Abstract
This paper (read: shitpost) proposes a novel solution to the early game quality farming boostraping problem. Through the use of speed beacon black magic and circuits so complicated that the author already forgot how it works, it is possible to use one building and one recycler to farm a recipe at all quality levels with a decent speed. It can constantly run the quality module 3 recipe at a speed of approximately 0.18/s, or achieving the equivalent of 6 machines (running the normal quality) plus 3 machines (running the green, blue, and purple quality level recipe). In other words, the proposed blueprint reduced the required number of quality modules from 49 down to 11.Introduction
At the time of this paper is written, there is generally two approaches to farm for legendary quality products. The maybe better approach is to insert quality modules throughout the product line, and hope that by the time of final product, there will be enough legendary components to craft the legendary end product. There are some obvious problems for this approach. To name a few: the hassle of managing belts with all different qualities, the lost opportunity cost when you could have used productivity module instead, and the hassle of setting up buildings for all different quality levels, for every product you have.The definitely easier approach is to use quality module for the end product only, and recycle it until you get the legendary version. The problem for this approach (and to a less degree for the first approach as well), is that quality modules reduce crafting speed while speed beacons reduce quality. To compensate, usually one needs multiple buildings for farming. Furthermore, a single building can usually only accept ingredients of a single quality. If one want to use the higher quality recycled components, one often needs to install additional buildings. While in general crafting buildings are not expensive, the quailty modules they require definitely are. Especially considering the fact that legendary quality modules are exponentially better in this craft-recycle loop.
All these observations convinced the author to invent a way that 1. allow the use of speed beacons[*1] for quality farming without sacrificing the probability, 2. use a single machine to handle all different recipes. The result is the Speed-Beaconed Recipe-Quality-Parametric Gacha Blueprints:
also a link to factoriobin: https://factoriobin.com/post/bygme2/1
Blueprint Description
This blueprint book contains two pairs of blueprints: the Gacha machine using EM plant and its design baseboard, and that using assembler and its design baseboard. For the rest of this paper, we will focus on the EM plant model.- example.png (1.05 MiB) Viewed 633 times
- batch size: how many crafts will the machine finishes before considering switching recipe, default to 2 for EM plant. This prevents the machine from reset recipe when there are still some productivity progress leftover. If you use a combination of productivity modules and quality modules, adjust accordingly.
- product: the product you want to farm for legendary quality.
After pasting the blueprint, one needs to manually change two things:
- change the machine to "set recipe" mode. I cannot enable it in blueprint because otherwise the parameter system cannot read the ingredient information,
- change the constant "5" to the number of different solid ingredients in the decider combinator between the EM plant and the recycler.
Mechanism:
The constant assemblers to the bottom left stores the ingredient information for all quality levels. the author wishes that there is a way to transform all signals to a specific quality level, but alas. The content in the requester chest, divided by the batch size, is checked against the constant combinator. It will pick the highest quality recipe with enough ingredients[*2] and store in the latch[*3] on the top-right area. The input green inserter[*4] will then insert the exact amount of ingredient for one recipe. Once the machine finishes, it emits a "F" signal which increments the counter to the right. If the counter reaches configured batch size, it will reset to 0, clearing the latch and reevaluate the available quality recipes. By default, only epic and legendary products are preserved, and lower quality items are recycled.
The speed beacons (an independent functionality from the quality-parametric circuits) operates in the following way: when the machine is not crafting (indicated by the missing "W" signal), the blue inserters will remove the modules from the beacon. The input green inserter is also disabled until all speed modules are placed in the steel chests. Once the machines starts, the speed modules are placed back into the beacon, accelerating the EM plant. Since the quality level is determined at the start of the craft, it speeds up the recipe without reducing quality.
The assembler blueprint operates in much the same way. I haven't found the need to roll for legendary items using other buildings yet, well I actually want to have some legendary tokamaks, but I have no where near enough ingredients for that I don't have blueprints for those buildings. But they should work much in the same way.
Caveats
[*1] I don't know when Wube will fix that, so enjoy while it lasts,[*2] there are two potential problems for the ingredient requesting/selection process:
- by default, the requester chest will only request one stack for each item. Thus, if "recipe amount * batch size > stack size" for any ingredient, it will not work.
- I abused one undocumented behavior of the selector combinator: if all inputs have the same value (1 in this case) and the same icon, the one with the highest quality will be chosen in the ascending selection mode. It might break with any update.
[*3] There was one time the latch and assembler content was out of sync out of 2 hours I was using this blueprint. I suspect it's due to the multi-cycle nature of the circuit network and some robot delivered something during the update cycles. I have no idea how to reproduce it or fix it.
[*4] This is one of the bottleneck (in addition to the recycler for some recipe). Upgrading inserter with higher quality helps a lot, but for some reason legendary inserter will break the ingredient counting/limiting mechanism. I suspect it's rotating faster than my combinators can compute. Maybe the machine content->stack size latency can be improved, but I have swore years ago to never touch latency optimization when I finished the computer architecture course.
Disclaimer
The author is not responsible for any wasted resources due to bugs in this blueprint or Factorio updates changing its dependent mechanisms.Citations:
The speed beacon part of this blueprint design is largely inspired by this reddit post: https://www.reddit.com/r/factorio/comme ... lity_farm/
