Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Post
  • Reply
M_Gargantua
Oct 16, 2006

STOMP'N ON INTO THE POWERLINES

Exciting Lemon
Maybe make it a earlier big sink in an improved iron plate or iron gear casting recipe?

Adbot
ADBOT LOVES YOU

Bhodi
Dec 9, 2007

Oh, it's just a cat.
Pillbug

seiken posted:

It might be good to have another early calcium sink but I'm not sure what. I tried to balance things with science production in mind so if you're only making science the amount of calcium you need changes through the game (the idea is to encourage automating switching between refined and non-refined metal smelting depending on your calcium status).

The main advantage of vehicle fuel is you don't need the cube to produce it and it means trains won't be drawing from the same source as your cube power station. Maybe that's not useful enough to be worth bothering with but I thought not having it could be annoying if you were going for a big train base.

:success:
I assume everyone did what I did, which is add in refined smelting with the current smelting as a backup, and then never have your refined smelting run. It's not that huge of an issue but no one likes seeing stationary belts for an entire science level. So we're looking at a sink for green but only green science because you don't want a steady calcium drain in the later game which would require more rebalancing. A few options as I see it:
  • Add concrete as a requirement to an intermittent/manual craft (such as a building) to encourage people to build out a few concrete fabricators in early-mid green and start stockpiling concrete for hand use and that would keep the calcium flowing (and, since you're stockpiling it, might as well pave the factory every time your calcium stops)
  • Add a new half-efficient recipe that's used within early green but not as efficient as later recipes. Maybe refined tar, since balancing recipes come later and I remember being very short of tar and having to void essential oils tanks twice.
  • Move a current tech earlier in the tree that uses calcium - Vehicle fuel would fit this but there's still little reason to use it at that tech level
  • Add a straight sink tech similar to sand (meh)
  • Do nothing, it's not that big a deal
I would definitely look at option one. If I had to pick a building, I'd look at the chemical plant or greenhouse. Speaking of greenhouse, it feels very off that it's build-and-forget it with nothing but Water+Power->Wood/Potato, I feel like there's some meat here left uneaten. It doesn't need to be some complex chain but the pack has carved away unnecessary tedium in all other places so this feels like a left over piece. This is a void asking for more (non-cube) complexity of a different sort. Maybe something focusing on fluid balancing or tanking or pipe or high item throughput? I need to think on it.

RE: vehicle fuel, that explanation makes sense late game but then they'll be using that better secondary vehicle fuel recipe (I think? I haven't unlocked it yet). Plus, condensed fuel is so energy dense I can't imagine vehicles being much of a drain at all, but I'm not there yet.

I'm midway though blue science now and my first reaction is that it's got the same "oil refining" issue as the main game that a lot of people have - too many things in too many different directions offered at the same time is a bit overwhelming, but I'll write up something more when I'm through it. I like the new cube split mechanic though! And putting a cube in your car is hilarious, extremely fun and satisfying and was needed because it took me ages to actually find an imersite vein. I just wish it could jump cliffs (Renai style)

Bhodi fucked around with this message at 17:45 on Jan 16, 2024

celestial teapot
Sep 9, 2003

He asked my religion and I replied "agnostic." He asked how to spell it, and remarked with a sigh: "Well, there are many religions, but I suppose they all worship the same God."

Harvey Baldman posted:

I have a question - it feels like there's a smart solution to my problem but I might not have enough experience with the game to recognize it yet.

Your unloading setup is the safe and standard version but it's over-engineered for what you're trying to do here, and that's making this problem look more complex than it otherwise should be.

Since your end goal is mixed belts you could unload only half belts, which makes the mixing simpler. Here's one implementation I just threw together to try to illustrate the idea.

This also has the benefit of using fewer inserters and splitters. In your screenshot many of these stack inserters will spend most of their time idle, especially once you have maxed the stack size. You don't need that many green inserters to flood a red belt!

Only registered members can see post attachments!

celestial teapot
Sep 9, 2003

He asked my religion and I replied "agnostic." He asked how to spell it, and remarked with a sigh: "Well, there are many religions, but I suppose they all worship the same God."
By the way, one green inserter can fully compress half of a red belt all by itself, if you have it unloading directly onto a splitter!

Only registered members can see post attachments!

LonsomeSon
Nov 22, 2009

A fishperson in an intimidating hat!

thanks i hate it

M_Gargantua
Oct 16, 2006

STOMP'N ON INTO THE POWERLINES

Exciting Lemon
Imagine being so inefficient.

celestial teapot
Sep 9, 2003

He asked my religion and I replied "agnostic." He asked how to spell it, and remarked with a sigh: "Well, there are many religions, but I suppose they all worship the same God."
I'm not trying to shame anyone - sorry if that came off wrong.

Majere
Oct 22, 2005

by Fluffdaddy

(and can't post for 11 years!)

Factorio is 90% shameful designs.

The Locator
Sep 12, 2004

Out here, everything hurts.





Majere posted:

Factorio is 90% shameful designs.

*Checks old screenshots*

Confirmed:

Harvey Baldman
Jan 11, 2011

ATTORNEY AT LAW
Justice is bald, like an eagle, or Lady Liberty's docket.

I'm slowly trying to level up my factory designs and I've discovered I really like how Space Exploration handles beacons and modules. Any areas where there are machines being hit by more than one beacon causes that machine to jam up, so you have to actually think about how to maximize your beacon effectiveness versus just dogpiling as many beacons around your production lines as possible. I have been finding it really enjoyable to try and compact my production lines down around beacons as much as possible with Bob's Adjustable Inserters.











I know space is basically unlimited, but I'm building within the parameters of city blocks I've designed so part of the fun for me is trying to build these out as compact as possible. I only have red belts unlocked because I haven't even gone to space yet in my SE run, so I'm sure I'll have to revisit some of these designs later with improvements, but I'm probably gonna enjoy that as well.

LonsomeSon
Nov 22, 2009

A fishperson in an intimidating hat!

celestial teapot posted:

I'm not trying to shame anyone - sorry if that came off wrong.

I want to be super clear: “thanks I hate it” from me at least is always intended as a complement

KillHour
Oct 28, 2007


LonsomeSon posted:

I want to be super clear: “thanks I hate it” from me at least is always intended as a complement

thanks I hate it

Harvey Baldman
Jan 11, 2011

ATTORNEY AT LAW
Justice is bald, like an eagle, or Lady Liberty's docket.

celestial teapot posted:

By the way, one green inserter can fully compress half of a red belt all by itself, if you have it unloading directly onto a splitter!



gently caress, I have a few designs I have to go back and poke at with this approach in mind now, this is so much easier. I thought I was a genius when I rigged this up:

Chin Strap
Nov 24, 2002

I failed my TFLC Toxx, but I no longer need a double chin strap :buddy:
Pillbug
My whole attempt at big rear end city bricks isn't working for me. Don't like how huge they are. Next time I'm going to try the more city block sized things and just steal intersections from someone else. What city block size do you guys like?

uPen
Jan 25, 2010

Zu Rodina!

Chin Strap posted:

My whole attempt at big rear end city bricks isn't working for me. Don't like how huge they are. Next time I'm going to try the more city block sized things and just steal intersections from someone else. What city block size do you guys like?

The area covered by 4 roboports in a square. Smaller than that feels too small and more gets BIG really fast.

Harvey Baldman
Jan 11, 2011

ATTORNEY AT LAW
Justice is bald, like an eagle, or Lady Liberty's docket.

I am by no means an expert and I have seen people go way bigger than this, but I have been working around a city block size that uses roboport spacing as its foundation. 3 Roboports for interior coverage, plus on the corners. I use the roboports to help align my block placement. I like this because it's just enough that I can see the whole block on my screen.



My baby base is in the top left but I started laying out blocks before I went into space (doing Space Exploration this run) so that I'd have a good base to operate from. Everything's serviced by 1-4-1 trains (both engines facing the same direction, though) and I made little maintenance corridors so they could pull out of the main thoroughfare and refuel/wait for dispatch. Haven't noticed much congestion on my 4-way intersections, too, though that may get worse as I continue scaling up.



If I really had my head on straight I'd have grouped my blocks up a bit better by product, but oh well. I can make XL blocks like what's on the right side with a little bit of adjustment, which I imagine will be where I drop most of my rocket launching and landing pad stuff.

My in-progress blueprint book, if you want to see:

https://factoriobin.com/post/KXavg9cx

The early designs suck but I'm understanding how to do things better as I go. Once I get blue belts and beacons I'll probably go back and remake a bunch of stuff like the green circuit production blocks. I really like my nuclear block - was able to design it with a big bank of auxiliary steam storage that I can flip a switch on if I need more power, like when I have to defend Nauvis from a Coronal Mass Ejection.

Radiation Cow
Oct 23, 2010

So I've finally conquered my restartitis and launched a rocket in vanilla. I'm still keen to keep going, which means megabase but I'm not sure how to manage the transition.

Does anyone have any best practices on how to make the transition as smooth as possible? I've noticed that my bottleneck is modules and circuits, but I'm running four blue belts of iron and copper and don't have much room to expand the bus, so I'd need to rejigger the entire base to make space.

Almost tempted to leave this base running and build up somewhere else but I'm not sure if there's sufficient infrastructure to support it.

Tamba
Apr 5, 2010

Radiation Cow posted:

So I've finally conquered my restartitis and launched a rocket in vanilla. I'm still keen to keep going, which means megabase but I'm not sure how to manage the transition.

Does anyone have any best practices on how to make the transition as smooth as possible? I've noticed that my bottleneck is modules and circuits, but I'm running four blue belts of iron and copper and don't have much room to expand the bus, so I'd need to rejigger the entire base to make space.

Almost tempted to leave this base running and build up somewhere else but I'm not sure if there's sufficient infrastructure to support it.

Megabases are usually built far away from your spawn, because the ore patches get richer the farther out you are.

Jabor
Jul 16, 2010

#1 Loser at SpaceChem
Basically the answer is "ditch the bus". A bus is great when you don't know what you're going to be building next and appreciate the flexibility of being able to tap off whatever the next thing needs, build the thing, and then route it back on to the bus.

At megabase scale, you already know what everything needs, so you don't need the flexibility - and the fact that the bus has a huge bottleneck right at the start where you're inputting the raw materials is a big problem.

If you're just scaling up to a little bigger than you are now, you can kinda patch around it by injecting more raw materials partway through the bus - but usually the better plan, since the amount of stuff you've already built is pretty small compared to the amount that you will be building, is to take the opportunity to radically redesign. Set up your existing base to churn out belts, inserters, chests, assemblers, furnaces, roboports, bots, rails, concrete etc. to give you the resources to build the base of your dreams.

Xerol
Jan 13, 2007


One thing that's very easy to get sucked into is converting your first rocket build into a many rocket build which is generally not great for long-term plans; having a couple more bot speed or mining prod upgrades seems great in the short term but that temporary build will easily and invisibly turn into a permanent build. Finish researching all non-space tech, then tear out any science production. Divert the resources to building things for expansion, I like to group this into a few categories:

-Expanding your rail network: Rails, engines, cars, tanks, pumps, inserters, chests, power poles, lamps.
-Building new science production: assembler 3s, beacons, inserters, modules. So many modules. You may even want to consider building new resource streams devoted just to making modules.
-Logistics: Belts, bots, roboports, chests. If you're going belt-based, an entire facility devoted just to blue belts is a must. It's almost entirely iron except for a very small amount of red + green chips and of course the lube. The circuit needs are small enough (just needed for splitters) that you can feed those from bots while moving the bulk of the iron via belt.

Bettik
Jan 28, 2008

Space-age Rock Star

Radiation Cow posted:

So I've finally conquered my restartitis and launched a rocket in vanilla. I'm still keen to keep going, which means megabase but I'm not sure how to manage the transition.

Does anyone have any best practices on how to make the transition as smooth as possible? I've noticed that my bottleneck is modules and circuits, but I'm running four blue belts of iron and copper and don't have much room to expand the bus, so I'd need to rejigger the entire base to make space.

Almost tempted to leave this base running and build up somewhere else but I'm not sure if there's sufficient infrastructure to support it.

I am hardly an expert but the approach I’ve both seen and taken once myself is to use your current base to build a bigass iron and copper mine, some smelting facilities (fed by trains from those mines) for end game quantities for both (probably built with modules in mind, even if you don’t currently have enough) and then a green circuit build fed from the results of those smelting facilities. From there it’s up to you, probably a bigass power thing, some red circuits, large scale module production, etc.

Xerol
Jan 13, 2007


Re: City blocks, I hated most designs until I changed perspective and started running rails up the middle of blocks instead of around the edges. Then rails only get placed where they need to go, not everywhere, and if a build needs more than one block to fit there's not going to be any rails in the way (I do try to keep the space between blocks open for walking/driving, but that's a 4-tile brick/concrete path that's easy to underground through).

My blocks are 128x128 with 3x3 (minus the center) roboports in a ring, which doesn't give complete logistics coverage to the middle of a block, so I'm designing my next generation with 192x192 blocks and an even number of roboports aligned to cover the whole block in a 4x4 spacing. This will also let me upgrade to 8 or 16 wagon trains (everything currently designed for 1-4 and that's becoming inadequate).

The only downside is I'm going to have to design a new solar block; my current 128x128 solar block is perfectly ratioed with exactly 1,000 solar panels. Well and I'm also going to have to redesign literally everything else too.

seiken
Feb 7, 2005

hah ha ha

Bhodi posted:

I assume everyone did what I did, which is add in refined smelting with the current smelting as a backup, and then never have your refined smelting run. It's not that huge of an issue but no one likes seeing stationary belts for an entire science level. So we're looking at a sink for green but only green science because you don't want a steady calcium drain in the later game which would require more rebalancing. A few options as I see it:
  • Add concrete as a requirement to an intermittent/manual craft (such as a building) to encourage people to build out a few concrete fabricators in early-mid green and start stockpiling concrete for hand use and that would keep the calcium flowing (and, since you're stockpiling it, might as well pave the factory every time your calcium stops)
I would definitely look at option one. If I had to pick a building, I'd look at the chemical plant or greenhouse. Speaking of greenhouse, it feels very off that it's build-and-forget it with nothing but Water+Power->Wood/Potato, I feel like there's some meat here left uneaten. It doesn't need to be some complex chain but the pack has carved away unnecessary tedium in all other places so this feels like a left over piece. This is a void asking for more (non-cube) complexity of a different sort. Maybe something focusing on fluid balancing or tanking or pipe or high item throughput? I need to think on it.

I think adding concrete as an ingredient to greenhouses or chemical plants would be ideal if it didn't push tar production further down the green tree than I'd like and make progression there more rigidly linear. I will have a think though.

You're bang-on that wood and potatoes mostly exist because they made thematic sense as ingredients for other processes I wanted to have, but their actual production isn't very interesting and could definitely use a revamp at some point.

Chin Strap
Nov 24, 2002

I failed my TFLC Toxx, but I no longer need a double chin strap :buddy:
Pillbug

Xerol posted:

Re: City blocks, I hated most designs until I changed perspective and started running rails up the middle of blocks instead of around the edges. Then rails only get placed where they need to go, not everywhere, and if a build needs more than one block to fit there's not going to be any rails in the way (I do try to keep the space between blocks open for walking/driving, but that's a 4-tile brick/concrete path that's easy to underground through).

My blocks are 128x128 with 3x3 (minus the center) roboports in a ring, which doesn't give complete logistics coverage to the middle of a block, so I'm designing my next generation with 192x192 blocks and an even number of roboports aligned to cover the whole block in a 4x4 spacing. This will also let me upgrade to 8 or 16 wagon trains (everything currently designed for 1-4 and that's becoming inadequate).

The only downside is I'm going to have to design a new solar block; my current 128x128 solar block is perfectly ratioed with exactly 1,000 solar panels. Well and I'm also going to have to redesign literally everything else too.

I like the prospect of not trying to run rails around the edges. Might adopt that next time. But 192x192 seems huge to me. My city bricks are 128x192 already and it feels too large and not modular enough. But I'm having more problems with rail than the sizing so maybe that would help fix things for me.

Really want to just power through my first K2 run before trying out something else. I have almost all non infinite techs researched I think. Just going to limp along to the end with a spaghetti second base with too many bots I guess.

Logistics bot question. How do you actually increase the bandwidth of moving stuff around for bots? I continually seem to have requests that are short, but plenty of supplies and idle logistics bots and I don't get it.

Xerol
Jan 13, 2007


Keep in mind I'm trying to make a multi-kspm belt base so blocks of this size probably aren't needed for most people. This design easily fits 8-lane smelter blocks (at the top), but I'm having problems fitting things like a full yellow belt of LDS into one block (bottom).



Beaconed builds will obviously not need as much space (even just going to t3 assemblers will reduce it by a lot). I also just want to have more space to run belts even if the builds take the same amount of space; running belts from another block into multiple smelter blocks starts to get really cramped. Better planning will help, but it's still nice to have the space to make belts neater.

192 will also make the roboport coverage less dense, which in this case will leave more room for builds (right now fitting 3 into 128, the spaces between them aren't big enough to fit 2 adjacent beaconed lanes of production). I want a system where I never have to move or remove a roboport to make builds fit.

e: Looking through a bunch of screenshots I'm also reminded of the other reason for wanting to go bigger, and that is to be able to fit more stations and bigger stackers into a single block. That 2-station, 6-stacker design in the bottom left is about as big as I can fit in a block, and this either fits two 1-4 or one 2-8 train per station lane, but if I have a 2-8 train in each station lane then I have difficulty fitting the output belts. And if I want one station block to be used for all of a single item's production, if that item has more than 2 ingredients I can't fit all of that in one station block.

Chin Strap posted:

Logistics bot question. How do you actually increase the bandwidth of moving stuff around for bots? I continually seem to have requests that are short, but plenty of supplies and idle logistics bots and I don't get it.

Most likely your network is too large. First thing is to make sure you have enough available logistics bots, then make sure your requests are large enough (and then go back to step 1). Bots will also pull from storage chests before passive providers, so if you have a lot of the item in your central storage which is farther away than your providers, bots are going to take longer to deliver. You could also try adding intermediate buffer chests so each individual bot trip is shorter, requiring fewer stops to recharge. You could also isolate the final network, using inserters to move items from a requester on the main network across a 1-tile gap into providers on the isolated network, then bots on that network don't have very far to go at all (still has the issues of latency/buffer size on the main network, but you've split the problem into two smaller problems).

e2: Found an (old) example of grid isolation.



If I was doing this now I'd just run belts through the gap and load into providers as close to the assemblers as possible.

Xerol fucked around with this message at 16:01 on Jan 17, 2024

Chin Strap
Nov 24, 2002

I failed my TFLC Toxx, but I no longer need a double chin strap :buddy:
Pillbug
I'm just in indecisive mode after abandoning my starter base and trying to start something proper I guess for my K2 run. Messed around a lot building a bot mall and mining all ore types into buffer warehouses but I don't want to megabase this. I just need to bite through the rest of the science and get the endgame and I think when I start over I'll keep it into consideration more.

But that is what is just pointing me to bot logistics to get to the end, it just isn't fast enough.

Chin Strap
Nov 24, 2002

I failed my TFLC Toxx, but I no longer need a double chin strap :buddy:
Pillbug
So with regards to running rails through the middle of a block does that mean you have a block that is just a NS straight and another that is an EW straight, and different turns and intersections? Isn't that a super empty block then?

Phobeste
Apr 9, 2006

never, like, count out Touchdown Tom, man
You can also use it for stations, which is a great idea. A problem with city blocks is that train stations can take a lot of space, so for especially modded runs that have some recipe stacks with a billion ingredients that need a billion inputs and a billion train stations, your blocks that are a reasonable size for a vanilla smelting stack can now have very little room for actual machines once you have a bunch of 8 car stations and stackers

ymgve
Jan 2, 2004


:dukedog:
Offensive Clock
I've never had trains and building beyond the first rocket launch "click" for me, so for me the game beyond winning is mods. Keeps the game fresh and interesting instead of having to actually plan and think at a higher level.

meowmeowmeowmeow
Jan 4, 2017

Chin Strap posted:

I'm just in indecisive mode after abandoning my starter base and trying to start something proper I guess for my K2 run. Messed around a lot building a bot mall and mining all ore types into buffer warehouses but I don't want to megabase this. I just need to bite through the rest of the science and get the endgame and I think when I start over I'll keep it into consideration more.

But that is what is just pointing me to bot logistics to get to the end, it just isn't fast enough.

I'm at a point in my k2 run where I've got yellow and purple science going and am feeling the 'arg gently caress I didn't leave enough space here either' that makes me want to rebuild things with more space but drat that seems like a huge amount of work to move laterally and I feel like I'll have the same issue in another tier of science or two when I have another bunch of new stuff unlocked...

Currently rebuilding my spaghetti mall into something bot based thats a lil more flexible and then going to re-evaluate, my smelting area is cramped and the start of the bus is a nightmare of trying to thread new ingredients into the bus but I'm finally getting bots really running beyond personal roboport construction bots which should make big teardowns a lot easier to stomach.

SettingSun
Aug 10, 2013

I'm 150 hours into a co-op space ex game and I've basically internalized outpost and secondary base creation. Our original base on Nauvis is a giant spaghetti monster with a nightmarish train network operating at the lowest efficiency possible, but exoplanet construction is clean, efficient, and sustainable. I've applied many of the concepts I've learned into my solo games.

Xerol
Jan 13, 2007


Chin Strap posted:

So with regards to running rails through the middle of a block does that mean you have a block that is just a NS straight and another that is an EW straight, and different turns and intersections? Isn't that a super empty block then?

I just have a "straight" block that I rotate if I want it to go the other direction. I also have a "straight" block that goes around the middle instead of straight through in case there's ores or oil in the way. And then I have two 90-degree blocks (one going through the middle, one at a 45 degree), a T-junction, and a 4-way. So these are mostly empty blocks, yes, but they are generally around the edge of my base, not running through the middle. In general everything flows inwards, from resource outposts outside the main base to smelters on the edge that feed inwards to science assemblers and then in from there to the labs.

Most of these pieces are in use here, you can see on the left and bottom it's mostly raw material inputs, and the top right is a kludge. The darker lines are stone/concrete paths between the blocks.



Having sparse spacing like this is also a benefit for your rail network - T-junctions are always guaranteed to be far enough apart to avoid deadlocks, so I can just have a single junction design that has a regular signal on exit.

Phobeste posted:

You can also use it for stations, which is a great idea. A problem with city blocks is that train stations can take a lot of space, so for especially modded runs that have some recipe stacks with a billion ingredients that need a billion inputs and a billion train stations, your blocks that are a reasonable size for a vanilla smelting stack can now have very little room for actual machines once you have a bunch of 8 car stations and stackers

And what I do here is the station block is devoted entirely to being a station (and stacker, since that also fits), and production blocks are dedicated entirely to being production. So there's (almost*) never any combination of functions, the station block unloads onto the belts and the belts then run to another block. I have some blocks that are entirely dedicated to running belts through to make production blocks easier to design (these are basically mini-bus blocks).

*The main exception is where I have active mines, I will generally build around them but sometimes it's hard to avoid it and so sometimes a station block loses a couple stacker lanes while a patch gets mined out.

Chin Strap
Nov 24, 2002

I failed my TFLC Toxx, but I no longer need a double chin strap :buddy:
Pillbug

meowmeowmeowmeow posted:

I'm at a point in my k2 run where I've got yellow and purple science going and am feeling the 'arg gently caress I didn't leave enough space here either' that makes me want to rebuild things with more space but drat that seems like a huge amount of work to move laterally and I feel like I'll have the same issue in another tier of science or two when I have another bunch of new stuff unlocked...

Currently rebuilding my spaghetti mall into something bot based thats a lil more flexible and then going to re-evaluate, my smelting area is cramped and the start of the bus is a nightmare of trying to thread new ingredients into the bus but I'm finally getting bots really running beyond personal roboport construction bots which should make big teardowns a lot easier to stomach.

I will say having just been there that all the advanced stuff (furnaces, assemblers, chemical, inserters, etc) can lead to some beefy output with relatively little input. Especially if you module it, you can pump out a lot of gears straight from enriched iron using just a single advanced assembler.
So maybe you can upgrade more compactly than you thought. Or when you start over from scratch it doesn't have to be a huge build to start, it can scale upward.

EDIT: Also for K2 you can make a bus stretch a lot further than in base thanks to tier 4 and 5 logistics. Suddenly your one blue belt of whatever can become 2 blue belts worth if you upgrade it to purple, without having to move a thing.

Chin Strap fucked around with this message at 17:39 on Jan 17, 2024

meowmeowmeowmeow
Jan 4, 2017
Thats good to hear, one of the next 'productive' things on my todo list is to get ore enrichment running, needing two belts of ore for a single belt of plates is a pain, getting that to 3 ore for 2 plates sounds much better. Haven't really looked at any of the advanced assemblers or chem plants yet, I'll take a look.

The scale of things keeps surprising me, like the number of electrolysis plants to get enough O2 to make rocket fuel for yellow science, that kinda stuff.

My initial rail network is 1-4 trains for ore, thinking of adding some 1-1-1 or 1-2-1 double direction trains for shuttling around smaller amounts of goods between parts of the factory when I can.

Bhodi
Dec 9, 2007

Oh, it's just a cat.
Pillbug
I've got thoughts! But I've been away creating something absolutely cursed, (thanks to seiken) which I'll show off when it's fully working.

seiken posted:

I think adding concrete as an ingredient to greenhouses or chemical plants would be ideal if it didn't push tar production further down the green tree than I'd like and make progression there more rigidly linear. I will have a think though.

You're bang-on that wood and potatoes mostly exist because they made thematic sense as ingredients for other processes I wanted to have, but their actual production isn't very interesting and could definitely use a revamp at some point.
so far IMO the progression has been very well designed so, even if I think early green is fairly linear already, trust your instincts.

Chin Strap posted:

Logistics bot question. How do you actually increase the bandwidth of moving stuff around for bots? I continually seem to have requests that are short, but plenty of supplies and idle logistics bots and I don't get it.
bots have theoretically infinite throughput, but the more volume you need and the further away it is the more material you will have suspended in transit "in the air" and unusable. If you're short and you have plenty of supplies, you have to request a higher volume. It's called Little's Law if you want to read on it.

If you want to mess with grid isolation but local bot logistics IMO, you should try heavily using factorissimo instead and isolate within each individual factory. Unless you're doing a "no bot" run, be aware that grid isolation entirely removes the ability to get materials across cells without using the train and is kneecapping yourself. The "edge transit box" does not scale to a city/grid system and without a mod that allows you to toggle between "local" and "remote" logistic networks, this is very much an either/or situation so it's important to understand what you are sacrificing if you want to use isolated networks.

Chin Strap posted:

My whole attempt at big rear end city bricks isn't working for me. Don't like how huge they are. Next time I'm going to try the more city block sized things and just steal intersections from someone else. What city block size do you guys like?
tl;dr if you don't want to do your own math and considerations, go with 1-2 trains and either a 2x2 or 3x3 roboport size with 2 lanes and parallel on/off spurs. It's a good balance between size and usability to start with and the throughput is enough to complete all advanced modpacks (k2, space exploration, seablock, etc). it probably will not not meet the needs of arbitrarily large megabases. To get a sense of scale, This is an example of 2x2 roboport sized cells at the end of a space-exploration run. this is a MINIMUM size for 1-2 trains and used was feeling cramped for some cells at end-game factorissimos were used to bridge the gap. You can see the side-spur waiting areas. Had we to do it again, we'd probably have gone 3x3 roboport rather than 2x2. My next game I went for even smaller for fun, which looks cool but IMO had very mixed results (deadlocking, throughput, all the issues you'd think)

the longer version, to add on to what others have said about city block size and planning, is that you can use any size you want but some sizes are just going to be better for what you want to do and are going to have higher throughput. It's complex but here are some things that go into the equation:

train length is the absolute most important consideration and you need to decide what size trains you want to run (1-1, 1-2, 1-4, 2-4, etc) which determine the signal block size which will then determine the minimum size city block you can use. best practices are that your intersections must be at least one train full length in size, and there also must be at least one full train length on the "side" between intersections, preferably two.

train stops must be on a spur. depending on the mods/throughput, you may or may not want/need a waiting area and this has to be calculated in the size because turnaround/waiting areas take up a lot of space within the cell. How many train spots will you need? If you don't have a good multi-use station you'll need to account for multiple stops per cell.

if you're designing your own layout, steal a top-rated intersection from https://forums.factorio.com/viewtopic.php?f=194&t=100614 that fits your needs. This is going to determine your max throughput of your base overall, you'll need to do the math on train number/minute times number of wagons to get your max cell input/output. I believe we used the "Super compact Celtic knot".

Additional note: the base assumption of city grids are that your blocks are square but they don't have to be. Rectangular blocks and designs that input/output from specifically the top/bottom might be a better fit.

Bhodi fucked around with this message at 18:58 on Jan 17, 2024

The Locator
Sep 12, 2004

Out here, everything hurts.





Radiation Cow posted:

So I've finally conquered my restartitis and launched a rocket in vanilla. I'm still keen to keep going, which means megabase but I'm not sure how to manage the transition.

Does anyone have any best practices on how to make the transition as smooth as possible? I've noticed that my bottleneck is modules and circuits, but I'm running four blue belts of iron and copper and don't have much room to expand the bus, so I'd need to rejigger the entire base to make space.

Almost tempted to leave this base running and build up somewhere else but I'm not sure if there's sufficient infrastructure to support it.

You've gotten some good advice already, but I'll follow up a little bit with how I've done it.

First thing is to do what others have said and convert your existing base into a 'make building supplies' base, as needed.

I usually then create some massive train-based iron/copper mine and smelter setups in order to create a largish module making setup. If you do this first, then you have modules being made while you build the other stuff that will need the modules and any time you spend expanding rails or whatever you will be building up a buffer of modules to use. This is obviously only if you plan to make your mega-base a beaconed design.

This is an example of a beaconed module factory which will make lots of modules and is fed only by the raw materials - iron, copper, oil and coal. The petroleum, sulphuric acid, plastic and then circuits are all made in this isolated build (the petroleum production isn't captured in the screenshot but it's done just to the right). Trains drop off the iron plates, copper plates and coal. If I hadn't built this right next to an oil field, then of course oil would be trained in also, and if the game went long enough the oil field would have to be replaced by a train station drop. Edit: Should be obvious, but this is an isolated bot-based build.



From this point, I just start doing the same sort of self contained train-fed modules for each science. Sometimes I set up these modules to be fed from trains delivering iron / copper plates, and other times I'll even include the smelting as part of the module build, it just depends on how you personally want to do it.

Here is an example of a green science build that takes in raw ore and spits out green science at about 2700 per minute.



Depending on just how big you want to scale up, you can expand in many different ways, but it will be really hard to squeeze a full rocket/minute (900 spm) base out of an original bus-based starter base unless you left plenty of room to expand. It can absolutely be done if you planned things right in the beginning and only built on one side of the bus so that the bus can be expanded as much as you need. I've done that before by simply ripping out the original science builds and replacing them with fully beaconed builds designed to churn out at least 900 SPM. It will require lots of modules and a blue-belt bus, but it can definitely be done.

I've never tried to go beyond a single rocket launching at max speed with a bus base, but I'm sure it could be done.

The Locator fucked around with this message at 18:06 on Jan 17, 2024

Xerol
Jan 13, 2007


For any kind of self-contained build like that which needs plastic, you can also use coal liquefaction. The coal also gets used to make steam for the liquefaction and if you really wanted to go nuts you could even have the whole thing self-powered.

The Locator
Sep 12, 2004

Out here, everything hurts.





Xerol posted:

For any kind of self-contained build like that which needs plastic, you can also use coal liquefaction. The coal also gets used to make steam for the liquefaction and if you really wanted to go nuts you could even have the whole thing self-powered.

Yup.. I've used coal liquefaction in builds plenty of times. I don't honestly remember why I went with the oil on the one pictured, it was quite a while ago. Sometimes I've built entire bases with 100% of my late-game plastic coming from coal, and other games I've never used liquefaction at all. I guess I just mix it up when I'm playing a vanilla mega-base game, and try to use a different core concept each time.

Edit: I probably used oil because I needed sulphuric acid for blue chips.

celestial teapot
Sep 9, 2003

He asked my religion and I replied "agnostic." He asked how to spell it, and remarked with a sigh: "Well, there are many religions, but I suppose they all worship the same God."

Harvey Baldman posted:

gently caress, I have a few designs I have to go back and poke at with this approach in mind now, this is so much easier.

Full belt variations :)

Only registered members can see post attachments!

Adbot
ADBOT LOVES YOU

meowmeowmeowmeow
Jan 4, 2017
When building a bot-mall do you keep a separate belt fed production zone for high volume logistics stuff like belts, splitters, rails, etc to keep throughput high? Multiple assemblers in the bot mall and tons of logi bots to keep stuff moving? RN I'm setting up the bot mall with all the inputs belts going into a passive provider warehouse with circuit controlled filter inserters to load goods, then doing your standard bot assemblers for everything and im not sure if I should throw my belts and rails into this or keep it separate.

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply