rails are logistics item (see which category do they fall in game?). I would not be surprised if they appear in recipe of an "advanced logistics science pack", if it exists. But for "production" science pack, hmm..., not fitting nice IMO.
That's why I proposed the new intermediate product.

[Avezo]

I just realized I had mistaken production with utility science packs earlier lol... Now I don't think that concrete is that good for 'production' science pack instead of rails, but I still think it would fit nicely into theme of 'utility' science pack.

[Rythe]

The original assumption that "there is no way to share it with a mall" is utterly exaggerated. A single yellow belt of rails will support a lot of assembling machines running at full speed that even the more megalomaniac new players are never going to have.
Again, having 2 belts or upgrading the belt does not sound like "there is no way".

In my factories I had rails built in a mall part of the factory and just belted all the way to the science build. On one belt. It was enough.

Waitwaitwaitwait... Is this the part where I learn everyone's been talking out the side of their neck on the aspect I didn't care enough about to figure out myself?

And I mean both sides here.

So at 1x speed, 1 assembly can almost produce enough rail for 3 production pack assemblies, and red or blue belt w/ quick inserters is more than fast enough to feed those almost 3 production pack assemblies?

What's the novelty/challenge here again? That you get to see an item zip by on the belts faster than usual? That it'd be neat to have a single rail assembly direct insert into 2 production pack assemblies? Why is the high throughput item side even worth mentioning here?

[Tricorius]

The original assumption that "there is no way to share it with a mall" is utterly exaggerated.

So, my diatribe on games development and realigning the “push to 1.0” was meant to point this out...but I probably didn’t do a great job of it.

As another said in previous pages, if I look back on my first couple of factories I didn’t build with a separate “goods” array (sorry, I dislike the term “mall”) and science areas.

In fact, most of what I did was either leech off production lines into chests (I actually still do this in early- to mid-game for a few things...but I use circuit conditions to be a bit fancier). I very much dislike hand-crafting things I know I’ll need moderate quantities of. Even then, my standard “rails” setup is a single blue assembler, set to a stack-limited chest. Generally with the top row available. So I think that gives me 2000, but I don’t exactly recall. What I do recall is thinking several times (even now) that those rails fill up that chest surprisingly quickly.

I feel like most new players likely still do that. There is a huge learning curve to Factorio (another thing I tried to point out). My current play style is nowhere near my first few factories. In fact, I realized I only launched a rocket in perhaps 2 of my first 5 or so factories. I started over a few times when I felt like I could do things a whole lot better.

I feel like new players are likely to not understand the factory demands for science. A likely consequence is that they don’t have a perfect ratio science setup, nor the sustainable raw production to support one.

This means science is going either be very slow (resource starved) or unbalanced (way too much iron, not enough copper). Either way, this leads two natural thoughts: scoop up a bunch of stuff on the line, or leech stuff you need into chests. I think the first time one would do this is for belts and inserters.

So I think it is a very natural thought to get to this science that needs a metric crap-ton of rails and do exactly what the devs assume: either leech rails from science (if it is uneven) or build a separate area for rails.

I think a “mall” is a pretty advanced realization.

Even a lot of “starter bases” use a fairly compact design with built-in “buffer stations”. A normal, pretty easy first shot at this is to just stick a buffer chest between the output “inserter” and the belt. Extra points if one realizes they can stick a green/red wire on the inserter between the chest and belt and set it to only enable when rails > 400.

I don’t think the typical new/casual player is going to be too upset by this realization. Oh well, my science will stall for a minute while it refills my rails now I want to try to build a train to that patch of ore clear up there. (And they are going to spend a lot of time getting that first (or tenth) rail line functional.)

An argument here would be that the chest might not fill properly after you take a bunch of rails out, if you don’t have a circuit condition. This is true...but then you just temporarily turn (or remove) the inserter if you have immediate need of more rails. (This cleanly transitions into “well this is a pain in the butt, there has to be a way to automate this..,then one hopefully finds the circuit network.)

The other, pretty easy first shot is to know you can “vacuum” up stuff on a belt. I think it is pretty easy to build an extra long belt which contains a buffered amount of a product you might eventually want. The game teaches you this pretty early when you run a bunch of iron out of a smelter. I’m guessing a lot of newer players vacuum up metal plates, gears, pipes, or circuits early on.

I certainly don’t think it is unlikely to be able to get an early, low-experience version of this science up that still allows you to grab a bunch of rails when you need them.

I think it takes quite a bit of Factorio experience to realize a full-throttle science setup. And before that, science is unsteady enough to easily have a leechable buildup of various goods.

[Zavian]

What does it "destroy"?
So did the intermediate product rocket control unit "destroy" the recipe of rocket part? Be reminded that it is a very simple recipe of just 1 processing unit + 1 speed module giving 1 rocket control unit. So why didn't the devs just cancelled this recipe, but just had processing unit and speed module as DIRECT ingredient at the rocket silo?

You misunderstand what I meant.

The new 0.17 production science recipe needs 30 rails per cycle. (Do note that that is only 10 rails per pack, and that since the cycle time is so long it is only 1.07 rails per second per assembler 2 making production science). Keeping this recipe adds variety to the different recipes in the game. Most recipes of them are 1-2 of something, and 1-3 of something else. This recipe is distinctly different. I view that as a good thing that adds interest to the game.

You want to

making the recipe looks more beautiful.

by removing the 30 rails and replacing them with 1-3 of something different. Why should the devs make another recipe that looks like that? Should all recipes be homogeneous? To me that is destroying what is unique and different about this recipe. (If you have been reading this thread, I've already said that I felt that the rocket control recipe was too simple and boring. "Rocket control units ... Speed modules + processing units. Again only 2 ingredients. Both of which are already automated for science. Only one step. Again too simple for this stage of the game".

[Serenity]

I certainly don’t think it is unlikely to be able to get an early, low-experience version of this science up that still allows you to grab a bunch of rails when you need them.

I think it takes quite a bit of Factorio experience to realize a full-throttle science setup. And before that, science is unsteady enough to easily have a leechable buildup of various goods.

Science doesn't necessarily run all the time until you get into repeatables. For much of the game research is way ahead of what you can implement and can idle for a while or be a bit slower.

You also won't be constantly laying down thousands of rail. Leeching off production lines is ok (up to a point) for things that you only need a little of now and then.

[Tricorius]

I certainly don’t think it is unlikely to be able to get an early, low-experience version of this science up that still allows you to grab a bunch of rails when you need them.

I think it takes quite a bit of Factorio experience to realize a full-throttle science setup. And before that, science is unsteady enough to easily have a leechable buildup of various goods.

Science doesn't necessarily run all the time until you get into repeatables. For much of the game research is way ahead of what you can implement and can idle for a while or be a bit slower.

You also won't be constantly laying down thousands of rail. Leeching off production lines is ok (up to a point) for things that you only need a little of now and then.

Exactly. A new/casual player is unlikely to calculate a ratio-perfect science setup and be able to feed it enough resources.

And it takes a while to learn to do pretty much anything in Factorio. An experienced player will have an idea of how to setup a “mall” and “science” area that don’t compete for incoming resources.

A new player will have unbalanced input, or be doing so much other stuff that science is idling a lot of time. And if you need a bunch of anything going into science a simple right click removes a feeder belt and allows a quick pile-up of whatever. Even if it doesn’t occur to leech into a chest or with the ‘F’ key.

[Tricorius]

The new 0.17 production science recipe needs 30 rails per cycle. (Do note that that is only 10 rails per pack, and that since the cycle time is so long it is only 1.07 rails per second per assembler 2 making production science).

Also, unless I’m not remembering correctly, or the recipe for rails changed, a single assembler makes 2x rail per half second. 4 rails per second. (I don’t care to adjust for the 0.75 blue assembler speed—I really don’t care, nor do I care while playing—I assume the simpler 1 second average—but I know it is higher than 1.07 ).

Regardless, the assembler speed is removed from the equation if you are running blue assemblers for the rails and the science packs. They both run at 0.75.

[jodokus31]

Let me try then from a different angle: having 30 rails as an input to this recipe is the same deviation as having 30 copper wire in the current yellow science recipe. It breaks away the typical logic of many basic ingredients to few intermediates to an even fewer final product(s). This is why some of us has proposed using concrete or refined concrete instead. It achieves the resource pull, but does not require the high number of ingredient input to a science pack, so it can be scaled better.

I don't see anything wrong to have that kind of deviation. I would wish, there was more deviation in the recipes. This also makes scaling more interesting.
If i think about, what happens in angel mods including loops and byproducts... but i also understand, that its not so feasible in base game.


I also think, that the rail-recipe itself (not the product) is a nice one: more advanced than other high end recipes and it includes an iron stick .

[Zavian]

I certainly don’t think it is unlikely to be able to get an early, low-experience version of this science up that still allows you to grab a bunch of rails when you need them.

I think it takes quite a bit of Factorio experience to realize a full-throttle science setup. And before that, science is unsteady enough to easily have a leechable buildup of various goods.

Science doesn't necessarily run all the time until you get into repeatables. For much of the game research is way ahead of what you can implement and can idle for a while or be a bit slower.

You also won't be constantly laying down thousands of rail. Leeching off production lines is ok (up to a point) for things that you only need a little of now and then.

Personally I rarely build a mall, and often either leech off science, or tack another non-science build onto the end of a science build with similar ingredients.

eg green science's belt and inserter assemblers both go into a chest, and then onto the belt for science production. (But one assembler of each is more than I need for science anyway, so those assemblers have excess production). I often have a chest of engines for trains/pumps/cars/tanks that is leeching from science.

I often end up building robot frames/solar panels/accumulator/laser turrets by leeching electric engines + steel + green circuits + batteries + copper by extending some of the belts from my science production for blue, purple and yellow science into a new production line located behind science.

I fully expect to either use one of the science rail assemblers to feed a chest of rails, or to add another assembler to the same iron/steel/stone belts to supply me with rails.

[Tricorius]

Personally I rarely build a mall, and often either leech off science, or tack another non-science build onto the end of a science build with similar ingredients.

I usually don’t build my “mall” section until I get to the point where I want to start building out my construction trains. This generally happens during late mid-game when I need to start a good rail network with larger, multi train stations.

Until then I leech. In fact, my latest holidays death world I didn’t even build my usual “starter factory” and it worked fine. Honestly, the “mall” is almost always built to slowly buffer a bunch of common goods over time anyway. So it is often very idle. In fact, I was recently trying to decide whether to add circuitry conditions to flip a power switch when the “mall” was full. Thus saving a reasonable amount of idle power.

[Avezo]

To chime into few above posts - when I've played for the first time (if my memory serves well), yellow inserters being part of science pack actually made it harder to use them for me - I wouldn't touch inserters that had automated production, because they were meant for science, so surely game didn't intend me to pick them away from there, and, I wouldn't set up separate automated production for them, because that would be a waste since I already had one assembler producing them. I ended up handcrafting much more of them that I ever should've.

I've learned a lot since then, but still I think putting any final product in science recipe makes it actually harder for new players - instead of hinting 'automate my production', it hints 'my production belongs to science packs, don't touch me'.

[Serenity]

For 14 yellow science assemblers you only need 6.67 electric engine assemblers. That leaves about a third of an assembler making engines for you even at full speed :p

[featherwinglove]

The original assumption that "there is no way to share it with a mall" is utterly exaggerated. A single yellow belt of rails will support a lot of assembling machines running at full speed that even the more megalomaniac new players are never going to have.
Again, having 2 belts or upgrading the belt does not sound like "there is no way".

In my factories I had rails built in a mall part of the factory and just belted all the way to the science build. On one belt. It was enough.

...default map settings? I'm asking because when I get into building rail networks, the demand for rail components rapidly gets beyond the point where sharing it with a science build which needs so many rails is going to be feasible. A yellow belt only practically feeds about a dozen AM2s, and in this situation with low machine ratios, using belt doesn't save any inserters, so direct handoff saves belt. Unlike with furnaces, assembly machines don't internally buffer much, making leeching via character interactions impractical, so I would still need other rail assemblers and derp, that's a separate line. If I belt buffer, the buffer is going to be either really tiny, or really impractical, and chests are cheaper, but an inline chest buffer doesn't work with this level of item throughput, and again, separate line. Does anyone ever get enough rails by 'F'ing a belt? I would be surprised. (With SL2, I can usually get enough inserters from the SP2 system belt, but conveyors are a different story. I tend to have a separate belt-making pair of machines - often one I've made before automating SL2 - just to make conveyor for construction.)

Exactly. A new/casual player is unlikely to calculate a ratio-perfect science setup and be able to feed it enough resources.

And it takes a while to learn to do pretty much anything in Factorio. An experienced player will have an idea of how to setup a “mall” and “science” area that don’t compete for incoming resources.

A new player will have unbalanced input, or be doing so much other stuff that science is idling a lot of time. And if you need a bunch of anything going into science a simple right click removes a feeder belt and allows a quick pile-up of whatever. Even if it doesn’t occur to leech into a chest or with the ‘F’ key.

I think it's a really bad assumption to assume only experienced players use math. I came into Factorio off freaking Kerbal Space Program! Joseph Anderson knows Factorio appeals to the mathematically inclined from the beginning. The only thing stopping such players from using perfect machine ratios is that it's difficult to keep up with a factory big enough to feature such ratios at a learning pace, and that will not be happening with the rails in this science pack! Back when SP3 was steel, battery, red circuit, and smart inserter (now filter inserter), the only item that wasn't near machine ratio perfect in my first game was the inserter line. And that's only because 0.5s crafting time. I think I had 12 labs? I can't remember, but it wasn't very many, that's for sure.