Friday Facts #275 - 0.17 Science changes

[featherwinglove]

I also agree that the sheer number of rails is huge, though I would say that *is* kinda consistently thematic for rails. I know I always significantly underestimate how many rails I will need for an expansion, even after having done many, many, many of them.

I wouldn't mind a lot of rails in a science recipe where the rails were thematically matched, say a "transportation" pack, but the rails should be cheap compared to the sum of the rest of the ingredients in the pack. In real rail transportation systems, and related experiments, rails don't wear out as fast as everything else in the system.

To the point where my end game solution is very likely the same as many others. I have a separate 4-car train with the sole purpose of expanding the rail network. Two (2) of those cars are completely full of rail.

FARL much?

But developers have to eat too. If not, they have to find a boring, soulless job which diverts much (most? all?) of their energy from making really awesome games. They have already made a lot of money, and I’m guessing they are doing fine, but this *is* the drive to 1.0.

There's a part of my mind that wants to mistake this for a commentary on the rest of the video games industry

And a lot of negative reviews that say “it was fun for a few hours, but then I couldn’t figure out how to get past that stupid purple science” isn’t going to look great for ongoing sales.

Think of who you would want to ask things regarding Factorio if you're new around here: Would it be the casual guy or the new guy, or would it be the players who have already been figuring things out and answering?

Again, agreed. But consider there is a flip-side to that. I have seen people come to these forums with a very simple “sticking point” question.

And half of the responses are something like “oh that is easy, but you’re stupid for building your entire base like that ... what you really want to do is build a production bus with jargon jargon jargon”.

And guess who tends to respond that way? We know we are being helpful, but what they really need is someone who hasn’t forgotten what it is like to be new

I recall a couple of Enjin Minecraft forums where Minecraft servers designated a "helper" class who were good at introducing new players to the server and to Minecraft in general. So I think that's part of the solution. While I remember my first freeplay game in some considerable detail, the frustrations that I liked (good gameplay progression) and didn't (little other than pipe mixing, TGIF), I haven't played enough vanilla since 0.12 to feel qualified as a designated helper. I don't have the currency.

The other part, and this is going to sound unfortunate, is that there are some people who will come to Factorio, but Factorio isn't really for them. I think a major part of the reason for the campaign overhaul, and to a lesser extent the pickaxe removal, is to warn such players off while they're still in the demo. Some will (hopefully) buy it anyway even being aware of this because they see an honest developer who deserves their support. I've done this with, see if I remember, I think Isomer (name of the game), Code Hatch (name of the developer), and Lunar Flight (name of the game, and the last purchase I ever made via Steam. ...actually, I kinda screwed up with Lunar Flight and learned rather the hard way that not everything Scott Manley likes is something I'm going to like.) Edits: I forgot to mention that these are games that I've purchased, but played very little and didn't like much. ...and somehow I left an unmarked quote block that I wasn't intending to use and got really lucky in noticing it.

[gleard]

When I first read about escape pod, I hopefully thought that we (players) would have to first launch a satellite-loaded rocket to get some science data represented by Space SPs (some measurments from space or just telemetry of the launch itself), then process this data in the research to design the escape pod that would allow to carry a living character in such a rocket (possibly depending on power armor as the only tech that deals with the character's body?), and then create this escape pod and launch the character into space, at which point it seems appropriate to have a choice dialog like the one quoted here ("Take me home, captain" / "This is my home now").
This way, processing of Space SPs is necessary to win the game. IMHO, that's good, because atm such a large part of the game appears only after you "win" it, and also Space SPs feel somewhat disjoint from the story.
It seems from this forum thread that my PoV is not very popular (most people who dislike this change seem to only object against win-by-empty-rocket part), so I wrote this to add some representation to the idea that

[Needing] space research to unlock the component of an escape ship <...> would close the game properly

~~~
Also, it seems strange that the Space SP tech does require Military SPs, while the Rocket Silo tech doesn't. If for some reason one and only one of these two is to use Military SPs, it would probably be more natural the other way around (I understand that it was probably done to make it possible to win without a single Military SP, but then - are they really needed for space science?).

[lacika2000]

Many good changes to the science packs, so I am with you on quite a number of things. It is great to see that deep thinking is dedicated to this: thank you for sharing not just the outcome but your and the dev team's thoughts on this subject!

My only main concern is the number of rails required: it just does not seem scalable to me.

I have tried to explain both in the article and here earlier that the “are rails advanced enough” is not just whether the recipe and item unlock consist of early parts (technically it is green tier without oil which is early). It’s also important when do you actually want to use the item - you can start setting up rails early, but there is never too late for that, and at any point adding in the game the rail network is almost certainly allow your factory to scale like crazy (and scale = production). That’s why I think the rail is a fitting ingredient into a science pack all about production.

Yes, this is true, and usually I have the rail assembly section around the perimeter of the main base, so I can access the rails easily when building the rail network and expanding for resources.

I would not mind the set up a dedicated production line of rails for a science pack and integrate the rail production to my main bus or set up a separate section for it, but the number of rails required just don't scale enough to allow a belt-based approach vs. robots or direct insertion. Ultimately, at higher science pack rates (anything above 50 per sec), direct insertion is the only viable option, which defies your argument on setting rail production up the first place for expansion anyway.

Why it is so many rails I also tried to explain. It’s not meant to be trivial and a full yellow belt of rails can get you 1,33 science per second - that’s a very decent amount for normal factories. Not to mention that Logistics 2 are a mandatory prerequisite for unlocking the science pack, so you can get double the amount quite easily.
Sure, if you want hundreds or thousands of science packs per minute, it gets a bit more complicated. Getting that much production is a bit more complicated in many other parts of the factory so I believe that is ok.

So if I want to have 100 packs per second, I will need 75 or so yellow belt, or 25 blue belt feeding this complex... and that production rate is not even close to some of the hard-core fans' ambition level. I am pretty sure you see the outrageous nature of a belt-based approach this recipe is forcing (or, actually, how it makes it impossible).

If I remember correctly, the dev team would like us to use more belt-based designs, so here is my plea: please don't make this recipe break this approach.

And please remember, I don't mind the rails (actually like the idea of a stone based item requirement), but I am very much concerned by the number of rails required in the recipe. There is no other recipe with such high item count to be delivered to the assemblers. Please reconsider it...

[morsk]

And please remember, I don't mind the rails (actually like the idea of a stone based item requirement), but I am very much concerned by the number of rails required in the recipe. There is no other recipe with such high item count to be delivered to the assemblers. Please reconsider it...

We need to remember that the recipe makes 3x at once, so it's asking for 10 rails / science pack. It's high but not outrageous. The rail recipe itself produces 2x, so it's asking for 5 crafts of the rail recipe, for each science pack.

The 30/21 has a nice coincidence it would be unfortunate to lose. It becomes 30/30 when under speed penalty from 2 prod modules, which is the first way everyone's going to build it. (Prod 1's are always available, assembler 3s never are.) It's so often that prod modules ruin nice ratios, but here's one case where they make it work out nicer, and it's fitting for a recipe intended to encourage the modules.

[Rythe]

Aw. Too late to neatly summarize the things everyone likes to gloss over and would ruin the easy poses and banter. But it's been...interesting watching the forum dynamic of slightly different groups of people relitigating the same problem over and over every few pages. By crossing the same ol' inane ground so often too. I have slightly more sympathy for the mods and devs.

But really, featherwinglove, you're ill disciplined and slightly too fond of yourself (related), which makes you easy and delightful pickings, which makes you less effective than one might hope at getting your points across. Of course, if the banter is the prize, then...carry on?

You don't need pipes in the logistic recipe because they're part of the engine recipe and that's fed into the Chem pack. Redundant.

[featherwinglove]

I would not mind the set up a dedicated production line of rails for a science pack and integrate the rail production to my main bus or set up a separate section for it, but the number of rails required just don't scale enough to allow a belt-based approach vs. robots or direct insertion. Ultimately, at higher science pack rates (anything above 50 per sec), direct insertion is the only viable option, which defies your argument on setting rail production up the first place for expansion anyway.

There is another problem with the direct insertion that I keep forgetting to mention (and one can safely forget logistics bots as they won't be available this early): It puts a 3x3 entity next to the science assembler, which complicates getting items off the belt. It's a workable problem, but it is annoying. 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.

[featherwinglove]

Aw. Too late to neatly summarize the things everyone likes to gloss over and would ruin the easy poses and banter.

You're welcome to anyway

But it's been...interesting watching the forum dynamic of slightly different groups of people relitigating...

Please say "masticate" or "ruminate" next time. I'd rather be called a cow than a lawyer

But really, featherwinglove, you're ill disciplined and slightly too fond of yourself (related), which makes you easy and delightful pickings, which makes you less effective than one might hope at getting your points across.

I don't understand how any of that works, but oh well. I do like your writing style a lot; you have a much better ability to convey the intangibles than I do. It took me a few days just to figure out just for myself why the axe removal felt so wrong. Despite you outwriting me, I think it was worth it to put it down once I did.

Of course, if the banter is the prize, then...carry on?

It's having something that makes sense appear under a purple name. It happens every now and again, so I don't always lose.

You don't need pipes in the logistic recipe because they're part of the engine recipe and that's fed into the Chem pack. Redundant.

Wait! (Did I break his brain somehow? ...flip flip...) Umm...



Okay, so that doesn't have a pipe anywhere near it and the "Chem" pack...



It's quite a ways further out there, actually a long time after I've usually automated pipes for handcraft assisting things like boilers, steam engines, UGPs, derricks, refineries and probably two or three other mid-game things I'm not remembering.

(The images have alpha, that is so cool!)

[Rythe]

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.

But yeah, totally falls apart with rails in production packs between the throughput demand and that it's easy to need rails before you need them in production packs. So the player will quickly learn to set up rail production separate from research far before they need them for research, which means the rail feed for research will likely be a separate bit of factory than general use, so doesn't do much toward the stated goal of prompting a player to automate them.

As opposed to ...[Things] like belts and inserters, it's so easy and habit forming to just craft them as you go early on that setting up production lines for them for logistic science packs is actually a bit novel.

I generally call this 'trying too hard', that is, overvaluing the idea that it's neat and useful to use science as an automation prompter for certain items that you take it past the point of silliness.

That 'un?

But I've enjoyed plenty of what you added to the pickaxe thing. It was more that some of the back n' forth in this thread you've been in got sidetracked over things that could have been said cleaner. Consider it a nudge that I bothered with from a mix of bemusement and general appreciation.
-
Yes, there's a fair gap between chem pack and pipes, but we are still talking a feature for new players here. They won't be crafting a lot of pipes until they set up oil production as their steam plant will likely be small until then. The step from oil production to chem pack production isn't that great, so it's only a little late. Having to craft all those pipes for their first oil plant and then automating them for the chem science pack might actually be the better prompt because you introduce the need that's felt before having them drop down the solution for research purposes.

And no, that logic doesn't counter the issue with rails in production packs being a bad prompter for automating rails in general. The spread of time between rail and production packs is greater and branching into train logistics more ad-hoc as the new player branches into various things at their own pace and as they find the need. Plus rail consumption issue. Really, automating belts should be signal enough for rails, if they're going to get the signal at all.

And pipes don't have the same feel as rails/belts because the fluid system doesn't get fully introduced until oil unlocks.

[Rythe]

You don't need pipes in the logistic recipe because they're part of the engine recipe and that's fed into the Chem pack. Redundant.

Wait! (Did I break his brain somehow? ...flip flip...) Umm...

Also, also - that wasn't actually a response to anything you said. Just me really lazily tossing 2 cents at the random discussion about adding pipes to logistics in order to clear the 'talk about science changes' directive for posting.

[featherwinglove]

That 'un?

No, something they said in an FFF or a comment here, it might even be a 0.15-era FFF for all I remember. Edit: I realize the point I was looking for a quote on didn't quite make it into my writing: Wube said they've been formulating the science pack recipes to encourage the automation of things that need to be automated for normal life in Factorio; something to that effect. One half of the point you apparently thought it was.

Yes, there's a fair gap between chem pack and pipes, but we are still talking a feature for new players here. They won't be crafting a lot of pipes until they set up oil production as their steam plant will likely be small until then. The step from oil production to chem pack production isn't that great, so it's only a little late. Having to craft all those pipes for their first oil plant and then automating them for the chem science pack might actually be the better prompt because you introduce the need that's felt before having them drop down the solution for research purposes.

It's mostly a pacing thing. The steam plant might be small, but it seems likely to me that some players are going to see expanding it, or possibly moving it if they get unlucky with their first coal patch. I just realized that it's a bad idea to put pipes in the first two science packs because if new players learn from the new demo/campaign to look for clues as to items they need to automate in the science packs, the pipe might lead them astray: should one take it as a clue to pipe the water to the coal patch and build steam power there, they're going to suffer friction losses in those pipes, something which doesn't happen on the conveyors. That moves me further away from having a pipe in the first two science packs, nowhere near to where I am regarding the rails in the production pack, but definitely a nay vote.

[Zavian]

Whilst I think the new science packs sound like a nice improvement for vanilla from a pacing and game balance/design perspective, I've always felt that late game science and rocket parts were too simple. Currently they don't feel satisfying. (Comment obviously based on 0.15/0.16, but I don't see the new recipes in 0.17 changing that much, although the addition of rails and flying robot frames will help). I'd prefer more complex recipe chains for those late game parts. (Of course this is something mods can change, so no big deal).

Now this is just confusing: Your complaint is about late-game stuff, which is things like processing circuits, rocket parts, and power armor. But the word "rails" appears. The running gag is of course this huge pile of rails to run one of hundreds of experiments in a lab. This is not late game, but mid game; the argument I see from the devs fairly consistently is that the item appears in a science pack approximately the same time you're going to be needing a lot of them anyway. It doesn't add up.

Ok. I'll elaborate on why I find the current late game recipes unsatisfying to automate. (Note that in this section I'm only talking about 0.16 recipes. i'll explain my thoughts on the 0.17 changes below).

Ok the late game recipes I'm talking about are processing units, prod science, hi tech science, and the various rocket parts. In most single player games I never bother to automate Power Armour MK2, and I'll only automate things like nuclear reactors, if I'm actually planning on using nuclear power on the map.

First some notes. I'm mainly talking from the viewpoint of an experienced player, starting a new single player base, and aiming for around 0.75 flasks per second, and a rocket in less than 8 hours. Sometimes I'll also mention my expectations of what a first time player might be thinking, and any new challenged he might face.

Prod science.
Only two ingredients. So actually simpler than blue science. What about those ingredients, anything interesting or new there? Well electric engines need lube, that is new, but also trivial to automate. (For a new player the largest challenge with lube is likely to be making sure he doesn't crack or turn into solid fuel too much heavy oil). The actual electric engine recipe is pretty simple as well. (Biggest issue for new players is likely to be realising that AM2s have a fluid input).

What about the electric furnace? 25 ingredients per cycle. But on the whole, that is still less ingredients (per assembler) per second than the mining drill.

So production science nothing really new or complex. Might need to scale up red circuit and steel production and add a few assemblers to engines.

Processing units.
Needs sulphuric acid, which is somewhat interesting to setup, also needs lots of green circuits, and that probably means adding more green circuit production. Can use direct insertion, ratio is good without modules, but ratios will be off if you later add modules. On the whole the decision over making more green circuits elsewhere, or using direct insertion is the most interesting aspect about processing units. (New players will probably miss the direct insertion possibility, and might not realise just how many green circuits this section of his factory needs).

Hi-tech Science.
First recipe that needs 4 ingredients. Processing units, batteries, speed modules and lots of copper wire. Processing modules I've already covered. Batteries is easy. Speed modules are easy. Assuming you are ok with using direct insertion, the copper wire is easy. On the whole the best (most complex) of the recipes I've covered so far.

Space Science.
The rocket silo introduces new mechanics and that in itself makes it interesting. Also requires concrete, which is something I might not have automated yet.

The satellite needs 6 ingredients, many of I won't have automated yet. On the whole it would be interesting, except if I'm only playing for the win in less than 8 hours, I only need one, so I'll just add a few assemblers somewhere for solar panels and accumulators, and hand craft the radars and the satellite itself. (If I'm playing long term, then this is one place where I use logistics robots for science production. For a long term game game I already want solar panels, accumulators and radars in provider chests, (to feed the train that will build the solar fields). Not worth the hassle of duplicating those setups, nor running belts across half the base, to make a few satellites per hour, so I tend to just use logistic robots to automate the satellite. In choosing to use bots to simplify satellite production I lose most of the interesting choices here).

Rocket fuel.
Boring recipe. Far too simple for this stage of the game. (Two steps, each with only a single ingredient. One step in 0.17 since we are already making solid fuel). At least make it more interesting by making Rocket fuel from say coal (and/or solid fuel) + petrol + light oil. (Only an idea off the top of my head).

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. Even in 0.17, where I won't have speed modules automated, it is exactly the same recipe as prod modules. Just add some more assemblers to the end of the belts. (Or for an 8 hour run, just change the recipe of the existing assemblers). At least add some more/different ingredients, eg processing units + speed module 2s or even 3 + a couple of green + red circuits and a few copper wires. (If that seems too expensive, then you don't need to keep the 10:10:10 ratio for rocket parts. 1 RCU controlling 10 fuel and 5 structure would be fine).

Low density structure.
3 ingredients, which makes this the best of the current rocket parts. But all stuff we are already making in quantity. Only one step. On the whole not interesting. I would like to suggest crushing stone + copper ore for rare earth ores, then alloying them with steel to make say titanium (or some other hi-strength light weight alloy). Then making LDS from that titanium plus glass reinforced plastic. (Make glass from bricks in a furnace, then make glass reinforced plastic from glass + plastic. Again only an idea off the top of my head. Could also go with carbon reinforced plastic with carbon made from coal in a chem plant).

Rocket parts
Uses all three of the above. Given that is 3 new ingredients, that is sort of interesting. Would prefer that the ingredients themselves weren't so boring.
Also why have a 1:1:1 ratio, and all 3 ingredients having exactly the same craft time? Yes, once you consider the satellite, this becomes 22:21:20, but that is so close to 1:1:1 that it only matters if you are actually able to keep all those machines constantly fed. Mixing up the ratios by varying either the rocket part recipe, and/or the craft time of the individual ingredients might be an improvement.

0.17 changes.
Production Science.
Now 3 ingredients. Would prefer 4 at this stage of the game, but 3 is better than 2. The third ingredient (rails) is not something I will already be making (at least not in sufficient quantities, so I can't just grab a belt from the bus). It is also a two step process, and that is more interesting than a one step process. Also rails seem like a candidate for direct insertion, which tends to be more interesting to design and build than just grabbing things from the bus. Takes more ingredients and has more stuff that needs to be automated than 0.16, so I expect this to be more interesting to setup.

Utility Science
Only 3 ingredients, but requires flying robot frames, which takes 4. And flying robot frames require automation of electric engines, which need lube, and batteries. Again more things that need to be automated, and again some of them have multiple stages, and so I expect I will find this more interesting to setup than 0.16.

[V453000]

...
So if I want to have 100 packs per second, I will need 75 or so yellow belt, or 25 blue belt feeding this complex... and that production rate is not even close to some of the hard-core fans' ambition level. I am pretty sure you see the outrageous nature of a belt-based approach this recipe is forcing (or, actually, how it makes it impossible).
...

If my numbers are correct, 100 science per second is 6000 per minute. How many people reach that number and how hard is it to reach that number in so many other parts of the factory? How many blue belts of iron plates do you need for that in comparison?

I find worth mentioning my assumption of science production:
- a typical advanced player builds 5-6-10-12-7-7 assemblers for the respective science packs, as Zavian mentioned that would be 0.75 with assembling machines 2
- it's not that uncommon to see x2 or x3 multiplications of the previous ratio, but both of them get so fast that you can't really build fast enough to keep up with your research, making the research eventually stop, waiting for you to get to the next science pack anyway.
- a newer player will likely build less han the previous ratio, or keep the actual production rate much lower (many machines will be idle)

A lot of the complaints I see about the amount of rails is generally about:
-A: I don't like to belt this much.
-B: I don't like to direct insert.
-C: It's too complicated.
-D: It's expensive.

-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.
-reB: Since there are other alternatives, nobody is forcing you to direct insert, but in my opinion it is a viable option to consider. (and making areas where the player can decide because multiple options are reasonable, is good)
-reC: It's the top tier science. Being simple also means boring, and majorly contributes to feeling tedious and uninteresting. If you don't have strict requirements for yourself (beacon setup, compact setup, ...), just building something that works (even somewhat sub-optimally which is probably enough as it's likely a new player doesn't have enough ingredients to support full speed production) is not that hard. This is especially valid considering the amount of time that you have already spent in the game, and the challenges you have overcome to this point (oil for example).
-reD: It's exactly as expensive in ore resources as Utility science pack. The rails are a significant amount of price in the Production science pack, but an electric furnace is about the same total price with a much more complicated recipe chain.

I guess I'm just repeating what I already said before and we are discussing in circles. Honestly I didn't feel very appealed to continue discussing after the recent few pages went into berserk of a few individuals about pickaxes and claims that we are (very rapidly getting to be) detached from our community - that one specifically I find personally offensive because of how much time I (and many other team members) spend interacting with the community. I would like to thank Koub for appearing and asking for order. I'm not saying we shouldn't discuss it, but are other places for it and there is a point of amount of posts and amount of threads that anybody of us could follow about this topic, if we hope it to be constructive and arrive to a real outcome. Neither major crossposting nor just repeating the same arguments over 20 thread pages is going to help. We are aware about the problem and as abregado tried to straighten up, we are trying to avoid shrinking of mod possibilities, and with that finding solutions that would satisfy even the most edge cases. (as Zavian noted, the 1% of modded and 0% of vanilla people - whether the numbers are precise is not super relevant, but you hopefully agree that pickaxes are a very minor thing)

Since there seems to be a very large portion of people who are very much in favor of all of these changes and the counter arguments I personally did not find big enough (you can consider me a short-sighted, detached a-hole who is only attached to his own idea - if you find it necessary and if it will make you happier), I myself will from just focus on the few loose ends:
-1: pipes in logistic science pack
-2: crafting time of chemical science pack (maybe also military)
-3: rocket silo recipe
-4: space science pack technology not requiring military science packs

- re1: The iron increase is negligible. The complexity added is very minimal. First time players tend to build 1 machine of each type, meaning that a 3-ingredient setup is not a problem - they've already passed that test with automating inserters. Pipe is basically an intermediate product because you use it in so many recipes (UG pipe, refinery, chemical plant, ...) so it's very consistent with the other intermediates in science packs in this regard. It's indeed been "fine" since forever, but refinery/oil has also been a giant problem since forever - this problem is not going to go away, and it's not even 100% wrong - complexity is respectful, challenging, daring and interesting - but slight bumps in the right direction but not giving explicit hints are good.
- re2: I already said it a number of times that I would like to change this, my current proposal is 16 seconds crafting time for both Chemical and Military science packs.
- re3: An interesting complaint I've heard is that the step between rocket parts and actually launching the rocket is fairly minimal and the Satellite was at least something on top of the rocket parts. However maybe the issue is instead that the rocket silo is a little too simple. I would not go full nuts here, but maybe replacing a few pieces in the recipe like concrete to refined concrete and maybe adding fluid to the recipe so you couldn't just hand craft it would be in order, and the pacing of the last step in the game would feel more appropriate.
-re4: I believe it should not require military science packs, we'll see what consensus do we get next week, I'm home this week as I'm ill so no news from me regarding that till then Sorry, I prefer mouth-to-mouth discussions with colleagues.

Thank you for your time and attention. I would like to thank people like Zavian - not for agreeing with me, but for having a thought out, polite discussion.

[quadrox]

This update suggests that the devs think a lot of players are using automated science production as a source of materials for further expansion, and expansion of their main factory to fuel science production. My play style definitely doesn't include this, nor does that of most of the people I've watched. When I build my first refinery, I am hand crafting all the pipes, and just working on something else while that process finishes. Automating red ammo for defense vs for science are two completely different concerns for me, one often coming long before or after the other depending on the game settings I'm playing. Using solid fuel from my main refineries to produce science is... not going to happen.

Am I wrong here? Are many people driving science production from intermediate products produced in your main factory?

I have to agree with this post. I generally set up one area of my base for science production and then don't touch anything that goes into it. If I need materials for base expansion, I set up production elsewhere.

Completely agree as well.

In fact, I have always felt that science was a huge distraction that prevented me from building out my factory and automation process. I have to invest such huge resources just to get science running that I barely have time and resources to get the actual factory built. And then you have to wait such a long time before you can improve the process with bots, which requires oil, which requires science, which... Let's just say that I haven't even played since the previous science change as it just seemed too much hassle.

Instead I started working on a mod that would allow me to automate much of the tedious process, but I got distracted by other stuff, so now I really want to play factorio, but I just can't be bothered anymore... Too bad, I really would like to play with artillery and other cool stuff.

[RocketManChronicles]

I really like these changes to the science in the game. Thank you devs for going to great lengths in your considerations and showing us the depth you are willing to dig to give us a great experience. Despite a lot of the naysayers here, I am happy with what you have proposed for us.

I say add the pipe to the Logistic Science recipe. It would give us an opportunity to figure something new out.

As for the "rails controversy" I see on this thread, 30 rails is not at all that bad. The crafting time for the science pack allows direct insertion as you propose, and that makes it more interesting to build that science pack. It keeps the ratios nice and clean.

And what people like to forget here, is that we have always had more assembling machines making the blue science pack than any other. If you were to use the easy-math 1spm, you would have 5 red, 6 green, 12 blue, 7 purple, 7 yellow. The proposal doesn't change that much for us to figure out, and I like what you have and cannot wait to experience it for myself to rally make any judgments on it.

[V453000]

My play style definitely doesn't include this, nor does that of most of the people I've watched. When I build my first refinery, I am hand crafting all the pipes, and just working on something else while that process finishes.

And that's fine, the science recipes aren't forcing you to change your behaviour. Yet, I would dare saying that there are less tedious and more automated methods.

In fact, I have always felt that science was a huge distraction that prevented me from building out my factory and automation process. I have to invest such huge resources just to get science running that I barely have time and resources to get the actual factory built.

If you as a player get everything for free, all recipes are cheap and available early, most of the content between the beginning and the end of the game is irrelevant, and none of the end game content feels nearly as powerful.
By all means I understand Factorio is a crazy game which not only lets you build big (with the infinite map and performance optimizations), helps you build big (construction robots and blueprints), but expects you to build rather big compared to other building games, while the exact amount of "your big" is still up to you of course.

Am I wrong here? Are many people driving science production from intermediate products produced in your main factory?

I have to agree with this post. I generally set up one area of my base for science production and then don't touch anything that goes into it. If I need materials for base expansion, I set up production elsewhere.

Even though this setup is completely fine and acceptable, I sometimes do it myself, the design is versatile and it's really up to you how you build it. I'd even say that if you are a player who builds the science setup separately, you by yourself are treating the science as any "another recipe" and therefore the changes have relatively little impact for you - except the prices and what does which technology unlock.

Small side noter: If you so desperately need Logistic system early, it's now accessible a bit earlier - without Production science pack.

[jodokus31]

...

In fact, I have always felt that science was a huge distraction that prevented me from building out my factory and automation process. I have to invest such huge resources just to get science running that I barely have time and resources to get the actual factory built. And then you have to wait such a long time before you can improve the process with bots, which requires oil, which requires science, which... Let's just say that I haven't even played since the previous science change as it just seemed too much hassle.

Instead I started working on a mod that would allow me to automate much of the tedious process, but I got distracted by other stuff, so now I really want to play factorio, but I just can't be bothered anymore... Too bad, I really would like to play with artillery and other cool stuff.

You could use creative mod, with instant research. Then you have time build a factory, maybe even with instant blueprint placing. It is very widely configurable. The cheats can be switched off later.
For myself, the process of evolving through the research is the most fascinating thing in factorio and gives a plot, but its also time consuming.
And I also enjoy building testing designs f.e. for rail networks with creative mod, to get to the point of interest faster.

[quadrox]

My play style definitely doesn't include this, nor does that of most of the people I've watched. When I build my first refinery, I am hand crafting all the pipes, and just working on something else while that process finishes.

And that's fine, the science recipes aren't forcing you to change your behaviour. Yet, I would dare saying that there are less tedious and more automated methods.

Automated methods which you won't have time to use if you want to get on with science. Of course you can just wait with science until you are done automating everything else, but then you have to wait even longer until you get bots, which means even more tedium. No thanks...

If you as a player get everything for free, all recipes are cheap and available early, most of the content between the beginning and the end of the game is irrelevant, and none of the end game content feels nearly as powerful.
By all means I understand Factorio is a crazy game which not only lets you build big (with the infinite map and performance optimizations), helps you build big (construction robots and blueprints), but expects you to build rather big compared to other building games, while the exact amount of "your big" is still up to you of course.

I understand that, and I don't want science to be completely free. But I wish that the big resource drain was associated with something where I get something more permanent, more visible out of it. Until the late game most effort/resources/assembler have the purpose of producing science. Only a tiny percentage goes toward defense and future expansion, and this imbalance feels extremely unsatisfying. I want to build large factories that produce something tangible, that allow me to expand evermore, and that make me feel that I am making progress, but always the science monster keeps my expansion plans at bay. I really wish the main factory was the larger part of the game instead of the science factory, but I am not sure what would need to change for that to become possible. In any case, it's probably too late to do anything about it now.

[Proculator]

I really liked the information you provided on science packs … very useful.

Too bad you did this thread before I posted: Entity icons for stack equations and Entity equations threads.

Then you would have had the math to illustrate how each science pack differed.

Hope you find this post useful.

[Serenity]

Rocket fuel.
Boring recipe. Far too simple for this stage of the game. (Two steps, each with only a single ingredient. One step in 0.17 since we are already making solid fuel). At least make it more interesting by making Rocket fuel from say coal (and/or solid fuel) + petrol + light oil. (Only an idea off the top of my head).

Low density structure.
3 ingredients, which makes this the best of the current rocket parts. But all stuff we are already making in quantity. Only one step. On the whole not interesting. I would like to suggest crushing stone + copper ore for rare earth ores, then alloying them with steel to make say titanium (or some other hi-strength light weight alloy). Then making LDS from that titanium plus glass reinforced plastic. (Make glass from bricks in a furnace, then make glass reinforced plastic from glass + plastic. Again only an idea off the top of my head. Could also go with carbon reinforced plastic with carbon made from coal in a chem plant).

I think for the end game recipes the game suffers from a lack of raw ingredients. When you get down to it everything is made from just iron, copper and oil. And while oil is complicated when you get to first (further increased by then needing to produce plastic and red circuits), it boils down to using a little light and heavy and turning all the excess into petroleum. The cracking is a great amount of complexity when you get to it, but later on it will mostly run itself.

Later in the game, introducing 1-2 more raw ores and 2 more fluids would be nice and allow more complicated chains. I'm not asking for Bob/Angel level of craziness here. Just maybe using something like aluminum for low density structures and rocket fuel (like it is used in real life for some solid rocket boosters). Maybe the requirement for some acid to refine/treat those metals in some way. Another fluid could be required for rocket fuel.

I know that it's not going to happen. Especially at this late stage in the development cycle. Using some more complex recipes from existing materials like you described would be very nice too though. In any case having two intermediate stages instead of just one would an appropriate level of complexity. As it is there is no ramp up in complexity for the rocket. It's just more of the same
There a market opportunity for a mod there though if the author can manage to stay close to the main game and not go over the top like the other industrial expansions.

Utility Science
Only 3 ingredients, but requires flying robot frames, which takes 4. And flying robot frames require automation of electric engines, which need lube, and batteries. Again more things that need to be automated, and again some of them have multiple stages, and so I expect I will find this more interesting to setup than 0.16.

By itself the build is very simple. If you already have a robot blueprint you can mostly just use that and change the number of assemblers a little:

YellowScience.jpg (160.64 KiB) Viewed 5799 times

There is one new material here in low density structures that will have to be automated and that may require scaling up copper, steel or plastic further.

Batteries are probably already automated at that point. A newbie will probably not have set up lube yet, but if you know what will come you will when you do your refinery.
Getting the required number of blue circuits maybe need some work

[lacika2000]

...
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.