Cleanroom ideas

[protocol-1903]

Melon, can we get a cut and dry description of your cleanroom idea as-is for a basis?

[notnotmelon]

Actually this was @przemo1232

[protocol-1903]

In that case, przemo?

[protocol-1903]

Well at any rate, I have some ideas I'll just dump in the meantime.
Pressure. Certain cleanrooms need to be kept at a controlled pressure, usually a vacuum. Certain processes can only create a specific purity of vacuum, say 95% or 99% or 99.95%, which can be fine for some cases but not others. Not keeping a good seal would do something bad, as simple as shutting the plant down or as complex as manufacturing bad parts or implosions. This requirement could possibly also be ignored if in space, or used to specify an argon rich, or oxygen deficient environment, meaning the player must filter and purify the cleanroom before use.
Bacteria. In addition to vacuum, certain cleanrooms cannot have bacteria or moisture present. Gameplay wise, this could be as simple as a recipe taking "impure" air from the cleanroom and filtering it using special filters that get better or last longer as tech level increases.
Both of the former ideas would be a great way to utilize the vacuum pump, in my opinion a heavily underutilized feature for gameplay.
Tools and molds. Certain operations require specific tools and parts that wear over time, more so than general machines. They would need replacing, and automatically. I posit either a system similar to the vatbrains, where subcomponents would not work unless supported by a similar device (where better "cartridges" produce better effects) or a beacon where the modules spoil. In either case, the tools/molds/parts are used for a period of time before being discarded and reprocessed or scrapped.
Other gimmicks could require the player to wear a bunny suit when in an operational cleanroom, requiring coolant to be run with pipes to certain machines, running "power" (a fluid) to machines through specialized pipes because Factorio's quasi-electrical connections "throw of the balance of electrons in precision manufacturing of microelectronics", or bots being disallowed in cleanrooms because of the pollution caused. There's also a number of fun API features that could be used here that would be rather annoying across all of Pymods.

Additionally, I'll be starting working in an electronics nanofab in a month's time, so I can hopefully bring some interesting insights from what an actual cleanroom needs.

[przemo1232]

What i came up with that in my opinion wouldn't be annoying was a compound entity with an assembler and some unit with an equipment grid, the player edits the grid like a puzzle to get bonus prod, speed or unique features in the end product. There are no inserters, belts or anything other than machines that can directly input or output components, and depending on where u input a component into a machine you'd get a different modifier (for example inputting a cpu from the left side of an equipment grid machine would give bonus prod, but inputting it from the bottom would give a chance to get a better item). The finished grid would be analysed and set the recipe of the assembling machine that actually makes stuff. It would be used in items not made in bulk so for example modules or maybe some augments to the player.

[protocol-1903]

I personally think that a factorissimo style building would be more interesting, and discussing the idea with some people elsewhere, they seem to agree. Given vector_to_place that was added to crafting machines, it would be trivial to have non-belting and non-inserter cleanroom configurations (although perhaps cleanroom inserters/item transport can be special item sets).

I also believe that a set of possible machines to produce different components (say, 4 different chipshooters) of different sizes, power draws, resource efficiency (productivity), and speeds, would be an interesting enough mechanic to replace the bonuses provided from inserting items from specific directions. Perhaps a small beacon and specific new modules can be used to augment this mechanic.

[MrSmoothieHuman1]

+1 on protocols idea, i'd much prefer a factorissimo style building where the player can enter and build the factory itself, while also having to keep it maintained with all the things it needs in the outside world (cleaning filters/clean or inert gases/sterilisation methods).
A random idea but i'll throw it out anyway, it might be neat if you have to sterilise the components you want to use in the cleanroom itself, wether that happens inside or outside im not sure - but sterilisation could be done by various different methods? gamma rays/different radiation techneqiues, cleaning solution, anti-bacterial materials, etc etc

[protocol-1903]

Upon further thought on the cleanroom design space I have come to a couple of conclusions:

1. If implemented, there should be enough freedom for designs to be unique and personalized while still being considered optimal by one metric or another. This means that a direct-insertion enforcement would be fine, as long as sufficient space is given so that a number of designs can be constructed.
2. I agree that belts should not be allowed in the cleanroom. I believe that items (non-fluid entities) should be transferred into the cleanroom via a passthrough, or a linked container to both the inside and outside of the cleanroom. It should be implemented similar to the linked chest mechanic for Factorissimo. Additionally, I believe that some form of logistical latency should exist so that items inserted into the passthrough cannot immediately be removed, which might require the use of monodirectional passthroughs instead of universal passthroughs. Certain fluids may be transported inside the cleanroom using specialized pipes (and inserted via external "injector" entities that isolate the material for cleanroom use), but I think it would be interesting if no underground pipes were allowed, or only short range undergrounds were allowed. Enough to get by an inserter or two, but not enough to get past an entire workstation. This way, where you insert fluids into your cleanroom is more than just an afterthought, allowing more unique "ports" on the designs.
3. Cleanrooms should have external infrastructure to manage the conditions inside the cleanroom. This could include one or more of the following: Temperature, humidity, air pressure (or lack thereof), air composition (normal Nauvis air, or mostly nitrogen?), possible airborne bacteria, light levels and makeup (yes, different light levels affect things. It's mostly just UV light, so specialized non-UV producing light sources must be used), gravity (maybe it speeds some processes up or slows others down, also not entirely realistic but hey maybe some late game stuff can affect the gravity of certain areas, call it gravitonium or something). This external infrastructure could be cycling the air in the cleanroom to remove contaminants, strictly managing the temperature in the cleanroom (via heatpipe or "thermofluid" that must be heated/cooled to the right temp range), removing air from the cleanroom to ensure the right level of atmosphere is present (via a pipe output from the cleanroom that slowly adds air (infinity pipe) and/or adds air due to external factors in the cleanroom, like gas leaks or processes completing), or for the UV stuff, making sure to have the right lamps in the cleanroom that have a specific wavelength range. Some of these systems could come down to placing the right entities in the right places, others could be an "upkeep" system where it's a resource drain on the factory.
4. Some recipes should be generic. Others should require specific cleanroom conditions to be met. Stuff like no bacteria present in the air, or the humidity must be within a certain range, or the atmosphere in the cleanroom can't contain a specific gas. This could be enforced by shutting down machines with recipes that do not have the proper conditions, but I think a more interesting alternative would be to have less precise results when using recipes in undesirable conditions. Read: you produce more "junk" or "broken" products and useful products are less common. Sure certain ranges of conditions might always produce undesirable results, but even just a little bit outside the optimal range should be, well, below optimal instead of completely impossible. This also opens the door for progression mechanics where you need to keep the cleanroom at a condition that is unreachable with your current materials access/tech level, but you can approach that threshold and get the tech/materials you need at a less optimal rate, then upgrade your setup to produce at more optimal rates (if recipe switching isn't implemented, I could see this being done as a base productivity of +400% [5x products per input] as a base value, on a recipe that has a 20% chance of success so it returns one good output per input on average [or lower if desired], then decreasing returns depending on the severity of conditions) Though this is one option, I can also see certain processes where the lack of optimal conditions doesn't change the failure rate, but makes the processes slower (you need to take more time to do something, like cooking wafers or some such) and/or more energy intensive. I feel like these would be sufficient penalties for improper systems as to promote proper conditions, but suboptimal conditions are still an option if need be.
5. I have other ideas that could be interesting systems if added (liquid gases for use in specific recipes that have abnormal upkeep mechanics, spoilage used as a half-life simulator, and more) but I think those should be presented on a separate, unrelated thread.

Sorry this turned into a bit of a brain dump lol

[protocol-1903]

Hey guys, I'm back...

Some new ideas to throw against the wall:

Back to przemo's idea about puzzling together buildings with arbitrary non-rectangular shapes, using proxy inventories and arbitrary effect scripted beacons it can be done, in-world, and rather efficiently at that. I can also see an implementation where linked containers between surfaces would be okay, but with proxy inventories it could be done much more cleanly, such as creating walls/surfaces on the outside of a cleanroom that directly interface with assemblers and the like on the inside (also, gonna put in an API request for proxy fluidboxes, but that one might not go through because linked pipes exist. but it would still be cool).

On the topic of arbitrary effect scripted beacons... How about assembling machines with an overheating mechanic? Use it too long, and it slows down significantly. Or, perhaps a bell curve where it speeds a little bit, then slows way down after that. (Or other effects like prod, efficiency, etc...) So that running a machine at 100% usage isn't optimal. Or the reverse, where machines need time to spin up to full speed, then slow down after not being used.

And on the topic of proxy inventories, how about making modular assemblers? You start with a core entity, perhaps 1x1 or 2x2, which is hotswapped via scripting based on the shape, count, and alignment of proxy inventories as 'modules' or 'components'. Each hotswapped assembler can craft different recipes with different properties. Like, say, adding a welder and an extruder onto a modular assembler allows you to make recipes using the welder, or extruder, or both. As opposed to having multiple entities for each type of production. It would also allow for more unique designs, especially with non-rectangular combinations.

Anyways, back to finding new things to tinker with...