so, I was thinking. Rather than having just static alt recipes from hard drives. It would make an interesting tech tree change to have them be researchable through actual production usage. and they could be upgradable with experimentation / research.
so, in my idea there would be some sort of R&D production building that you could feed with sets of materials and there would be a chance It would generate an alt recipe using those components. so like “here play with this stack of wire and some plates and see if you come up with something.” It would have a chance after every x number of input cycle to make a new recipe. This would allow a TON of alt recipes rather than just a few based on collecting hard drives. Also, once you had a given recipe it can start at a lower un optimized level, and you can research it over time to get more efficient / faster at making that thing. I don’t know if the improvement should be in less materials used, or more output items or production time (or maybe have 3 separate R&D projects to improve your choice)
Regular recipes could also be improved like this so say you have your research division crunching on super computers, and after 10 cycles it gets a multiplier of 1.1 to the production time, then after 50, 1.2 then 100 1.3 so it takes exponentially more and more research to fine tune and perfect a given process while the initial low hanging fruit improvements are easy.
This would add a large amount of granularity to the tech tree and let people improve on their factory rather than being being able to just calculate number of machines and resources, or just looking up boring ratios and that’s it.
Its not unlike the infinite research chains in late game factorio, to allow improvement