Hacker News Reader

Comment by lupusreal | Hacker News Reader

That's pretty cool, but balancers are cope for anything other than loading trains from ore patches.

Any time you find yourself dismissing something many people put a lot of work into, you should be suspicious that maybe you aren’t operating in the same context they did. For instance a problem some balancers solve (but many do not) is lane balancing, such that each input has an equal chance to show up in either lane of the output. This solves problems with balancing the pull (input) side, as opposed to the push side, and can recruit idle machines that otherwise wouldn’t get work. This in and of itself doesn’t apply to fully utilized production chains but in times of lower utilization it may matter. As another example, maximizing freshness of components is a factor in the latest DLC and this is much easier to do if you can guarantee equal drawing from all assemblers.

Balancers are conceptually cool so I understand the appeal of trying to find aesthetically pleasing, compact or otherwise interesting designs, however I just don't think they have practical use outside of the two niches I listed (loading ores, and fixing mistakes.)

For instance on Gleba, you better have a plan to either use or burn the full belt. It doesn't matter if you're balancing the lanes, you need a plan for the same throughput of spoilage.

Equal consumption of all the miners on an ore patch. The opposite ore loading, unloading at an equal rate so trains spend the minimum at a station (probably the more critical part to balance since you can solve the input side with buffering). Taking multiple less than full lines down to fewer full lines in a way that is evenly distributed, either to conserve space or, in larger bases, UPS. On Fulgora, with uneven recycler load, balancing the flow of an unfiltered belt before you separate it. Those are examples I just pulled from my current play through but I could probably find more.

You need them for loading, unloading, and smelting.

Space Age (the most recent DLC) required me to use several 6:4 and 6:3 balancers.

I usually start by hacking it. Once I place a premade balancer, throughout dramatically improves.

I can see them coming in handy for large production lines and main bus inputs.

Bus inputs maybe, I usually just put a few splitters before and after tapping off of main bus lines. For large production like circuits I design for some integer multiple of my train length, usually 1x or 2x of four cargo wagons, so I stopped using balancers for that and haven't missed them. The only place I still use belt balancers, except to patch over a design mistake, is loading ore into trains. And for that it's almost always something to four, rarely something to eight. I have a book for all sorts of balancers but almost all of them go unused.

> I usually just put a few splitters before and after tapping off of main bus lines

You are literally describing balancing. You are using an ad hoc solution as opposed to a blueprint, which can be fine, but I wouldn’t dismiss the convenience of being able to belt off a line of variable consumption from your bus and then guarantee each line down the bus will continue to have a fair flow of components. You clearly appreciate that this is desirable, and are solving for it in your own way.

I wouldn't call priority output splitters a form of belt balancing, but I'll accepted there is some conceptual overlap. However:

If you're tapping off the side most lane it doesn't matter what balancer you have at the head of the bus, you have to "balance" it after that point.

Furthermore, if you need real balancers in the middle of your bus to prevent the later stages of the line from being starved, that's a failure to plan for sufficient capacity. Balancers are a convienent way to patch over that mistake, but that's the sort of thing I'm talking about when I say they're cope.

The balancers could literally be part of the plan in your bus. Maybe you branch off a line of iron and two of copper to a green circuit plant that is then coming back onto the bus and being used downstream by both modules and red circuits. You know that for the next few hours those lines will be completely consumed by the number of modules you’ll be creating but after that your red circuit production will be able to go full speed and feed electric engines for robots. Your argument seems to be that you should just have enough lines to do both at full speed. That’s reasonable for the late game, but earlier the needs of your production are variable.

Reply on news.ycombinator.com