Remove requester chests

[Plop and run]

Lads and gents, may I present to you

The layout

*Ignore the suboptimal left side, focus on the center part*
A simple blueprint is enough to add a new item in production:

blueprint

All you need is configure the requester chest, the assembly machine and the smart inserter. Ok, in rare cases you need to do some piping.

Blah Blah Blah...

Before robots, you need to solve the problems that belts throw at you. With robots, there are no more decisions to make. When you played the game for the first time, could you imagine that the game about factory optimization turn the whole factory design process into "put down the blueprint, configure the components". The next step would be a tool that allows you to specify an item and it's quantity. Once specified, it would follow the said process of creating the item.
Maybe this is the desired direction the game should develop, and self-expanding factory sounds nice, but there has to be something a simple algorithm can't handle.

The obvious solution would be to remove the requester chests from the game. Bots will move items from providers to storage, from storage to player, build and repair things, but will not be a part of assembling.
Obviously, belt based factories are pain beyond description, but there are lots of things that can replace requesters: chest-on-a-belt, trains, automated cars and what not.

Your thoughts?

[muzzy]

I believe the problem is that energy costs involved in this are too cheap.

Maybe logistics bots should require more energy? Maybe roboport (or bots) should cause more pollution, or maybe flying bots should attract enemies in some other ways? Perhaps massive energy production should be more unreliable or otherwise dangerous?

Transport belts are free energy devices and theoretically they should always be better than anything else because of this.

[Coolthulhu]

I'd say robots should be limited to very low numbers of items. Removing bots' most important role right now would be very painful. We'd need something to replace them.
If we had smart splitters, smart belts or other such deterministic (so not smart inserters) methods of fully diverting item flow, bots wouldn't be needed, just handy.

If I were to solve this, I'd change bots like this: logistic bots get huge speed boost when not loaded, but huge speed penalty when carrying stuff. No energy cost for moving self, but prohibitively big cost for moving items. It would also mean they'd be limited to short-distance jobs (no taking jobs that would kill them). Roboport would only carry like, 32 bots, but would have huge range to offset that.

[Plop and run]

It is not the bots who made the said layout possible: it is requester chests. So, instead of nerfing bots by breaking their wings or numbers per port or energy usage and what not (witch will result in MOAR ROBOPORTS and MOAR BOTS and MOAR POWERPLANTS), I suggest removing the abusive use of them. I see no problems with bots transporting things from provider to requester (so that finished goods storage can be centralized) and supplying player with items (so I don't have to worry about running out of capsules/inserters/ammo as long as my factory can support it).

[just_dont]

I totally agree that currently logistic bots are WAY too good compared to other factory-building options. However, I too (as posters before) think it's the bots that should be addressed, rather than chests.
Requester chests are currently the only way to properly organize your storage for finished goods. Without them, producing a large number of items in a single factory would lead to a HUGE mess in the resulting storage. So I don't like the idea of just removing them.

But the easiness of setting up a bots-powered factory could be completely offset by some bots' nerfs. At the moment, they are way too good to even compete with other solutions. If they were limited, energy-hunrgy, and less efficient in quickly moving lots of items at once -- it would be quite balanced.
At the moment, bots provide not only design simplicity (which is already a great boon), but also more energy-efficient operations AND more speed (as inserters picking items 1-by-1 from belt can't compete with bots picking up stacks of several items).

[Deathmage]

I think the launching speed should be nerfed, they get released all too quickly. If they where released slower, should you have an execs of orders coming in they would take a wile to all go and sort it. I think the other elements with them are fine. Another suggestion would be that they cannot be too close to each over and need their own space instead of flying on top of each over. Just my thoughts on this.

[Lee_newsum]

jast one word NO

[Coolthulhu]

I suggest removing the abusive use of them.

It's not "abusive" when it's the most important role that actually made them exist.
Bots were created with provider->requester transfer in mind. Everything else was built around that.

supplying player with items

Having an exception just for the "manual inserter"?
I'll just quote a good answer:

jast one word NO

So, instead of nerfing bots by breaking their wings or numbers per port or energy usage and what not (witch will result in MOAR ROBOPORTS and MOAR BOTS and MOAR POWERPLANTS)

Only if you do it badly. I suggested a solution that would work. The key phrase is "prohibitive energy cost". Bots transporting blues is fine - blues are scarce enough that you could be moving them by hand.
And when players decide to spend more energy on transporting stuff than on everything else, cripple their economy and ignore useful features, it's their choice. It's only really bad when it's the optimal way. It's slightly bad now because it's too easy, it would be really bad if it made sense to transport all goods to player character just to put them in manually, because the only other alternatives are belts and trains.

[muzzy]

I think bots flying in the air should attract enemies over long distances. Under the current mechanics that means flying bots should generate pollution, enough pollution that excessive use of bots would call in enemies over ridiculous distances... Maybe bots should indeed spend more energy for the number of items they're carrying, and thus more pollution per number of items they transport.

[kovarex]

When I was playing for the last time, I used bots much less than before. Bots are quite expensive to build, and you need lot of roboports to support bigger flow of robots. Also without the very expensive upgrades of speed and cargo, they are even less efficient.

What I learned in the last game was, that using robot transportation too much made bottlenecks very quickly, and whenever there was bigger flow somewhere, especially over longer distance, it was better to use transport belt.

Maybe my strategy assumption were just wrong, but I believe that it is just matter of balancing, I would definitely not remove them.

The balancing might consist of:

• Robot battery storage size
• Robot energy consumption
• Robot recharge speed
• Robot acceleration (it is instant now)

I believe that choosing between robot and belts when both is available should be strategic part of the game, where for prototype production robots might be used, for larger scale, the player can switch to belts or trains.

[ssilk]

I have one more:

Distance to which the bots choose another roboport vs. remaining charge.

[Coolthulhu]

Instant bot acceleration is fine. Bots move chaotically enough already with their "pause job to recharge" behavior.
It would look kinda stupid if they had to do any of:

• Decelerate to 0 before changing direction
• Keep speed while changing direction despite requiring to accelerate at start
• Swing around like hi-tech boomerangs (like combat bots do now)

Unless you intend to add an actual pathfinding system (like I suggested in https://forums.factorio.com/forum/vie ... f=6&t=3283) or significantly change recharging behavior, keep bot physics simplified.

If energy consumption/storage/recharge are to be nerfed, remove the bot suicide (at 0 energy) first. Preferably also while fixing the job pausing thing, so that they don't spend more time breaking off to recharge than working. Late game bot hordes turn kinda probabilistic (can't exactly control their number or density) in the worst way possible.

The worst issue I see with bots being "OP" at the moment is the fact that bots can carry more than 1 item at a time, while belts can't. This makes high speed assemblies require either bots+chests or multiple inserters to maintain 100% usage and predictable supply. The latter requires significant redesigning (hard and tedious when the factory is compact) or building with it in mind from the start. Chests-on-belts, multigrab inserters, item packets/stacks or cheap trains available from the start could all help with bots being the ultimate local transportation system.

[ssilk]

The pros and cons of current transport medias


1. Belts:
Pro: medium long distances (500 tiles), constant flow, two distincted types of items (per lane), useable as buffer, cheap parallelization, medium space requirement, needs no energy, very cheap for basic belts.
Con: slow, medium transport capacity, curves nerve flow, sometimes complicated to handle, may block in rare situations, buffer function is sometimes not wished, not intelligent, one direction only, medium slow building.

2. Trains:
Pro: very fast, very high capacity, high throughput, long distances (300-5000), semi intelligent (automated mode), distinct transport per wagon, in average constant flow, low need of energy, cheap in long run.
Con: uses big space like nothing, complicated setup (train station, filling, emptying, routing, signals), not useful for short transports, fuel for the locos, short peaks when loaded/unloaded, high first investment, slow building..

3. Logistic bots:
Pro: intelligent transport, extreme high capacity (with stack size bonus) within short distances (0-100), good charge behavior when used periodically, intelligent transport, fast building.
Con: medium-high need of energy, limited usage area (roboports), limited number of charging points, bad behavior when overused, danger of loosing bots when overused, very bad behavior when transport high capacity over long distances (>300 tiles), expensive.

I also want to mention, that we have the car, which is extremely recommended for fast transport over medium distances with an astonishing high capacity and inserters, namely the long arm inserters are an interesting replacement for low capacity, but long and fast transport in one direction.


So what's missing in my eyes is an intelligent transport over medium (and eventually up to high) distances.

In consequence it should be a development target to have game mechanisms, which will make the player very clear, that the logistic bots are not very useable over high distances.

[just_dont]

Thing is, short-distance transportation (i.e. inside one's factory) has MUCH more weight and priority than other forms of transportation. Short-distance transport is the only thing you need as a part of your factory.
Long- and medium-distance transportation currently (mostly) boils down to moving 4 basic resources (everything excluding stone) in large quantities back towards your factory. Oh, and also moving items (bots, ammo) to player, but that part is not currently "required" in any way.

[ssilk]

Thing is, short-distance transportation (i.e. inside one's factory) has MUCH more weight and priority than other forms of transportation. Short-distance transport is the only thing you need as a part of your factory.

The only need you have.

Long- and medium-distance transportation currently (mostly) boils down to moving 4 basic resources (everything excluding stone) in large quantities back towards your factory. Oh, and also moving items (bots, ammo) to player, but that part is not currently "required" in any way.

What I forgot in my list was the time to built up the transport. Which is slow for train and belts and fast for logistics. I edit that. That brings some players to the thought, to use logistics only, until they come to a point, where it doesn't work anymore.

A fourth main transport system should be built up very fast, because what your description misses, is how to built up the distant mines and all the stuff you need to defend them. This is currently a very difficult part and over distances of 500 tiles it's not so much fun anymore.