Factorio Forums

Dune wrote:

abregado wrote:So far I would say the testing versions of the NPE feel less like a tutorial and more like a playable trailer (but I am biased).

Can you not use acronyms please? I've no idea what a NPE is, and the closest I get is NBE (Non-biological-extraterrestrial) if anyone gets the reference.

In my experience as a software developer, NPE stands for NullPointerException.

Oktokolo wrote:The readability of the science tree could benefit from using slightly diagonal lines wherever horizontal lines are used now. Not connecting implied dependencies to reduce the amount of lines also is a good idea.

+ 1 for autodetecting implied prerequisites and removing them (be it data or visual in the gui only)

Oktokolo wrote: But i would also like to get rid of science packs alltogether!
Let the science centers consume real intermediates like all colors of circuits, modules, inserters, belts, combinators, wires, batteries, plates, ores (for mining research), guns and ammo (military research)...

I actually thought of that myself yesterday when pondering what to do about the broken "science bottle" mataphor. In-game they're called "science packs" but they look like bottles. And there's quite some mods that prove that the whole "science juice" methaphor is a weird one and doesn't fit into the scenario very well. Renaming them to "research material pack" or something and changing the icon accordingly could fix that. And while i personally like the idea of just directly shoving the materials into the labs without packaging them first there are a few problems with that (some of which you consider positives):

1. Every lab needs every ingredient. So you have to supply each lab with everything. Would be easier if you could supply item_a to half the labs and item_b to half the labs, even if there was a speed penalty for that. (Actually a speed penalty for that sounds like a nice way to encourage late game optimization while keeping early game simple.)
2. You'd have to shove 13 materials (16 for military) into every lab. This is firmly into the territory of advanced logistics (i.e. circuit controlled sushi belts) and not at all suitable for new players. And the system has to work for everyone...
3. Displaying 13 different icons on every technology is both difficult/ugly to layout and makes for very bad readability of the gui.
4. Research material is used up gradually, but not every item supports that kind of granularity. I.e. gears, circuits, etc have no concept of "half used up", and thus it would need to be added. Which would probably have an impact on performance as every item now needs to track its status.

I started playing Factorio in the mid-0.14 days. loved the campaign and am eager to see updates made to it. In particular the train campaign was really good and showed how trains can be loaded / unloaded. I was eager to jump into freeplay after that.

As for science packs I'm happy to see the changes proposed but I think research could go through additional overhaul. In my first playthrough my base was an utter mess. Trains helped but the core of my base needed help and this is where logistics robots came to the rescue. However, I couldn't figure out how they worked until I found the tech that spoke of Requester Chests. This got me building robots and pushing to that chest so I could fix my spaghetti.

I don't use logistics robots for throughput. To me they are temporary help with spaghetti parts of my base that I don't feel like completely redoing. In 0.14 they felt like a fluid next step after unlocking oil: something to use heavy oil for and solve an immediate need.

I was very sad to see the chests pushed out to late game and locked behind both purple and yellow packs. I really don't get why requester chests are late game and feel they should be unlocked at the same time as logistic robots are unlocked.

I'm sad to see logistics robots unlocked so early but be useless until requester chests. Back in 0.14 days it was a single hop from logistics robots to requester chests and only required blue science. I think they should require batteries but not blue chips.

I think there are some key technologies:
- plastics (red circuit)
- batteries
- blue chips (blue circuit)
- nuclear

I feel blue, purple, and yellow science should each focus on one of these. Maybe batteries should be required for purple science and logistic robots moved over to purple. Yellow can focus on the really advanced stuff that requires blue chips and modules.

It's interesting how different people view things differently. To me, flying robust solve a logistical challenge that should be mid-to-late game and speed modules and beacons are for really late game. Thus robots should be purple and beacons yellow in my eyes. Plus robots are fun and you want your players to have fun.

As for nuclear science I almost feel it could be its own track and requires uranium to research. The science pack can also unlock the centrifuge needed to produce U-238 which can be used in the research.

But special care should be taken when crossing research streams. Request chests sjouldnt be both production And high v-tech. I don't like that armor requires military science for example, because I want it for the modules. There should almost be separate path for modular armour with roboports and military armour with combat drones and lasers.

I think the only thing that should require multiple science packs beyond blue is the rocket silo which should nicely tie together the military, nuclear, production, and high-tech aspects.

Just throwing some ideas out there.

The line between production 2 and engine has a one pixel offset. As a hobbyist graphics guy, this would make the game literally unplayable.

On a (more) serious note, I like the fact that you will research the science packs separately, and that you won't have to restart in the campaign. I might even properly complete it after this!

eradicator wrote:Every lab needs every ingredient.

You don't need raw ores for anything but mining research.
You don't need ammo for anything non-military.
Circuits will not need steel, belts or inserters (you can safely assume, that the labs come with some equipment and in Factorio, stuff does not need maintenance).
It should be possible to rearrange the tech tree so that you would end up with parallel branches that overlap in some ingredients but mostly need different stuff. That way, you could do the old belt feeding with 10 or so ingredients for any of the four to five resulting lab clusters.
But you could also do sophisticated setups to feed all 20 to 30 ingredients into one big lab cluster using multiple, potentially even braided, rainbow sushi belts.
And most of the stuff would be low-quantity.

eradicator wrote:You'd have to shove 13 materials (16 for military) into every lab. This is firmly into the territory of advanced logistics (i.e. circuit controlled sushi belts) and not at all suitable for new players. And the system has to work for everyone...

Recipes could be altered to get the material count down to an arbitrary number for each research branch. There is nothing that prohibits the devs to get the ingredient count for military down to 10.
I would recommend all three plates, stone bricks, explosives, belts, inserters, circuits, handgun ammo.
Better belts, inserters, ammo, or circuits would be needed for higher tiers of research and lower tiers would just progress faster when given better stuff than requiered.
That are only 9 - so let's add gears or engines to reach the 10.
The same can be done for any arbitrarily chosen cluster of recipes and desired ingredient count. It is a game after all.

eradicator wrote: Displaying 13 different icons on every technology is both difficult/ugly to layout and makes for very bad readability of the gui.

Yes. That is a challenge. Would need some rather distinct icons to be able to grasp what is actually needed for wich recipe. But most recipes of a tech tree branch would need the same base ingredients plus some that are specific to twigs of a branch. So maybe, clustering recipes requiring the same ingredients visually together branches and twigs could help.
I would leave some space between the tree branches and a bit less space between the twigs - and every branch gets its set of ingredients while every recipe of a twig shares exactly the same set of ingredients (differing only in needed quantities).
Higher tiers would require more advanced versions of some ingredients, but lower tiers could be researched feeding that higher versions too and would get a speed boost if doing so.
I think, the tech tree could benefit from a more clean separation of branches and twigs. It would be less complex and therefore easier to grasp. At the same time, science gets more demanding logistically - wich i consider a good thing as Factorio is all about logistics.

eradicator wrote: Research material is used up gradually, but not every item supports that kind of granularity. I.e. gears, circuits, etc have no concept of "half used up", and thus it would need to be added. Which would probably have an impact on performance as every item now needs to track its status.

Minimum quamntity of ingredients for each research step would be 1. If you change research, the unfinisched current step would be lost. But that is no big deal as most research needs hundreds of bottles today and that would just be translated to sets of ingredients. so the potential loss is negligible.

Zavian wrote:

Avezo wrote:My proposal is - remove 'flying robot frames' from the game, produce straight construction bots with ONLY red/green science materials OR logistic bots with higher tech materials.

I would rather they went with something like nanobots. (Basically expendable mini-construction robots that can be built with red+green science). That way they could move real logistics/construction bots to blue science.

I never used the nanobots, so correct me if I'm wrong, but don't they make the normal bots useless, since you already have bots?

Oktokolo wrote:Minimum quamntity of ingredients for each research step would be 1. If you change research, the unfinished current step would be lost. But that is no big deal as most research needs hundreds of bottles today and that would just be translated to sets of ingredients. so the potential loss is negligible.

Back in 0.13 (or maybe even until 0.14?) science bottles worked like this. And it was bad. It basically removes the ability for the player to freely decide that they want a different tech now. Especially in low-supply situations where you want to quickly switch to some green/red reserach because you're out of blue it's bad. Quite frankly bringing that sort of behavior back is simply not an option.
Regarding the rest of your post... your premise is that a move to raw ingredients would also result in complete reordering of the tech tree into groups with similar requirements. And you said yourself that one would need 4-5 different blocks of labs, meaning 80% of all labs would do nothing at any given time because they're supplied with the wrong ingredient group. That's not a fun gameplay mechanic. I purposely assumed that even *if* the science packs were replaced by direct use (projected chance of that happening: <1%), the overall requirements wouldn't change. Because completely redoing the whole tech tree from scratch is not something that is realistically going to happen. Btw, all of this is quite easy coding if you want to make a demo-mod ;).

And now...back to campaign talk :D.

I never played the campagin and i never launched the rocket. I have hundrets of hours in the game. I didn´t spend 1 minute in the campaing. I only play multiplayer. I wonder if there are stats how many percent of the players play the campain(longer then 10 minutes) and how many percent do play singleplayer.

Can you play the campain in multiplayer? This would be a feature where i would play it. Parallel to the main-game i am playing - multiplayer of course.

But i never would start with campaing - even if it´s multiplayer. Because i don´t wan´t to get spoiled. The campains are often made to have super quick and short outcomes. So you have in a short time seen every part of the game. In multiplayer we do science very slow - because of fighting each other. Or in coop play we don´t have more then 2 science labs. So we can live the phases of the game longer and have longer the fun of wondering what the next things are in the game.

Bokkie wrote:I never used the nanobots, so correct me if I'm wrong, but don't they make the normal bots useless, since you already have bots?

The Nanobots mod allows you to reasearch an early game (red science) "weapon" and "ammo" wich together somewhat act like a personal roboport if selected as current weapon (but pulling the trigger only wastes your bots so you have to switch to a real weapon to kill bugs). Each bots can only construct/pickup/repair one thing and disappear after the job. They also have rather limited range.
It is more or less a QoL mod wich makes early game blueprint use a lot less annoying. There also is research for equippment that lets the nanobots interact with the logistics network - but you are better off using regular bots with a real personal roboport at that time.
The mod is currently the second most downloaded of its category with 107k downloads.

If you want to make the tech tree so it's more understandable for new players, please change it so you never unlock a recipe that you can't use. It boggles my mind that the current tree is "intended behavior."

viewtopic.php?f=23&t=33612
https://www.reddit.com/r/factorio/comme ... d_require/

GuiltyBystander wrote: please change it so you never unlock a recipe that you can't use.

That issues is more complex than you think it is. More lines make the tech tree more difficult to read, and many technologies would end up containing only one recipe because the others have different requirements.

Have some threads on the topic:
viewtopic.php?p=339162#p339162
viewtopic.php?f=6&t=58079
viewtopic.php?f=6&t=59692

@eradicator, not to mention, many mods create additional dependencies .... that will be huge mess

i'm glad that game at least checks for circular dependencies on load

j3robins wrote:I'm sad to see logistics robots unlocked so early but be useless until requester chests. Back in 0.14 days it was a single hop from logistics robots to requester chests and only required blue science. I think they should require batteries but not blue chips.

For me, Logistic robots are two things; Personal Automated Supply from your base, and Robot based Factories (requester chests). This makes early Logistic Robots tech far from useless. Separating these two uses actually means you can get access to them earlier and start experimenting. Getting requester chests earlier would also invalidate a lot of what the player has learned by providing a clearly better option over spaghetti belts.

GuiltyBystander wrote: please change it so you never unlock a recipe that you can't use.

eradicator wrote:That issues is more complex than you think it is. More lines make the tech tree more difficult to read, and many technologies would end up containing only one recipe because the others have different requirements.

While I agree with GuiltyBystander, the end result of this is actually that every recipe has its own research item. Also, interestingly maybe to some is that when I surveyed players (young,old, new and vets) most people actually enjoy the chaotic nature of the tech tree. I made a "clean" tech tree mod and most people preferred the stock one.
https://mods.factorio.com/mod/tech-trunk

j3robins wrote:In my experience as a software developer, NPE stands for NullPointerException.

Haha! The other option was New Player Content... but it umm... was also already taken. I will endeavor to just write all three words until we come up with a new name for it.

eradicator wrote: "Scroll Direction":

Firstly, thanks for your effort in replying! The image we posted was just a quickly made one to show the concept. Actually, planning out the areas at the beginning is a terrible idea, so we first make dependancy diagrams to show which areas need to be connected where, and then later we can lay out the level. Be sure it will not just be in a single direction. Your diagram is nice, but I think you will be pleasantly surprised with what we come up with.

eradicator wrote: Remnant Factories:

These are super tricky but I think you have the right idea. There are a lot of opportunities to demonstrate concepts to the player. The idea of giving players items before they can make them themselves is also tricky as it really messes with the dependencies. The game already has several layers of dependency (Tech, Recipe, Concept) and the Campaign adds to that a layer of Geographical dependancy not seen in Freeplay.

eradicator wrote: Science Pack Dependencies:

In 0.16.x there are so many hidden prerequisites to the technologies. Some "require" other unlinked techs because they contain a recipe needed for a recipe given in the unlocked tech (see Concrete + Rocket Silo). The other type of hidden prerequisite is the science packs needed. This was taken into consideration with the new change. Having new technologies which are not linked anywhere to the main tree makes them incredibly hard to locate, so we decided against having them in their own tree. The same goes for "technologies which act differently, aka auto unlock". This late in development it would be prudent to change how technology works.

eradicator wrote: Voice Acting:

We cant find any voice actors of the species that the Avatar comes from...

eradicator wrote: Btw, i don't even play on steam, but if there was a statistic for it, i'd probably have something like 1 rocket per 1000 hours . I usually overengineer my factory before the rocket stage, and then start over before actually launching one. And i haven't done a single "infinite technology" yet.

Well, I based my research on Steam statistics because there is no other data. There is a lot of discussion in here about that being a terrible way to go, and I agree. My thought process on this was:

• People who choose to keep their data private (offline mode) probably wont expect a dev to make something specifically for them.
• Those that do will voice their opinions in the FFF Thread
• Players who use mods (and therefore dont have the achievement) are more likely to be Sandbox players, and therefore less interested in "Winning the Game".
• A player seeking A Guided Experience will probably be playing online and getting the achievements.
• There is no other data to go on.

That given, Id say that my statement that players seeking Guided Experiences are underserved in Factorio was correct.

Alice3173 wrote: Going by Steam achievement numbers is pretty unreliable anyways. It's not uncommon for popular games to have an unavoidable achievement at the beginning of the game yet it has a 60-70% rate of game owners having gotten it. For example:

This means that more than 3/10 owners of Portal 2 haven't even gotten through the first five minutes or so or the game.

While I find them silly, these "Achievements just for loading the game" do help with the stats! Pity our Achievement Art is so top notch, otherwise id ask for a "Start Any Game" and "Crafted your first Red Pack in an Assembler" achievement. Some things which would be unlocked easily by players just loading a factory.

An Achievement just for "Loading a Modded game" would also be nice.

abregado wrote:

GuiltyBystander wrote: please change it so you never unlock a recipe that you can't use.

eradicator wrote:That issues is more complex than you think it is. More lines make the tech tree more difficult to read, and many technologies would end up containing only one recipe because the others have different requirements.

While I agree with GuiltyBystander, the end result of this is actually that every recipe has its own research item. Also, interestingly maybe to some is that when I surveyed players (young,old, new and vets) most people actually enjoy the chaotic nature of the tech tree. I made a "clean" tech tree mod and most people preferred the stock one.
https://mods.factorio.com/mod/tech-trunk

tree for high tech pack might look like this for example (mind this is not exactly vanilla tree) and for science 3 if i connect all science packs to every technology they unlock, what it will do with this (except science tree on right side, this is vanilla advanced electronics, and it's not cleaned, nor properly reconnected) while i don't enjoy disconnected technologies, because sometimes takes time to find prereq, connected are also problematic

and this is what i had in mind, when it comes to mods ... see that lines on top? so yes, it's not that easy to actually "cleanup" or made properly connected tree

abregado wrote:

Alice3173 wrote: Going by Steam achievement numbers is pretty unreliable anyways. It's not uncommon for popular games to have an unavoidable achievement at the beginning of the game yet it has a 60-70% rate of game owners having gotten it. For example:

This means that more than 3/10 owners of Portal 2 haven't even gotten through the first five minutes or so or the game.

While I find them silly, these "Achievements just for loading the game" do help with the stats! Pity our Achievement Art is so top notch, otherwise id ask for a "Start Any Game" and "Crafted your first Red Pack in an Assembler" achievement. Some things which would be unlocked easily by players just loading a factory.

An Achievement just for "Loading a Modded game" would also be nice.

"kill your first bitter" achievement?

I like the split between high science and production.
Dislike the Logistics System still being locked behind yellow science.