Add support for VoidProtection of Items to MultiSmelter

[Worive]

I may lack context as I'm an early LV player, so I'll describe quickly what I've done (to be sure I'm missing nothing):

For each possible recipe, I'll check the max possible parallel according to the VoidProtectionHelper, and limit them according to it. If after checking each recipe I end up with an empty output (so everything at 0 parallel), I assume it's because there's not enough space.

Close GTNewHorizons/GT-New-Horizons-Modpack#17086

[HoleFish]

This "void protection" doesnt provide void protection. It will void stuffs if output bus is about to be full; it only provide protection when output bus is full already.

[Worive]

This "void protection" doesnt provide void protection. It will void stuffs if output bus is about to be full; it only provide protection when output bus is full already.

@HoleFish I did not manage to reproduce any voiding of materials, but I've noticed I missed a miss calculation of remaining cost, which would make it do 32x stone then 15x stone then 15x glass (3 steps), instead of 32x stone then 15 stone and 15x glass (2 steps), I'll push this fix, but I would gladly like if you could tell me how did you void items 🙏

In my setup, I'm using a LV energy hatch with Nichrome Coil Blocks, with 1 input hatch and 1 output hatch, but I've also tried with 2 output hatches.

[HoleFish]

2 random types of smeltable item in input bus and a ULV output bus. Multismelter reasonably start processing and void one of them with this void protection enabled

[Worive]

2 random types of smeltable item in input bus and a ULV output bus. Multismelter reasonably start processing and void one of them with this void protection enabled

@HoleFish Alright, I noticed I was only trying with same item slot voiding kind (test 2 -> 5), and now I fixed the case of test 1 which is yours (somehow did not happen to me to test this situation, sorry). I've put down all the tests I've been doing to be sure I did not forget about another situation

Context: Basic Multi Smelter with a LV Energy Hatch and Nichrome Coil Blocks, with 1 input and 1 output bus and I smelt sand and cobblestone to have two types of output.

Test 1 - 2 Items with one slot (HoleFish's test)

• Output bus with 1 available slot
• Input bus with 1x of sand and 1x of cobblestone

Process

1. 32x Glass
2. 15x Glass, 17x Stone
3. 30x Stone

Test 2 - Simple one output

• Output bus with 1 empty slot available
• Input bus with two stacks of sand

Process

1. 32x glass
2. 32x glass
3. Not enough space

Test 3 - Non-stack output available

• Output bus with 1 slot with 1 glass, the rest with cobblestone
• Input bus with two stacks of sand

Process

1. 32x glass
2. 31x glass
3. No enough space

Test 4 - Testing normal functioning with multiple recipes

• Output bus with 1x glass, 1x stone, and two slots of cobblestone
• Input bus with 64x of sand and 64x of cobblestone

Process

1. 32x stone
2. 1x glass, 31x stone
3. 32x glass
4. 30x glass

Test 5 - Testing non-stack with multiple recipes

• Output bus with 17x stone, 17x sand, and two slots of cobblestone
• Input bus with 64x of sand and 64x of cobblestone

Process

1. 32x Glass
2. 15x Glass, 17x Stone
3. 30x Stone

[Worive]

Current known issues:

• Batch mode's time is not adapted to the limited output rate by VoidProtectionHelper
• Some configurations of VoidProtectionHelper will give incorrect getMaxParallel() results (wrong usage of VoidProtectionHelper surely)

Everything working well now ✅

[Worive]

Ended up adding the ability to the VoidProtectionHelper to use virtual output bus items, to properly check for cases like multi-smelter.

Seems to be the more fitting solution and can be used in the future by any other machine requiring it

[HoleFish]

2 stacks of sand, ULV output bus
result: 128 glass processed and 64 voided

[Worive]

As it it would be so simple to implement virtual stacks output to VoidProtectionHelper, fixed that and did all my 5 tests listed above, all working 👍 (in both normal and batch mode)

[HoleFish]

Stocking input bus, 192 random ores + infinite sand, LV output bus
result: 192 ingots + 192 glass processed, 128 glass voided

[Worive]

Ready for another checkup 👍

[HoleFish]

I still suggest you move this void protection part and merge into Calculate parallel part, so that subtick energy cost will be correct

[Dream-Master]

@Worive any news?

[Worive]

@Worive any news?

Sorry I only talked about it on discord, forgot to mention it here too:

I apologize, but I don't feel like I'm able to finish this task, just looking at the code regarding what is left to fix simply bricks my mind and prevents me from thinking further. I would greatly appreciate it if someone with more experience could finish what is left of it and explain their thinking while doing it 🙏

Wish I could have finished this task by myself 😞

[Dream-Master]

@HoleFish can you help here ?