, computers.
While there are a few automall designs flying around none quite did what I wanted them to, so I made my own design to fit certain criteria. In my mind the advantage of an automall is you only need to configure it once and it maintains high throughput by using all its assemblers to produce the items there might be a spike in demand for. That immediately ruled out designs that crafted logistics network requests and designs that recursively crafted recipe ingredients at a slow pace.
So here's mine:
- Screenshot 2025-01-01 233657.png (478.65 KiB) Viewed 2184 times
- Will craft the recipes you enter in the recipe list and keep them stocked if they drop below the desired buffer quantity
- Maintains high throughput by reducing or eliminating idle time.
- It tries to pick a new recipe if the inserters aren't inserting ingredients, if the materials aren't available in the logistics network or if demand is sated.
- Infinitely tileable.
- Designed to fit neatly in a substation network.
- Has a fancy off switch.
- Recipes must be chosen once manually and entered in the constant combinator (which would have been less of a chore if I bothered to do it myself before sharing the blueprint, but oh well).
- You must manually select the idle inserter timeout period depending on how fast bots deliver ingredients to the chest. However it can usually be set very low if there isn't a severe shortage of everything.
- Will drive people nuts with my misuse of technical terms in the combinator descriptions (I've no background in anything pertinent to computer science, programming or anything of the sort).
- The automall checks ingredients for availability only after the recipe is queued up. This reduces throughput by queueing up invalid recipes until they're cancelled. I've already devised a way to prevent that by checking before the recipe is queued but it was too bulky so I didn't implement it until it's refined.
- The single selector combinator + clock combo can be replaced with a signal iteration device such as seen here. This would make the blueprint usable before advanced circuits are researched. However I've already spent too long making this thing so this is another improvement for a future version.
- Can be made to work with fluid recipes by replacing one of the active providers with a sushi pipe and a filtered pump. This I don't intend to bother with in any future version.
Tileable stage. Align it with the initial assembler and place as many as you'd like. More machines=more throughput.