Factorio Forums

This is kind of a hot topic right now, so here are my 2 cents.

I think belt performance (as in throughput) is currently on a good level, however the process of building belt contraptions makes you run into different kinds of QoL issues. My passion for belts lasts only so much of them before giving up and switching to bots.
Here is some of the things that would greatly inspire (at least me) to use belts more often in the endgame:

1. 3-way splitters
I think there was a mod that allowed you to merge several splitters into a big one, essentially allowing for balancers with unlimited amount of inputs/outputs. I'm not sure if that one is a good idea to add in vanilla, because it is unclear how fast the items should travel horizontally across the splitter and the difficulties of creating arbitrarily large buildings on the fly. But in a world with 6-block trains and fastest belt having 3x speed of the slow ones 3-way splitters are definetely necessary, as a separate entity and research. Also, hello new balancer contraptions!

2. (Un)loaders
Loaders already exist in the map editor, and Klonan made a cool mod for those suffering without belt compression in 0.16, which is essentially a 2x1 chest with a built-in loader and unloader.
Right now it takes 6 stack inserters to unload a container onto one blue belt and fully compress it. This may be called a feature by some, but the designs and layouts involved are often quite clumsy. Also, inserters only unload on one lane of an away-facing belt (and it's better to leave this as it is). Loaders will be a good addition as a new high-tech research and will work quite well with the new splitters too.

3. New chests: 2x1 and 2x2, optionally 3x1 and 3x3
Even sizes will allow for more convenient item transition between trains (I found myself using Klonan's buffers for that as a 2x1 chest). Bigger sizes can gain more inventory cells per tile cell and therefore be used in a storage area - an application where logistic bots completely dominate atm.

4. Belt compression
This needs to make its way back into the game. Splitters are too expensive and too big, while old method of compression was cheap and also left you with some extra spacing for power poles.

Ubertwink wrote:This is kind of a hot topic right now, so here are my 2 cents.

I think belt performance (as in throughput) is currently on a good level, however the process of building belt contraptions makes you run into different kinds of QoL issues. My passion for belts lasts only so much of them before giving up and switching to bots.

I have to disagree on this first point. Once you try to start building 8x8 setups with belts you quickly run into throughput issues (in particular with with iron, copper and green circuits). Whilst it is possible to work around them, doing so often involves compromises that are undesirable (eg 4.25 craft speed, instead of 5.5, feeding setups from both ends, because you can't get enough throughput from blue belts, and can't route another belt under the inserters, because you need multiple inserters, since inserters picking up from a belt are slower than inserters picking up from chests). Often it's just not worth the hassle trying to build a high throughput setup with belts that is as efficient as the equivalent setup would be with bots. Lose of underground compression also hurts some belt builds, since we don't have another way to compress belts this is as compact, except by using the circuit network to make timed swings.

All these things could be addressed with stacked belts. If a stacked belt could carry stacks of 4 items, throughput would be 4 times higher, inserters would pickup faster and deposit faster. We only need some items we eligible to be stacked on a belt that would make high throughput beaconed setups much more practical. Support for just iron and copper plates and green circuits would help a lot. Lower priority would be plastic, copper wires, red circuits, gears, steel, pipes, bricks, iron ore, copper ore, stone and coal. I think that covers enough ingredients that in practice you would no longer find it difficult to route items for 8x8 beaconed assembler rows of reasonable length.

I've brought up some points before, which I'm sticking to.

viewtopic.php?f=38&t=56519&start=400#p334355

bobingabout wrote:Love the splitter
Stack inserters should be able to add stacks to belts
a higher tier belt and inserters is a good idea
give us base game loaders.

if you read, the point about the loaders makes sense. They can load/unload to/from a belt to/from a factory/train/chest, but one side MUST be a belt, meaning there's no chest to factory/train links, this means that robots still need to use inserters to load/unload train/machines, where belts can use the loader, giving you a reason to use belts.