[Suggestion / RFC] Change ME Storage Components beeing extensive recursive crafting

[alphaNOVAircraft]

What do you suggest instead/what changes do you propose?

Currently ME Storage compnents dont bring any benefit when beeing upgraded in terms of capacity (they even get more expensive but not more capacity). Furthermore the high tier cells take insane amounts of cheap compnents that just slow down crafting (1 16M drive is 64 000 ulv circuits e.g.)) Maybe add direct recipes for storage cells higher then 1k that get aviable later so that a alternative to the currently recursive crafting only exits #

What is your GTNH Discord username?

alphaNOVA#9875

[Alexdoru]

I would like to note, a 64k ME storage drive stores the same number of items as a storage drawer with 4 emerald upgrades and a void upgrade.

ME storage components should be cheaper, not more expensive. Players should be incentivized to use AE2 for server and game performance.

But that’s only for storing one item, otherwise you would need 63 drawers with upgrades to do the same thing as the ME cell. And your argument also applies to vanilla recipes, a bunch of wood and emerald will always be cheaper than me components. So what’s your point ? We should make them dirt cheap ? No

And the point of this ticket is to make them cheaper, I proposes recipes that makes them way cheaper than what it is now

[Yarntheory]

That is a 64k drive storing a single item type. A storage drawer with all 5 upgrades holds 655,360 items, so it holds more than a full 64k drive. A 16M storage drive, with a full set of 63 types, should store 68,157,440 items, or ~1.08 million items per type.

[Sphyix]

I like the recursivity at 2, but I think it makes the 16384k way too cheap. What about just splitting recursion and doing recursivity at 2 for 1k to 64k and at 4 for 256k+?

Threw down some numbers just to see what's the result, 16384k are expensive as needed, rest is cheaper. Numbers for the 256k+ can be tweaked by adjusting the number of circuits and processors used. I'm just not sure about the 64k, maybe it's too expensive compared to the 256k?

[0lafe]

drawers have a lot of downsides to the drives though. You make a stack of drives and shift click them into a single block, but with drawers you need to place them down individually and fill them with upgrades. Not to mention they can't dynamically allocate their overall capacity between item types, giving some larger inefficiencies.

One issue here is that when cost is the only component, the balance has issues when looking across tiers. You generally are gaining quite a lot of processing ability each tier in the early game, and drawers are craftable in stone age. The upgrades are LV/MV, so by the time you reach EV and get AE, your processing ability has become so tremendous in comparison to your increased need for storage. There isn't a good way to give it a good cost for the utility within the tier.

Imo an easy solution is to give drawers other downsides that make them unusable by AE tiers, forcing the player into making AE drives. This allows us to give them any cost we want, and it would theoretically still be worth it. Personally I think putting more epoxy in them would be a fun change, as we have yet to really reach possible EV+ levels of oil derivative usage.

I think a very simple change could be to throttle the throughputs on drawer controllers. Let them check a drawer once per second or once per few ticks. That would allow them to work well for ores and dusts in early game, however any automated input would probably struggle and cause the player to look for alternatives. If the drawers controllers search less often, that would probably also be better for performance on what could be very large drawer networks.

[leagris]

Throttling drawer controllers, is one tool to reduce drawers spam, the other tool is to limit the number of controlled drawers and the area size, which are both configurable options (As I already featured the former. I can certainly turn controller search tick rate into an option as well).

Although, reducing the service of mod features shall remain a last resort option.

Priority shall be to preserve services and improve but not degrade content. (You don't cut-out bits of feet when shoes are undersized).

So, lets continue to focus on improving the service and cost/benefit ratio of ME disks and drives before cutting-out bit of the drawers.

[Dream-Master]

lets use a 3:1 ratio not mix it up. And yes plates are to cheap

[Prometheus0000]

ix?

[Dream-Master]

mix

[Alexdoru]

The lord has spoken, what should we put instead of the plate for 3:1?

[johnch18]

It should be NaqAlloy if we want to keep consistent

Plates should match the tier of circuits, and be the tier material. And replacing steel with naquadah alloy, really? And no tier uses this

I meant that the ZPM component should use Naquadah Alloy rather than Naquadria because ZPM components use NaqAlloy.

[NoX-programer]

So, how the recipes will look after all the decisions/suggestions? (less recursive (alt recipes with field gens are good! - maybe moar of them?), 3:1 ratio, maybe some fun chem like polycaprolactam, epoxy etc.)

[repo-alt]

The recipes being put to vote are pure madness, we should promote higher tier storage, not forbid them. Otherwise we should forbid storage buses on super chests, because no sane person would spend a lot of UHV circuits on a cell, that does the same as storage bus (a bit more efficiently for the server, but the player doesn't see that)

[Prometheus0000]

So comparing them to SC/QCs, the 256k holds ~2M, 1/2 of a SC I, the 1024k ~8M about the same as a SC II, the 4096k ~32M about the same as a SC IV, and the 16384 ~128M which is the same as a QC I.

So why should they be more expensive? You're going to be formatting them to hold one thing at this level, so multiple items in one cell is irrelevant. Why not just add a SC/QC -> Component recipe?

[mitchej123]

Why not just add a SC/QC -> Component recipe?

<3

[botn365]

So why should they be more expensive? You're going to be formatting them to hold one thing at this level, so multiple items in one cell is irrelevant. Why not just add a SC/QC -> Component recipe?

i for sure dint ae was general storage al spesific mass storage was QC/infinity schest i would stil have liked QC more sinse they have everflow void but they have trouble saving their contents

[repo-alt]

So, essentially you say that we don't need larger storage cells and promote Quantum chests instead? That is probably ok

[Prometheus0000]

I say add voiding to cells (workbench option, like formatting) and and even larger ones. And/or single type ones, and/or voiding ones.

[repo-alt]

Add a separate issue about that, maybe that is not so hard to implement

[repo-alt]

If some non-recursive version of recipes will be accepted, I think recursive recipes need not be removed, because they are useful to recycle low tier cells.

[KamenriderZO]

In my opinion : No matter which type of components, they all can hold only 63 types of items. Higher tiers only increase the storage.So the recipe just need 1 lower Tier component. We should use 1:1 recursion. And other material could be expensive:maybe +4 different type of chest or single type item storage (super chest) and other things. To consider function.1k and 2048k they all just save 63 types. so it means just upgrade the capacity. If you use 4:1 or 3:1 recursion,why the higher tier components can't save 63*(4n) type items???

[querns]

Perhaps making circuit assembly line recipes for ME storage components might be of help? Not only does it allow for easy batching, but there can be SoCs involved to ease the material requirements (or, heck, even entirely new wafers made with fluix/certus/nether lenses or whatever.)

[Glease]

So, do we have a consensus now?

[github-actions[bot]]

This issue is stale because it has been open 90 days with no activity. Remove stale label or comment or this will be closed in 3 days

[bombcar]

https://github.com/GTNewHorizons/NewHorizonsCoreMod/pull/241

[ArtherCasan]

This is too cheap now. I think multiply recipe time by 20 or 30 times and add a same voltage level Field Generator each recipe will be ok.

[Sphyix]

This is too cheap now. I think multiply recipe time by 20 or 30 times and add a same voltage level Field Generator each recipe will be ok.

Adding same tier field gen would make them go tier+1 Multiplying recipe time by 30 just doesn't make sense. The whole point of this is that right now they are kinda cheap already, but they take a lot of time to make because of the recursion.

[repo-alt]

I think the recipes are (finally) ok, because the situation when infinity chest is way cheaper than 256 max size cells is weird (and bad for servers).