Logistic Network Research

Post all other topics which do not belong to any other category.
meganothing
Fast Inserter
Fast Inserter
Posts: 238
Joined: Thu Sep 15, 2016 3:04 pm
Contact:

Re: Logistic Network Research

Post by meganothing »

Tricorius wrote: But I don’t think “just use mods” is the best answer to everything.
No, but shouldn't they at least be an answer to some problems? As I see it the vanilla game has the right tech tree for beginners but not for experienced players. It can't be the other way around, obviously. IF mods are the answer, it MUST be the answer for experienced/advanced players. Wube could add an (hidden?) option as a shortcut if the demand is really high to warant this, but the vanilla game has to keep the beginners behind the "security tape".
I also disagree that bots are a “cop out” design. They have their own trade offs to use well. Anyone who has spent a reasonable amount of time with them understands that. Perhaps bots should be made more interesting to use, not just more punitively difficult to use.
I didn't say the bots are a "cop out" design, with "cop out" design I was refering to the specific design Hedning1390 was showing. Unlike the train supply design it is a VERY obvious method. As soon as requester chests are available I would guess a very high percentage of beginners immediately would get the idea. Using train wagons to supply assemblers easily is something I for example didn't see at all. Someone had to tell me.
Honestly, I don’t see that much fundamental difference in setting up an early “mall” with logistics bots vs a set of five vertical belts. Other than style or preference. Why is a huge vertical parallel line more interesting or prettier than nicely laid out rows of assemblers and chests? Is striping five rows of belts endlessly upward a feat that Michelangelo would commend, while scoffing at the dude using robots? (Well, maybe he would...I never knew him.)
My first playthrough I did a spaghetti design. In further playthroughs I experimented with vertical sets of belts, belts that transport the basic materials in a big circle and a manhattan design. NONE of them ever were problem-free, all needed adaptions and redesigns to supply intermediate products, redesigns because supply wasn't enough. All kept me occupied tinkering and balancing.

That simply doesn't happen if you build the "cop out" design of Hedning1390. A beginner might never experience the fascination of tinkering or transitioning into building mega factories, never see the interesting problems Factorio offers for the advanced player if he sees the solution to the mid-game immediately before him. Imagine playing factorio as a beginner and shortly after setting up your first research stations you just have to wait for for any new tech and except for combatting the aliens and adding new mining fields there is nothing much to do. If a new tech becomes available, you just put down 1-4 assemblers plus two chests and inserters and you are done. Factorio doens't work if you don't have anything to do, but a beginner isn't able to set his own goals yet.
Again, I simply don’t get the fundamental mindset of “the way you want to do it is bad, so no one should do it that way”. (Also, I’m not saying this is the attitude of the devs, if that were, they’d remove the bots altogether and “damn the torpedos”.)
Probably there are some people around with that attitude, who knows. I'm simply saying the devs have (in my opinion) a reason for not giving the beginner requester chests right away for a very similar reason to why they don't make 10-lane belts, every inserter a filter inserter or belts that directly feed into assemblers. And it is the same reason why any RPG game doesn't give you the Sword of Ultimate Destruction at level 1.

zOldBulldog
Smart Inserter
Smart Inserter
Posts: 1161
Joined: Sat Mar 17, 2018 1:20 pm
Contact:

Re: Logistic Network Research

Post by zOldBulldog »

Wouldn't a simple answer be to allow robots and logistic chests from the earliest resources... if you already launched a rocket in a previous game.

Key word there is *allow*, as it could be configured as an option in the map settings that you have to manually turn on, but can't do so unless you've launched a rocket in a previous game.

meganothing
Fast Inserter
Fast Inserter
Posts: 238
Joined: Thu Sep 15, 2016 3:04 pm
Contact:

Re: Logistic Network Research

Post by meganothing »

zOldBulldog wrote:Wouldn't a simple answer be to allow robots and logistic chests from the earliest resources... if you already launched a rocket in a previous game.

Key word there is *allow*, as it could be configured as an option in the map settings that you have to manually turn on, but can't do so unless you've launched a rocket in a previous game.
Yep, could be a solution. One drawback is that launching the rocket is not a reliable indicator of beginner status. In another thread a lot of long-time players posted that they never launchend a rocket, and I would assume many first-time players will launch a rocket on their first play-through and still should be considered beginners. But since the option would still be turned-off by default, the "damage" might be minimal.

Another hypothetical drawback: An option that changes the research table might make it necessary for many mods that also want to change the research table to add conditional code for it (at a minimum the option just wouldn't work with a mod anymore, it might even produce serious bugs). That naturally depends on how modding the table works in Factorio, I didn't look at it yet, but in another moddable game I know something like this might produce problems or at least more work and complexity for modders.

There is still the uphill battle to convince Wube that they should implement this when another feature they put considerable effort in (modability) already provides a solution (even if such a solution might be refused by some players for security reason or because they think they can't handle the install).

FrodoOf9Fingers
Fast Inserter
Fast Inserter
Posts: 109
Joined: Sat Apr 29, 2017 11:13 pm
Contact:

Re: Logistic Network Research

Post by FrodoOf9Fingers »

I agree on the nerfing principles, that we should make they more interesting rather than more expensive. I would like to see earlier logistics, but maybe limited to not being able to cross roboports for logistics until some research is done.

zOldBulldog
Smart Inserter
Smart Inserter
Posts: 1161
Joined: Sat Mar 17, 2018 1:20 pm
Contact:

Re: Logistic Network Research

Post by zOldBulldog »

FrodoOf9Fingers wrote:I agree on the nerfing principles, that we should make they more interesting rather than more expensive. I would like to see earlier logistics, but maybe limited to not being able to cross roboports for logistics until some research is done.
You already can't get requested and buffer chests until Yellow Science. Those are necessary for logistic factories.

The only logistics currently available before that are simple deliveries to the player and trash collection to storage chests.

What is there to Nerf?

Tricorius
Filter Inserter
Filter Inserter
Posts: 266
Joined: Fri Jul 01, 2016 9:04 pm
Contact:

Re: Logistic Network Research

Post by Tricorius »

zOldBulldog wrote:Wouldn't a simple answer be to allow robots and logistic chests from the earliest resources... if you already launched a rocket in a previous game.

Key word there is *allow*, as it could be configured as an option in the map settings that you have to manually turn on, but can't do so unless you've launched a rocket in a previous game.
In my opinion it is a similar request to those wanting an end-game reset with a varied or simplified new game start. The main difference being an end-game wrap up sequence (congrats, you launched yourself into space but crash landed again) vs a new game setting (skipping the narrative pretenses).

I get why Wube don’t want to maintain two different versions of the tech tree though. I admit it isn’t a simple swap, it may very well have additional ramifications to move Logistics Network research forward in the tech tree.

Zavian
Smart Inserter
Smart Inserter
Posts: 1641
Joined: Thu Mar 02, 2017 2:57 am
Contact:

Re: Logistic Network Research

Post by Zavian »

I sort of expect that if you had the bots, and the logistics chests etc, then you could build bot based infrastructure and it would work, even if you don't have the tech yet. (Admittedly the bots would be slow with no upgrades, but I expect you could still build a logistics network that would work. That seems to be my experience from resetting tech when using the creative mod).

So a mod that wanted to respawn you on a new world, but set the tech tree back to zero, you should be able to bring bots + logistics chests + solar panels + accumulators etc, provided the mod lets you keep your inventory, and get a head start on the new map.

mrvn
Smart Inserter
Smart Inserter
Posts: 5681
Joined: Mon Sep 05, 2016 9:10 am
Contact:

Re: Logistic Network Research

Post by mrvn »

Zavian wrote:I sort of expect that if you had the bots, and the logistics chests etc, then you could build bot based infrastructure and it would work, even if you don't have the tech yet. (Admittedly the bots would be slow with no upgrades, but I expect you could still build a logistics network that would work. That seems to be my experience from resetting tech when using the creative mod).

So a mod that wanted to respawn you on a new world, but set the tech tree back to zero, you should be able to bring bots + logistics chests + solar panels + accumulators etc, provided the mod lets you keep your inventory, and get a head start on the new map.
When I crash landed I managed to salvage the nuclear reactor from the ship, 8 roboports, 20 construction bots,10 logistic bots and 100 nuclear fuel cells. Time to figure out how to get more fuel before I run out. I think figuring out how to build more bots will be benefitial too. ..

Proculator
Burner Inserter
Burner Inserter
Posts: 10
Joined: Wed Dec 19, 2018 8:28 pm
Contact:

Re: Logistic Network Research

Post by Proculator »

Viewed a few of the exchanges in this topic in order to see what your perspective is.

Any suggested sub-forum(s) that you (an experienced player) might suggest a beginner should hang out?

My first topic posting should show what my perspective is. Thanks for your future help.

sathill
Long Handed Inserter
Long Handed Inserter
Posts: 52
Joined: Mon May 29, 2017 10:16 am
Contact:

Re: Logistic Network Research

Post by sathill »

Also agree logi network its too far in tree.

mrt144
Inserter
Inserter
Posts: 43
Joined: Thu Sep 13, 2018 9:30 pm
Contact:

Re: Logistic Network Research

Post by mrt144 »

libik wrote:
Sun Jun 03, 2018 10:09 pm
Hi,
I have started once again after some time. What I have noticed is that researching of logistic network became quite difficult. You can get roboports etc. quite easily, so I was hoping that for recipes that requires lot of special ingredients, I can start using the robots... However I was surprised that to get requester chest, you have to invent logicist network which requires 5 different science packs with one of them being quite late-game material.

I know the game is focused on belts and trains and these robots should be for late-game usage, still its "too late" in my opinion. I would be fine with expensivity - like it costs i.e. 500-750x red, green and blue, so you really have to invest heavily to get that. But it would allow you to get these options before having already being forced to create almost all science packs types with belts.

Whats your opinion on that?
I disagree with the opinion only because I think it's actually trivial to beeline logistics techs.

I recently restarted a megabase attempt after going a bit crazy and deep into trying to prebuild an entire front to back 1000 SPM solution with belts...it was effing terrible and I over explored the map so save and load times were in minutes.

The thing I prioritized in my current game was getting all logistics and welp, 48 hours later with a nearly airtight yet ponderously slow research setup I easily scooped them up.

Like, is it a cop out to leave your game running unattended for 2 days to acquire resources and beakers because your humble base is that good (and you turned biters off because they interfere with megabase builds haha).
Last edited by mrt144 on Thu Jan 24, 2019 12:51 am, edited 2 times in total.

mrt144
Inserter
Inserter
Posts: 43
Joined: Thu Sep 13, 2018 9:30 pm
Contact:

Re: Logistic Network Research

Post by mrt144 »

Hedning1390 wrote:
Mon Jun 04, 2018 12:18 pm
Robots solve any and all logistics problems. They "solve" the game, so to get them you have to first beat the game.

This is what bots turn your game into. https://i.imgur.com/jv0hQ6f.png There's no challenge building with bots at this level. Completely game breaking.
I don't believe this is correct because while bots solve logistics efficiency related to path finding of materials and better optimization of material distribution they still have their own set of issues with power logistics (which are made easier with bots) and traversal time and actual logistic bot manufacture. If you have a network of 3000 logistic bots they didn't simply appear out of thin air and there was a time and material cost to that.

More importantly though is that at a certain size, logistic bots simply can't deliver the throughput that other means could for a much lower material and energy cost. An army of 3000 logistic bots is great for point to point transmission of materials in a large manufacturing base but you still aren't going to use them to get 50k ore or 50k iron plates to your base from a far flung ore patch.

I think if you go whole hog on bots, the impetus for that is that you're playing the game at a more macro level than micro level. Raw materials converted to goods is where more of your thought is devoted than the specifics of converting a basic good to a more intermediate or advanced good.

Has there been conversation about where logistic bots start to really become detrimental to overall efficiency based on route distance for available and requested materials? I would love to read about that before I go logistic bot heavy for a 1000 SPM base.

Edit 2:

Now that I've hopped back into my current game at Phase 1 of a Mega Base, the time to solve some of the game's problems are trivial for me at this point only because I've played the base game experience exactly once to completion and then have gone through several iterations of building a slightly spartan research base to open up the options for a megabase. This isn't trivial to neophytes who don't have 80 hands on hours doing the same task faster and with more direction in each iteration.

Thinking on that first run through the base game, the logistics robots and construction robots came in at just the right moment when some stuff became tedious to do by hand to the point I was getting super antsy and pissed that I would have to walk out everything everywhere all the time. On this 5th iteration of a sandbox 1000 SPM base, I saw how quickly I could get to the logistics techs and it was relatively quickly.

But then what? Phase 2 is going to be building a 'Factory of Base Building Materials' that also functionas as scale model of how the larger 1000 SPM factory will operate and I'll learn lessons in that.

I guess this is just evidence that this game is what you make of it ;)

mrvn
Smart Inserter
Smart Inserter
Posts: 5681
Joined: Mon Sep 05, 2016 9:10 am
Contact:

Re: Logistic Network Research

Post by mrvn »

mrt144 wrote:
Thu Jan 24, 2019 12:31 am
Hedning1390 wrote:
Mon Jun 04, 2018 12:18 pm
Robots solve any and all logistics problems. They "solve" the game, so to get them you have to first beat the game.

This is what bots turn your game into. https://i.imgur.com/jv0hQ6f.png There's no challenge building with bots at this level. Completely game breaking.
I don't believe this is correct because while bots solve logistics efficiency related to path finding of materials and better optimization of material distribution they still have their own set of issues with power logistics (which are made easier with bots) and traversal time and actual logistic bot manufacture. If you have a network of 3000 logistic bots they didn't simply appear out of thin air and there was a time and material cost to that.

More importantly though is that at a certain size, logistic bots simply can't deliver the throughput that other means could for a much lower material and energy cost. An army of 3000 logistic bots is great for point to point transmission of materials in a large manufacturing base but you still aren't going to use them to get 50k ore or 50k iron plates to your base from a far flung ore patch.

I think if you go whole hog on bots, the impetus for that is that you're playing the game at a more macro level than micro level. Raw materials converted to goods is where more of your thought is devoted than the specifics of converting a basic good to a more intermediate or advanced good.

Has there been conversation about where logistic bots start to really become detrimental to overall efficiency based on route distance for available and requested materials? I would love to read about that before I go logistic bot heavy for a 1000 SPM base.

Edit 2:

Now that I've hopped back into my current game at Phase 1 of a Mega Base, the time to solve some of the game's problems are trivial for me at this point only because I've played the base game experience exactly once to completion and then have gone through several iterations of building a slightly spartan research base to open up the options for a megabase. This isn't trivial to neophytes who don't have 80 hands on hours doing the same task faster and with more direction in each iteration.

Thinking on that first run through the base game, the logistics robots and construction robots came in at just the right moment when some stuff became tedious to do by hand to the point I was getting super antsy and pissed that I would have to walk out everything everywhere all the time. On this 5th iteration of a sandbox 1000 SPM base, I saw how quickly I could get to the logistics techs and it was relatively quickly.

But then what? Phase 2 is going to be building a 'Factory of Base Building Materials' that also functionas as scale model of how the larger 1000 SPM factory will operate and I'll learn lessons in that.

I guess this is just evidence that this game is what you make of it ;)
For a mega base the biggest factor is UPS and nothing else. So the question isn't if bots are cheaper to build and transport goods more efficiently but the question is do they use up more CPU time doing so.

That said lets ignore the CPU cost for now and look at the efficiency. You can pretty much get a limitless number of bots to transport goods in parallel but the limiting factor is recharging bots. Each bot can only travel a set distance before it needs to recharge (N tiles) and each roboport can only recharge a set number of bots per minute (M bots) and each bot carries K items. Lets assume the bots travel in a horizontal line from A to B over a great distance. You then put a line of roboports one next to the other on that line. That allows (N / 4 * M / 2 * K) items / minute to be moved from A to B.

In the same space you can put 4 parallel belts, 2 car/tank belts or a 2 track train line. For large distances trains will win by a landslide.

And now look at the CPU cost. A train will certainly use less CPU than thousands of bots. That only leaves the loading/unloading stations and factory itself to worry about. Make them bot or belt based? Apparently bots are better there, especially since they use less space making beacon setups compacter and therefore more efficient. Not sure if belt setups will actually use more space because bot setups need lots of roboports. It's just that you can place the roboports outside of the beacons.

mrt144
Inserter
Inserter
Posts: 43
Joined: Thu Sep 13, 2018 9:30 pm
Contact:

Re: Logistic Network Research

Post by mrt144 »

mrvn wrote:
Thu Jan 24, 2019 1:43 pm
mrt144 wrote:
Thu Jan 24, 2019 12:31 am
Hedning1390 wrote:
Mon Jun 04, 2018 12:18 pm
Robots solve any and all logistics problems. They "solve" the game, so to get them you have to first beat the game.

This is what bots turn your game into. https://i.imgur.com/jv0hQ6f.png There's no challenge building with bots at this level. Completely game breaking.
I don't believe this is correct because while bots solve logistics efficiency related to path finding of materials and better optimization of material distribution they still have their own set of issues with power logistics (which are made easier with bots) and traversal time and actual logistic bot manufacture. If you have a network of 3000 logistic bots they didn't simply appear out of thin air and there was a time and material cost to that.

More importantly though is that at a certain size, logistic bots simply can't deliver the throughput that other means could for a much lower material and energy cost. An army of 3000 logistic bots is great for point to point transmission of materials in a large manufacturing base but you still aren't going to use them to get 50k ore or 50k iron plates to your base from a far flung ore patch.

I think if you go whole hog on bots, the impetus for that is that you're playing the game at a more macro level than micro level. Raw materials converted to goods is where more of your thought is devoted than the specifics of converting a basic good to a more intermediate or advanced good.

Has there been conversation about where logistic bots start to really become detrimental to overall efficiency based on route distance for available and requested materials? I would love to read about that before I go logistic bot heavy for a 1000 SPM base.

Edit 2:

Now that I've hopped back into my current game at Phase 1 of a Mega Base, the time to solve some of the game's problems are trivial for me at this point only because I've played the base game experience exactly once to completion and then have gone through several iterations of building a slightly spartan research base to open up the options for a megabase. This isn't trivial to neophytes who don't have 80 hands on hours doing the same task faster and with more direction in each iteration.

Thinking on that first run through the base game, the logistics robots and construction robots came in at just the right moment when some stuff became tedious to do by hand to the point I was getting super antsy and pissed that I would have to walk out everything everywhere all the time. On this 5th iteration of a sandbox 1000 SPM base, I saw how quickly I could get to the logistics techs and it was relatively quickly.

But then what? Phase 2 is going to be building a 'Factory of Base Building Materials' that also functionas as scale model of how the larger 1000 SPM factory will operate and I'll learn lessons in that.

I guess this is just evidence that this game is what you make of it ;)
For a mega base the biggest factor is UPS and nothing else. So the question isn't if bots are cheaper to build and transport goods more efficiently but the question is do they use up more CPU time doing so.

That said lets ignore the CPU cost for now and look at the efficiency. You can pretty much get a limitless number of bots to transport goods in parallel but the limiting factor is recharging bots. Each bot can only travel a set distance before it needs to recharge (N tiles) and each roboport can only recharge a set number of bots per minute (M bots) and each bot carries K items. Lets assume the bots travel in a horizontal line from A to B over a great distance. You then put a line of roboports one next to the other on that line. That allows (N / 4 * M / 2 * K) items / minute to be moved from A to B.

In the same space you can put 4 parallel belts, 2 car/tank belts or a 2 track train line. For large distances trains will win by a landslide.

And now look at the CPU cost. A train will certainly use less CPU than thousands of bots. That only leaves the loading/unloading stations and factory itself to worry about. Make them bot or belt based? Apparently bots are better there, especially since they use less space making beacon setups compacter and therefore more efficient. Not sure if belt setups will actually use more space because bot setups need lots of roboports. It's just that you can place the roboports outside of the beacons.
Thanks for the reply!

I think your post further highlights that the dimension of problem solving changes depending on your goal as a factory builder. If it's to do a run through the base game and stumble through everything, logistics bots might be necessary crutch to completing it at some point because the alternative might be a more streamlined attempt from basically nothing - the builder may have painted themselves into a painful corner they didn't know was a corner due to their inexperience with the game.

For those that have climbed to that vista, then its much more self determined on what to do after that. I basically saw the loco megabases and said "I want that!" Other folks are enamored with simply building elegant circuits to control things, others want to do speed runs, etc etc.

The fundamental thing to identify as a builder is what are your goals and then what are the limits you butt up against in achieving those goals. As you state for megabases its CPU/UPS, for a minimalist it might be conservation of materials, there's so much self determined opportunity.

As I said above, thinking back on that first run through the game, both types of bots appeared seemingly at just the right moment to fight tedious actions (I need to build 20 smelter array and have a good design already. Do I really need hand install each? No. I found that for my first nuclear plant, logistic bots felt like the right call for the fuel containers that get used in the plant and then can be recycled. And that's solely what I used them for there). For the factory builder with experience and with a CPU/UPS limit on the horizon, they come way too early because you have the knowledge of how to use them just that much better and earlier than a novice.

PTTG
Long Handed Inserter
Long Handed Inserter
Posts: 51
Joined: Sat Nov 18, 2017 7:47 pm
Contact:

Re: Logistic Network Research

Post by PTTG »

If logistics bots had to path on the ground, they'd be perfect. buildings would need to be set up with accessibility in mind, traffic would be a problem, but a solvable one, roads and paths would be cool-looking elements of factory design.

Koub
Global Moderator
Global Moderator
Posts: 7175
Joined: Fri May 30, 2014 8:54 am
Contact:

Re: Logistic Network Research

Post by Koub »

PTTG wrote:
Sun Jan 27, 2019 4:33 am
If logistics bots had to path on the ground, they'd be perfect. buildings would need to be set up with accessibility in mind, traffic would be a problem, but a solvable one, roads and paths would be cool-looking elements of factory design.
And megabases would switch from UPS (Updates Pers Second) to SPU (Seconds Per Updates). You'd need a supercomputer to run a big base.
Koub - Please consider English is not my native language.

mrvn
Smart Inserter
Smart Inserter
Posts: 5681
Joined: Mon Sep 05, 2016 9:10 am
Contact:

Re: Logistic Network Research

Post by mrvn »

PTTG wrote:
Sun Jan 27, 2019 4:33 am
If logistics bots had to path on the ground, they'd be perfect. buildings would need to be set up with accessibility in mind, traffic would be a problem, but a solvable one, roads and paths would be cool-looking elements of factory design.
Install the AAI mods.

brutalbombs
Burner Inserter
Burner Inserter
Posts: 7
Joined: Thu Jan 24, 2019 5:41 pm
Contact:

Re: Logistic Network Research

Post by brutalbombs »

libik wrote:
Sun Jun 03, 2018 10:09 pm
Hi,
I have started once again after some time. What I have noticed is that researching of logistic network became quite difficult. You can get roboports etc. quite easily, so I was hoping that for recipes that requires lot of special ingredients, I can start using the robots... However I was surprised that to get requester chest, you have to invent logicist network which requires 5 different science packs with one of them being quite late-game material.

I know the game is focused on belts and trains and these robots should be for late-game usage, still its "too late" in my opinion. I would be fine with expensivity - like it costs i.e. 500-750x red, green and blue, so you really have to invest heavily to get that. But it would allow you to get these options before having already being forced to create almost all science packs types with belts.

Whats your opinion on that?
Keep in mind this is when you play the STANDARD game mode. The whole point of the STANDARD mode is the progression designed by the developers; it is a natural progression that is in place in this game mode to FORCE players to make working - and smart - systems made of belts and rails before they get to the technology that renders them useless. And this is the reason that you would play the STANDARD game mode. If you want to focus on building smart and effective systems with bots, you should play with the SANDBOX or any of the free-play derivatives (or maybe a MOD that gives you full bot network earlier?).

Point is, changing this designed progression will potentially remove the need to make the most advanced/complex belt/rail system nearing endgame, which is a really, really, important exercise for new and intermediate players to this game.
Regards,
brutalbombs

Post Reply

Return to “General discussion”