-
Notifications
You must be signed in to change notification settings - Fork 20
/
TODO.txt
23 lines (20 loc) · 1.76 KB
/
TODO.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Write schema documentation for data model jsons and model tier jsons.
Refactor Simulation Chamber to have Training Mode and Inference Mode, so you don't gain data and run the model at the same time.
* Needs an additional button and GUI element to display the current mode
* Needs additional fluff text during each mode
* Doesn't really need many logic changes beyond locking the generation of data / items past the current mode flag.
* Current mode has to be synced (at least via menu) to client.
Update Loot Fabricator to have better selection process - fabrication lists, round robin mode, etc.
* Needs a new GUI element to select the production mode
* One of: Fixed, Round-Robin (between selected, defaults to all selected), and Production Queue
* Production Queue allows specifying a list of target outputs that will be executed in-order, looping when finished.
* Needs new internal storage to hold the specified mode as well as the additional data used by the mode.
* Technically, round-robin can be treated as a subset of Production Queue, so we could skip it.
* Needs new logic to pick the target output and continue along the line.
Add the Data Center, a multiblock combining the work of the Simulation Chamber and the Loot Fabricator.
* Holds up to 25 data models, of which it can simulate and perform loot generation for.
* Only works with Self Aware models (configurable?)
* Total multiblock is 7x7, where the center is air and the walls are some kind of glass.
* Visually, each model being worked on by the data center will be shown spawning and dying in the middle over the runtime of the block.
* Automatic processing of output items will be done through Production Queue rules, similar to the Loot Fabricator.
* Uses 150% of the normal simulation power for each model.