Friday Facts #227 - Rendering, Trees & Scenario talk

[ske]

If you want different builds, what is needed is different recipes for each item.

This idea sounds very interesting. Randomizing recipes requires different setups each time.

I see another way to come to a very similar result: We permanently introduce multiple alternative recipes to obtain the same item by different pathways.

With (advanced) oil processing we already have one example implemented. With burners we can burn many different things for energy (wood, coal, solid fuel, etc.).

Unfortunately some parts of the game (handcrafting) are not made with thought of those alternatives. When handcrafting, you select the item you want and it automatically selects the recipe to craft it. If there are alternatives, how would you select them? I'm sure there is a solution to this.

[BlackKnight]

Benefits from separating shadows from entities...

Seperating the shadows from the entities means hopefully we get a moving sun.

This would be totally awesome - creating a much more realistic day/night cycle and giving every object a more 3D feel.

Which may be a result from dynamic shadows:

I'm wondering, wouldn't it be easier to generate shadows in-game from object sprites instead of using separate shadow sprites? With lots of sprites overlapping it might be much less taxing on hardware.

Please help us make the right choices in the graphics settings by adding more descriptive tool-tips and for every option!

This would be really useful to help save those of us who like to play with the settings but end up making things less efficient! On top of that even a little side note mentioning what combos would be good for what level of system.

I still think that we need to have some sort of "cosmetic" designation for mods such as WaiTex, Colorblind Circuit Network, and KingIon's mods such that they don't need syncing or prevent MP connection.

I second that. Even those like EvoGUI which provide a way that helps get basic (IMO) non-cheaty info. So you could use it in singleplayer without achievements being disabled.

[Lubricus]

If you want different builds, what is needed is different recipes for each item.

This idea sounds very interesting. Randomizing recipes requires different setups each time.

I see another way to come to a very similar result: We permanently introduce multiple alternative recipes to obtain the same item by different pathways.

With (advanced) oil processing we already have one example implemented. With burners we can burn many different things for energy (wood, coal, solid fuel, etc.).

Unfortunately some parts of the game (handcrafting) are not made with thought of those alternatives. When handcrafting, you select the item you want and it automatically selects the recipe to craft it. If there are alternatives, how would you select them? I'm sure there is a solution to this.

You chose a recipe for handcrafting so there is no problem with it. (angels mods often have several recipes for one thing)
Implementing random recipes is a little tricky because of the everything have to be deterministic rule so it have to be dependent of a seed generated by the map creation or something? There is a mod for random recipies. https://mods.factorio.com/mod/zRandomRecipes

[BlackKnight]

GAME SCENARIOS

Tower Defense Scenario PLEASE!
I am a huge fan of Tower Defense games and have been wanting to play Factorio with a TF mode since the beginning as I can see it being an absolutely perfect environment to make this work. I know there were mods ages ago but always severely lacking.

• IDEAS
• It should have the same typical rules Tower Defense games have..
• Concept - It would be very similar to what the Wave Defense scenario is now but have some differences
  - Walls would be impenetrable so aliens would have to find the route with the shortest distance - the game would require you to maintain at all times a 1x1 tile path to a "Core" area of the base. Alien AI may need some improvement in pathfinding so they dont get stuck. Using the natural formations of Water and Cliffs would also be useful for integrating into path making.
  - Path display - The path the aliens would take would be displayed somehow to provide the player with insight in where to build up defenses and reconfigure routing. It may also display how many tiles in a GUI the most efficient path is.
  - Location - You might start out at the site of a crashed spaceship? and would have a limited space to setup your base.
  - Turrets would be key and possibly additional turrets required (oil slicks, (short range) poison/slowdown capsule launchers, etc) and traps?!! Aliens would perhaps prioritize the most powerful defense buildings first.
  - Gates could be used along walls (yet not breaking the previous rule) allowing the player shortcuts but could also let in biters if the player wasn't careful.
• Goals
  - Player Win Condition - Survive all waves or build a spaceship and launch yourself into space?
  - Alien Win Condition - would be to destroy a building at the center of your base / all buildings and or the player (and or could have other requirements).
• Additional options
  - Night Waves mode - Each night the wave hits - increasing in strength each time.
  - Could provide the player with a number of building items to start with (enough to speed the game up past a specific stage in the usual game to not make this a slog)
  - Could either have a monetary reward system from waves (with much less/no manual building required) to allow purchasing (as is currently the case with the Wave Defense scenario) or start from scratch with nothing and build your way up just like Wave Defense. The trick is to somehow meld the two (Tower Defense concept + Factorio) together.
  - One idea on early game building - provide construction bots with perhaps a speed boost? to allow using map mode to control items more like via RTS..
  - Gain the option for immediate research upgrades after passing a wave

OTHERS
Segregated Co-operative Scenario
An interesting idea for a co-operative mode where individuals or groups have to share with other teams to succeed. Each has one or more resources but not all and they must Share by belting and or pumping resources out of their fixed area under a wall to their neighbor(s) area and vice versa to complete specified goals.

Capture The Flag Scenario
(as suggested earlier viewtopic.php?f=38&t=57174&start=20#p338586)

Transportation Scenario
Between Factories/on a Factory Floor, emphasizing use of trains or belts - You would receive orders of certain items/in specific time/of specific throughputs and you would have to fulfill these by building the most efficient solution with the items/raw material you have access to.

Quick Start mode
Just like preselecting resource levels before starting a game, this would be a config tab that would allow you to select what items (ie. personal roboport, construction bots) and research levels were preset before starting.

[Avezo]

was that the one where you control tanja? in the building? <3

I don't remember it that well anymore It's over 20 years after all. Little as I recall it was some rescue mission with just soldiers.

[ToneMcStone]

With the competitive aspect, what if there to make a RTS style game with team based biters?

A lot of people have mentioned bots and AI controlled cars and tanks but I haven't seen much mentioned about using biters to fight another side.

Here's a few ideas and points to consider

Research Tiers of spawners


-Spawners could be like Assembly machines with higher tiers having to be researched to be able to make higher tier biters/more frequent spawns
-They could be like Oil refineries so the same spawner but higher tiers need to be researched.

Pollution based Tiers of spawners

-Evolution based on the other teams pollution so this may balance the ability to attack. The other team has a larger factory but you have more attacking power.
-Evolution based on your own base. Make the biggest, dirtiest base you can.
-Evolution is global.

Maybe there's a mix of research and pollution. You can have to research the tiers of spawner but your ability to produce higher tier biters is based on pollution.
Maybe the spawners are pretty much team based, played deployed versions of the biter spawns in vanilla?

Spawn rate/producing biters

This could be based on research/pollution/evolution and/or having the spawners need to be supplied with materials (don't ask me what though! maybe a liquid as 1 part)


Biter movement

At a minimum I think players would be able select the biters and move them to a point and after gets to the point the AI kicks in
There's a loads of potential movement commands, formations, aggression stance ... but that sounds like a whole load of work


God Mode

God mode would make it feel like other RTS so maybe but this is Factorio so maybe it just be normal or a research. That would be a big leap in ability.

Map Design

Like it has been mentioned with tower defence, with cliffs and water there's impassable points or these also could be overcome.
Thick woods could provide places to be consealed.

[NotABiter]

I see another way to come to a very similar result: We permanently introduce multiple alternative recipes to obtain the same item by different pathways.

Except that doesn't achieve the desired result at all.

This is the problem as originally stated:

So it is a battle, who has the best blueprints loaded from Internet, instead of who is the best factorio player.

If you consider the general form of that problem, it is not just a PvP problem but a single-player problem as well. That is, if I start up a new game, the optimal set of builds is going to be pretty much exactly the same as it was the last time I played the game (and the time before that, and the time before that). All I need is a good set of blueprints and I can not only play the game without really doing any thinking at all, but I can play it near optimally that way, with my brain basically set to the "off" position.

Your suggestion, having alternate fixed recipes, does not change that situation. Any time such a fixed recipe is added to the game that may mean (for a given tech level) some new build is now the optimal build, but that only means that new build gets added to the "optimal builds blueprints book" (and maybe an old build which is now suboptimal gets tossed out) and then you're once again done - no more real designing/thinking required for all future games. (Or even no designing/thinking required ever - just download blueprints others have made.)

Only by having the recipes (and/or other things that affect builds like assembler module count, assembler/belt/inserter/robot speed, assembler/beacon size, max robot/inserter capacity, power pole/substation range, which techs/items are available to research/build, etc.) change every single game does it result in the player really taking a fresh look at build designs every single game.

[LotA]

Scenario suggestion :

• Tower Defense
• PVP Attack-Defense
• Car Race
• Train : escort the convoy (could be either pvp or pve)

[Rayfie]

On the topic of scenarios, I would like to refer to my earlier post. This combines greatly with the topic about bots vs. belts and creating variety when restarting the game. Of course this could also apply to PVP.

[ske]

Your suggestion, having alternate fixed recipes, does not change that situation. Any time such a fixed recipe is added to the game that may mean (for a given tech level) some new build is now the optimal build, but that only means that new build gets added to the "optimal builds blueprints book" (and maybe an old build which is now suboptimal gets tossed out) and then you're once again done - no more real designing/thinking required for all future games. (Or even no designing/thinking required ever - just download blueprints others have made.)

The alternate recipies are optimal depending on the resource availability. E.g. in a level with lots of iron you would go a different route than you would in a level that is lacking iron. When you have only little coal you might need to go via oil. There are only few resources and I think you are right. There is a good variation for replaying in single-player without downloaded blueprints but in multiplayer with many resources you need maybe 5 different sets of blueprints to cover most configurations.

Having a set of recipes that randomly change will result in much more variation, especially when the ingredients change and not only the ratios.

For example for a green circuit you could need:
* 1-6 copper wires
* 0-1 iron plates
* 0-1 wood (instead of iron plates)
* 0-0.1 copper plates
* 0-0.1 coal

Where the value for each item is pseudo-randomly chosen and 0 means that this item is not needed.

[doppelEben]

Having a set of recipes that randomly change will result in much more variation, especially when the ingredients change and not only the ratios.

For example for a green circuit you could need:
* 1-6 copper wires
* 0-1 iron plates
* 0-1 wood (instead of iron plates)
* 0-0.1 copper plates
* 0-0.1 coal

Where the value for each item is pseudo-randomly chosen and 0 means that this item is not needed.

So on a very lucky round you could spam you factory with green circuits because they only cost 0.1 coal? great

[TheRaph]

I see another way to come to a very similar result: We permanently introduce multiple alternative recipes to obtain the same item by different pathways.

Except that doesn't achieve the desired result at all.

This is the problem as originally stated:

So it is a battle, who has the best blueprints loaded from Internet, instead of who is the best factorio player.

If you consider the general form of that problem, it is not just a PvP problem but a single-player problem as well. ... All I need is a good set of blueprints and I can not only play the game without really doing any thinking at all, but I can play it near optimally that way, with my brain basically set to the "off" position.

Your suggestion, having alternate fixed recipes, does not change that situation.
Only by having the recipes (and/or other things that affect builds like assembler module count, assembler/belt/inserter/robot speed, assembler/beacon size, max robot/inserter capacity, power pole/substation range, which techs/items are available to research/build, etc.) change every single game does it result in the player really taking a fresh look at build designs every single game.

...

The alternate recipies are optimal depending on the resource availability. E.g. in a level with lots of iron you would go a different route than you would in a level that is lacking iron. When you have only little coal you might need to go via oil. There are only few resources and I think you are right. There is a good variation for replaying in single-player without downloaded blueprints but in multiplayer with many resources you need maybe 5 different sets of blueprints to cover most configurations.

Having a set of recipes that randomly change will result in much more variation, especially when the ingredients change and not only the ratios.

Where the value for each item is pseudo-randomly chosen and 0 means that this item is not needed.

Nice ideas ... but not the problem I originally stated. As NotABiter said my original state was about the competitiveness of that game.
If you allow blueprint library, a player who is a absolute newbie may beat up an "old" player just because he is better prepared in terms of had downloaded more blueprints. But he may never had created a factory design by his own.
I think it is like doping in sports - the question is what do I want to compare? Did I like to compare who is the strongest sports man or did I like to compare who has the best type of doping?

The easiest way to solve this is to get an option to disable all global blueprints for an specific competitive session. So every one is free to decide if he like to compare player or blueprint-libraries used by players (there are also truly differences between people who can do bp-libraries and those who don't).

The suggestion to have different recipes every time a game starts, effects booth - single player and multi player.
On the other hand, this suggestion sounds nice on the first hearing, but if I think about it, it may gets lack in terms of training or practice. You cannot train yourself how to setup things, but they alter to much in every game.

I think we should go a way in the middle. What about to have different recipes / ratios / efficiencies depending on what BIOME you're building in.

• So it seems to be consequential if furnaces located in hot desert biomes needs less energy / coal and be faster than in cold biomes.
• On the other hand, chips should be produced in cold but dry regions - if you locate a green chip assembler in a swampy biome it my need an extra ingredient as deoxidizer.
• Reciepies who needs whater may need more in dry biomes and less in swampy areas.
• Offshore pumps may placed on land-tiles if it is wet enough, working as a well.
• Steam generators should also run better in cold biomes because (as in reality) the generate energy (steam engine movement) out of difference between hot steam and colder surrounding.
• There should be an reason why copper cable assembler should not be affected by the same perk than the green chip assembler - so the ration copper-wire-assembler to chip-assembler (currently 3:2) should not be the same in desert as in swamp. (my be in desert the copper-cables will speed up and chip fabrication may slow down in swampy biomes it may be vice versa.

The whole benefit from that is, a player can learn and practice to have the best setup for every biome - and he will be forced to rethink his setup in terms of having different items produced on different places (connected by train) to be most efficient. But due to biomes a very random every map has its own optimum. (And large array blueprints may not fit in strange shaped biomes )

Someone may say against that: "It means only you need more blueprints (for every biome one set)." Yes that will help whit the ratios of starting bases (did I need copper wire / chip - ratio 3:2 or 5:2 or 1:1 you may name it "swampy base" or "desert base"). But to get most effective base, a player need to know in which biomes hecan produce which thing (and not only which setup is the best for a given biome) - and that knowledge you can't store in blueprints.
So in this case blueprint libraries may help start, and also may help later but the player need to get rid of it and how biomes works to do a good job.
And in this case only those player may win, who have the best of both skills, knowledge about mechanics and biomes of factorio AND knowledge about how to efficient use a bp-library.

But I don't like to hurt anybody, why not put in two new options - "no blueprint library" and "radom reciepies" - so you can switch it on for a given session - and on the other hand put a biome-depended-assembling in game which be on by default but my be switched of (like bitters).

[ske]

Having a set of recipes that randomly change will result in much more variation, especially when the ingredients change and not only the ratios.

For example for a green circuit you could need:
* 1-6 copper wires
* 0-1 iron plates
* 0-1 wood (instead of iron plates)
* 0-0.1 copper plates
* 0-0.1 coal

Where the value for each item is pseudo-randomly chosen and 0 means that this item is not needed.

So on a very lucky round you could spam you factory with green circuits because they only cost 0.1 coal? great

1 copper wire would be mandatory in this list, so without extra rules it would be 1 copper wire + nothing else.

I was not very clear, so 1 iron OR 1 wood (or both?) would be an additional rule that I'd choose.

You could make up extra rules as you like.

[ske]

What about to have different recipes / ratios / efficiencies depending on what BIOME you're building in.
...
Someone may say against that: "It means only you need more blueprints (for every biome one set)." Yes that will help whit the ratios of starting bases (did I need copper wire / chip - ratio 3:2 or 5:2 or 1:1 you may name it "swampy base" or "desert base"). But to get most effective base, a player need to know in which biomes hecan produce which thing (and not only which setup is the best for a given biome) - and that knowledge you can't store in blueprints.
So in this case blueprint libraries may help start, and also may help later but the player need to get rid of it and how biomes works to do a good job.
And in this case only those player may win, who have the best of both skills, knowledge about mechanics and biomes of factorio AND knowledge about how to efficient use a bp-library.

I don't think that this would have the effect that NotABiter wants. He wants a wide variety of combinations that no blueprint library can cover. As you already see, those five biomes don't cut it.

What you are talking about is something else that I'd like. Making the biome matter. Doing this on the recipe level while keeping the buildings the same would be very confusing for many players, though. But you could do it on the building-level where you cannot place certain buildings on certain biomes. I.e. the swamp cannot carry heavy structures, you have to build them on solid ground. Level 1 assemblers are light but level 2 and up are heavy and need better ground to support them.

[Kyralessa]

Please help us make the right choices in the graphics settings by adding more descriptive tool-tips and for every option!

Yes! I can't figure out what half of the settings are or what effect they'll have except by tedious experimentation...and sometimes not even then.

To add to the idea above: Have a few presets to make things easier. "Optimize for speed", "Optimize for graphics quality", etc.

Or maybe just a couple of buttons: [The game is too slow] [The game is too ugly]

The first would reduce graphics quality, whereas the second would improve it at the expense of speed.

[TheRaph]

Doing this on the recipe level while keeping the buildings the same would be very confusing for many players, though. But you could do it on the building-level where you cannot place certain buildings on certain biomes. I.e. the swamp cannot carry heavy structures, you have to build them on solid ground. Level 1 assemblers are light but level 2 and up are heavy and need better ground to support them.

About "being confusing" - I don't think so if it is properly implemented. But if poorly implemented it would be confusing either it is on production level or on building level. I imagine the questions her in forum like "why can i sometimes place buildings and sometimes not, is it a bug?". The same may happen with different raw material/energy consumption as well - but nos as early in learning the game.

So I remember my first beginnings: I don't start to calculate what needed to craft certain things. I build a green circuit assembler and a copper wire assembler, then I wait. Some time later I registered that circuit assembler stops and see that it start to wait for missing copper wires, so I add a second copper wire assembler. Eventually I see a little bit later, that there are to much copper wires and find out, that there might be a 3:2 ratio.
As I start to get more knowledge of factorio and seen some let's plays on YT, I first notice the tooltips which shows what are needed resources and what time does it need to produce. Later I notice the sidebar with given informations about effectiveness and so on.

If the ratios on different biomes are different, means a newbie will take a little bit longer to understand whats happen ther, but there are two things that may help:

• Fist (in my opinion they should implement it right now) there should be the current needed time in tooltip inside an assembler. So if inventory tooltip tells me 2 copper cables need 0.5 seconds, the tooltip in gray assembler should show me 1.0 sec instead. For blue it should show me 0.7 sec and for yellow it should display 0.4 sec.
• In sidebar there should be the real crafting speed and (+xy%) or (-xy%) behind that. And then there might be an extra line for biome-for-reciepe modifier, which tells you the gain for current recipe in current biome.

So e new player may read this and gets behind the rules. Or he don't and have to try and error every time he build up a factory (but it is possible to riddle that out). But he can build and it will work (not as effective as it could but it still works). In case of yours, a new player like to place a specific type of building on ground and it don't work because of wrong biome ... that may be frustrating. Even if there is a message "wrong biome for this building" it might be frustrating because he might have build entire infrastructure (belts inserters etc.) and then can not place assembler. He has to deconstruct every single item (in early game are no bots!) and have to build up a new setup in an other place.

For my style of gameplay it is nice to have to riddle things out. Most things in minecraft (and especially in factorization addons) are not as well documented in-game as in factorio (therefore a big thumbs-up for the factorio dev team!), but I've played a long time hard core modded MC - just to do some "science" and riddle out how mechanics work. So biome depended speed and material cost in factorio would not over complicate it.

[Tricorius]

... snip ... (about 8 really complex posts)

So, it starts. Balancing for “PvP” could have some pretty drastic complexity implications. I think the biome complexity discussion would be *very* confusing for new users. It’s an interesting idea, and those small tweaks to crafting times (overall) probably wouldn’t matter much to a new player in the first few games. But all of that to try to “balance” a multiplayer “issue” is...I’m going with interesting.

[TheRaph]

I’m going with interesting.

Just a bit of brain workout. I don't think they will ever implement it like suggested (if they will I'll be like a happy small child jumping up and down trough my living room), but I like to think about what is possible and what may not.
And I like to have here people who also like to discuss the pro and cons seriously. I think there are great ideas in between and devs reading it may find new ways to think about their game.
I like this more than that fiery discussion about belts and bots.

[NotABiter]

The alternate recipies are optimal depending on the resource availability. E.g. in a level with lots of iron you would go a different route than you would in a level that is lacking iron.

I concede the point - when there is sufficient pressure on a resource (despite them technically being infinite -- i.e. some factor such as time, distance, biters, etc. makes acquiring more of the resource difficult, at least in the short term), then even just having fixed alternate recipes which favor different resources will result in the player having to think. However, that thinking is then just between "blueprint A" and "blueprint B" (or maybe "blueprint C") to suit the condition.

Having a set of recipes that randomly change will result in much more variation

Yes, and I think there's some critical point where there is sufficient variation that just picking the best blueprint from some set (rather than doing actual design) becomes suboptimal/infeasible enough that players would choose to actually do design. The game should support at least that much variation.

I was thinking about how this should work, and it's important to recognize that while variation is good, people also like familiarity - they like to take their knowledge from previous plays and apply it, and that would remain true even if some things were different so they had to do at least some new builds if they wanted to be anywhere near optimal. So it occurred to me there could be a "randomness setting" that goes from 0 to 1. At 0 you get vanilla as it exists now. At 1 everything is randomized (chaos!). Somewhere around 0.05 though, you get a game that has most of vanilla's existing recipes (and vanilla's existing ranges/speeds/resistances/etc.), but a few crucial things have changed (to some random input ingredient, to some random ingredient amount, etc). Those changed values could even be highlighted in the research and crafting UIs in some other color so it's easy for veterans of the game to see what the differences are this game. (There could also be a new summary screen which just lists all the differences for the current game.) As an example, just think what happens to all of your "mega factory" builds if just one critical setting changes, such as beacon range. It would only take 3 or so such major random tweaks (selected from a much larger pool of possible tweaks) to make each game involve significantly unique builds and at the same time let the player reuse most of their existing Factorio knowledge (and not feel nearly as lost as someone playing Bob's or Yuki's mods for the first time). Add in a few minor tweaks as well for flavor, and I think the experience would be a good one.

The easiest way to solve this is to get an option to disable all global blueprints for an specific competitive session.

Your solution only changes it from "use premade blueprints" to "use memorized builds". It still does not require any actual thinking during the game. Furthermore, it does nothing to address the greater problem of the game being highly repetitive which includes single player. (Sadly, I often find myself using the "use memorized builds" method, because there's simply no in-game reason to make a new build. So far I haven't really built up a blueprint library, probably in part because doing so would be admitting that Factorio will never be fixed and will always be repetitive.)

The suggestion to have different recipes every time a game starts, effects booth - single player and multi player.

Yes, and that is the point. I don't care if PvP gets fixed - it doesn't affect me. I would really like to see single player fixed though, and single player suffers from the same repetitiveness as you are complaining about in PvP. (Though now I see that apparently you're not at all bothered by the repetitiveness and you just want people to memorize builds, and only for PvP which you don't even play.)

You cannot train yourself how to setup things, but they alter to much in every game.

Yes, and that is the point. I don't want to just use memorized builds (or watch others use memorized builds). I want to have good in-game reasons for designing new builds. It need not affect every single build in the game (see my idea above for a randomness slider and where I suspect the sweet spot is), but it should significantly affect a number of builds so I have something new to do in every game.

What about to have different recipes / ratios / efficiencies depending on what BIOME you're building in.

While I find those ideas to be interesting (and also ske's take on the biome idea), they don't really address the issue at hand. (I.e. they can be solved using multiple blueprints. Furthermore players will tend to locate each build in "prime biome" for the given item being produced in the mid to late game.)

Also note that there's a bit of a problem here (and with my earlier idea of making landfill and cliff explosives prohibitively expensive), and that is it gets in the way of scaling your factory in size via automated construction. I'm facing this issue now in my current play through - I've been going solar (no nuclear) with massive beaconing (large power requirements), and I would really like to have a nice automated system for sending trains full of materials to the solar area and having robots take those materials and expand not only the solar itself but all of the solar-creation infrastructure (tracks, stations, unloading, roboports, etc.). But those damn cliffs! I have to go out there and get rid of cliffs by hand before any of that can work, which means I can't scale the whole problem up with automation because there is a required manual step. Having builds change every time they hit a biome boundary would present similar issues. (Actually, solar is probably not an ideal example of the problem because in many cases cliffs won't matter much - most the time it just means some solar panels or accumulators don't get placed. But when it gets in the way of trains or robotports it causes major problems.)

The whole benefit from that is, a player can learn and practice to have the best setup for every biome

Ugh, no. You're almost explicitly saying everyone should just memorize the best builds and then use them like mindless zombies. That's boring. I want to 1) come up with totally new (for me anyways) builds that are optimal for the given game, and then 2) scale those solutions up via automated construction. And the next game I want to do it all over again. The only way I can do #1 is if what is "optimal for the given game" changes significantly from game to game.

(And large array blueprints may not fit in strange shaped biomes )

Right, which like I said above, means it breaks scaling up your factory via automated construction.
(And like I said, I'm of two minds on that - it does increase thinking not in terms of the builds themselves but in terms of logistics, but breaking automated construction is a big price to pay for that. Frankly, Factorio's logistics story is not good - the relative scale between train stations and resources/builds tends to be very bad and completely unrealistic. In the real world it would be like having simple train stations that are 100 square miles in size. And that makes having these types of logistics issues inserted into the game less fun than they could otherwise be. Plus, your original complaint, and my generalization of it, was about the builds themselves - logistics is a whole different topic.)

But to get most effective base, a player need to know in which biomes hecan produce which thing (and not only which setup is the best for a given biome) - and that knowledge you can't store in blueprints.

Again, you are talking about memorization and recall as if that is compelling gameplay. It's not.
Furthermore, you're simply mistaken about "that knowledge you can't store in blueprints" - you can most certainly store that knowledge in blueprints and blueprint books via a simple naming/organization convention.

why not put in two new options - "no blueprint library" and "radom reciepies"

Because the "no blueprint library" option already exists? Any player can choose to not have or not use a blueprint library for a given game. (And PvP games can have house rules - I'm not a follower of such games but surely there have been PvP games where they have house rules like "no solar", "no lasers", "no logi bots", etc.? Single players do such things all the time, as I have - it's one of the few ways to get real variation in vanilla.)

[Avezo]

This and recent FFF makes me think that weather really needs to be implemented prior to 1.0. At first I thought it would be better off left for future expansions with different planets and whatnot, but dynamic weather (and other planetary conditions, gravity, daylight time, etc.) seems to solve several issues, i.e.:
-Could cause dynamic changes in recipes in some way (maybe even more than recipes, i.e. varying daylight time making default solar stamp blueprint unreliable)
-Could affect bot travel, effectively nerfing them in an interesting way (and also not divide playerbase so much since weather could be disabled or adjusted in settings)
-Could add up to interesting events you need to fight against, instead of just biters (think of weather disasters in Sim City, except there's a tornado or lightning strike destroying your base, or even others, i.e. solar storm disabling entire electric network like in real life)