My problem with infinity research productivity in 2.0

[Carbon]

In 2.0 there shall be an infinity research for the productivity.

I think it would destroy factorio for me.

The goal to have a coordinated production in which everything is coordinated is hardly possible with it.

After each research the production would have to be adjusted. This forces one to a box construction that is easily expandable.
With the new quality system and this research, one is gradually forced to a transport system with robots, because assembly lines are too slow.

A coordination of the production processes is therefore no longer possible. in any case, this is also an endless task.

Yes, i don't have to implement this, but it will then always be in mind it can be in the normal game.

what do you think?

[Galileus]

I think it's really horrible the gaem put a gun to your head and forced you to play the way you dislike.

If only there was a way to play a gaem in a way you want to play it!

How is one forced to use robots? How can you conclude that from how little info we have?

Also - 'coordinate' your factory on every 1x, 5x, 10x, 15x... research milestone. Design in a way to allow for the next research before milestone to be fully used, as in, design with progress in mind. Problem suddenly solved. It's amazing!

[Khagan]

I think it's far too early to jump to any conclusions.

You are _probably_ right that one effect of productivity research and quality will be to end the idea of designing a once-and-for-all optimum production line for any given output. I'm inclined to regard that as a Good Thing.

In some cases this may _perhaps_ shift the balance of advantage from belts to bots. Since I currently find no real use for logibots outside of the 'mall', that might also be a Good Thing.

[Hurkyl]

It's not actually endless. There is a hard limit on the usefulness of productivity researches: how fast you can move items in and out of an assembler with inserters.

So if there isn't an infinite research for inserter stack capacity, then in principle there is an endpoint where further productivity research cannot increase the throughput of an assembler.

In practice, I imagine the cost will grow exponentially, so there are practical limitations of how far along you can go.


I get the impression there will be less pressure to use 'fully beaconed' designs, because with quality machines and quality speed modules, you will already be able to get the machines to blinding speed.

So, this means you have more space for designing your production lines, e.g. to arrange for direct insertion for greater throughput.

[Fel]

It was already very tough to feed end-game setups with belts anyway.
Filling machines with productivity modules and having speed modules in beacons made them go through ingredients really quickly and produce more as well.

Belts quickly become unreliable for those setups as inserters interracting with belts take time to pickup/deposit items from/on belts.
This time is not needed when going between inventories only (chests, machines, wagons).

As far as I understand FFF 376, it will be in the expansion rather than 2.0 and many things about the expansion seem to scream that you won't be able to rely on always using the same blueprints.
Instead, it seems to reward a different approach, which is closer to the effects of recipe tiers in bob and angels for example.

So less copy/paste and more tinkering, seems like a good way to really set apart vanilla and the expânsion.

[Odonata]

There are good reasons for it. As mentioned in the blog post, only 2 infinite technologies
currently exist that are not military-oriented. This is... really not a lot, especially when a lot of megabase-builders play with biters disabled and those 2 are now the only ones worth looking at. It certainly makes infinite techs more interesting and less of a monotonous resource sink, with how the player is given a reason to work around them and concoct new blueprints that take extra production into account.

I've also realized that there is another purpose to adding these techs that wasn't even mentioned in the FFF; performance issues and the effective science cap. With all that's being introduced in SA it's near impossible that performance won't take a hit. Even in vanilla though - after you build a large enough factory that UPS starts to drop, the game might as well be over because expanding further will chug the game too much. Productivity research is a workaround that would allow a post-endgame player to keep ramping up production (with massive cost of investment) for a while longer, especially with the lab productivity.

[brian_va]

Originally posted by Carbon:

With the new quality system and this research, one is gradually forced to a transport system with robots, because assembly lines are too slow.

what do you think?

all your science doesn't have to be in the same place.

5 locations churning out 1k/m science can be more logistic efficient than a single 5k/m science factory. obviously it not necessarily that easy, but its not really all that difficult either.

[knighttemplar1960]

It shouldn't be an issue for you. You'll still have the upgrade planner. Once you tweak a set up you can simply add it to your upgrade planner and save it with the rest of your blue prints.

Once you saturate your outputs further productivity researches will still reduce the amounts of inputs you require. All of that will get your subfactoies to a point that all belts are saturated at all times which will save UPS.

The changes to robot pathing will also make robots more UPS friendly so using them in more places won't be as big a drain.

If the devs have changed the game engine for 2.0 to make more use of additional cores bots may become the go to for all megabases going forward.

[RiO]

Originally posted by Carbon:

In 2.0 there shall be an infinity research for the productivity.

[..]

what do you think?

I think you didn't read the blog post properly.
There won't be a generic infinite productivity research.

There'll be infinite productivity research for a few choice end-game recipes which have long production times; take many expensive ingredients; and require so many buildings that it's impractical to scale them beyond one or maybe two full output belts per sub-factory without going into the unwieldy stupidly sized.

Basically; only for those recipes where it makes sense and where the vertical scaling will help alleviate the need for horizontal scaling to more acceptable levels.

At some point or other you'd also stop researching them, because the gains are simply no longer worth it. Assuming they run on diminishing returns, and eventually you'd get like +0.01% productivity out of them.

The nightmare scenario you're sketching where belts will no longer be able to move items fast enough, doesn't apply.

The nightmare scenario you're sketching where you'll be continuously rebuilding because the ratios are off, doesn't apply either. The worst that will happen? Some of your already built machines will stop running 100% of the time because output is saturated. And you'll use less of the input.

There will never be a situation where you'll need to rebuild because you're no longer meeting output expectations, or where you're exceeding input expectations.

You'll just be revisiting a few facilities to strip off some assemblers; after every 5th level of research or so. And then every 20th level or so, you'll figure: "hmm.. my provided input has enough slack left that I can fit another production lane."

[TSP]

Originally posted by Carbon:

what do you think?

I think direct insertion is the best way to play the game in terms of UPS. This means that, with this in mind, it doesn't matter at all that productivity gets improved for setups, I'm not going to adjust those setups because they're already DI'd and maximized on UPS. In which case the input demands for those setups get reduced and output supply gets increased.

So what do I think? It's an absolute win-win.

[AlexMBrennan]

At some point or other you'd also stop researching them, because the gains are simply no longer worth it

No, I don't believe that that is true because the point of Factorio is to maximise production; most players chasing the maximum SPM would be just as happy to shove the science packs into a void chest without the tiny benefits they get from infinite research.

[DaBa]

What do I think? I think you are making a lot of assumptions that aren't backed up by anything concrete. Also I simply don't see the correlation you're trying to describe here. Weird conclusions.

[Ferox_Stormdragon]

Factorio already has infinite productivity research for miners, why not expand it to other machines as well?

[Seth_Haveron]

Originally posted by tcmitchell957:

Factorio already has infinite productivity research for miners

you sure you dont confuse things with a mod ??

[brian_va]

Originally posted by Seth_Haveron:

Originally posted by tcmitchell957:

Factorio already has infinite productivity research for miners

you sure you dont confuse things with a mod ??

nope

[Ferox_Stormdragon]

Originally posted by Seth_Haveron:

Originally posted by tcmitchell957:

Factorio already has infinite productivity research for miners

you sure you dont confuse things with a mod ??

no

https://wiki.factorio.com/Mining_productivity_(research)