Make salvage module scaling linear (and nerf single blocks)

[Valiant70]

Test: I built two salvage arrays. Each had the same number of groupings. One array had one module per group and the other array had 30 modules per group. I tested both on an asteroid. The array with 30 blocks/group was faster, but it was CLEARLY not 30 times faster.

Proposition: The scaling of salvage modules is inconsistent with weapon blocks which now scale linearly in performance. It also makes honeycomb salvage arrays of single blocks absurdly overpowered, especially when combined with a cannon support system. I think salvage module performance should scale linearly with the power of a single block heavily reduced from what it is now. This will give the honeycombs appropriate performance for the number of blocks they contain and make salvage performance scaling consistent with the rest of the game.

On a side note, the cannon slave effect may not be to blame for the overpowered arrays. If the power of each block is appropriate, such arrays should be balanced, at least theoretically.

 

[FunnyBunny14]

Calbiri has said that he is aware of the cannon linking problem and is going to change it.
And I doubt they will change the scaling of salvage cannons. Bigger arrays would otherwise seriously create lag. Smaller arrays would also be way to weak.

The power already stays the same. I do agree that non-linked salvage arrays should require less power. And I definitely agree that cannon linked salvage arrays need to require more power in order to be balanced.

 

[Valiant70]

And I doubt they will change the scaling of salvage cannons. Bigger arrays would otherwise seriously create lag. Smaller arrays would also be way to weak.

Currently, the most powerful arrays are the lag-causing bazillion-single-blocks-in-a-honeycomb arrays. These need to be nerfed HARD while keeping the block-to-performance ratio of medium sized groupings similar to what it is now. If people get in the habit of building huge arrays that lag the snot out of servers due to too many block updates, then the power requirement of these arrays should be increased.

 

[FunnyBunny14]

You mean exponential power usage? That was tried before, didn't end well. It got removed and it now is the linear power usage as we know it.
You also obviously haven't tried the current salvage - cannon combination. If you thought massive arrays were bad, wait 'till you see how tiny arrays work. Trust me, you are in for a shock. No downside, power usage is 5 times as low and the overall efficiency is about 5 x 5 x 10 = 250 times bigger.

 

[jayman38]

It sounds like the slave weapon slot on salvage systems simply needs to be removed. Then you can't link anything to your salvage array to increase performance.

Sadly, you would also lose the ability to create "shotgun" salvage beams. Maybe that could become the new normal salvage beam system (shotgun beams from a single emitter), with multiple beams per salvage beam group, with each individual "shot" in the shotgun group having weaker power.

As for requiring more than one salvage block per beam, maybe subtract the effect of two salvage blocks from every beam, so that you require an absolute minimum of 3 connected salvage blocks to have any effect whatsoever. (Hmm... A short-range, no-effect beam for decoration could be interesting....)

 

[FunnyBunny14]

It sounds like the slave weapon slot on salvage systems simply needs to be removed. Then you can't link anything to your salvage array to increase performance.

Sadly, you would also lose the ability to create "shotgun" salvage beams. Maybe that could become the new normal salvage beam system (shotgun beams from a single emitter), with multiple beams per salvage beam group, with each individual "shot" in the shotgun group having weaker power.

No, the effects were just literally copy/pasted from the weapons. And since salvage doesn't do any damage, all the normal nerfs (and in the case of the damage pulse, buffs) don't apply. Same goes for the astrotechnician, power supply/drain beam and shield supply/drain beam.

 

[Valiant70]

You mean exponential power usage?

No, something like a gently-sloping linear increase in per-block power usage would work better. Exponential power usage is a bad idea.

You also obviously haven't tried the current salvage - cannon combination.

Incorrect. I have tested them quite a bit.

If you thought massive arrays were bad, wait 'till you see how tiny arrays work. Trust me, you are in for a shock. No downside, power usage is 5 times as low and the overall efficiency is about 5 x 5 x 10 = 250 times bigger.

I'm well aware of that, and it's a different topic.

It sounds like the slave weapon slot on salvage systems simply needs to be removed. Then you can't link anything to your salvage array to increase performance.

That's a different topic.

 

[FunnyBunny14]

So what you're saying is that you want the speed of salvage cannons and power usage to change?
With the speed being (initial speed)/(amount of salvage modules in an array)?
And the power to be initial power + (X amount) x ( (amount of modules) - 1 )?

But what do you want the initial power to be? And what do you want the initial speed to be?

Those are the only questions I still have.

 

[Valiant70]

I'd like to see something more like this:

speed = (number of blocks) * (base speed)
where the base speed is the salvaging speed of one block.
Base speed should be a config option, and set by default to about 2 sec./block.

Power usage per block = (base power)+0.002*(number of blocks)*(base power)
capped at 5 times the base power usage

Thus an array with 500 blocks would use roughly twice the power per block as a single block. This is a gentle, linear increase per block. Since salvage arrays are usually used in a lot of smaller groups, it might be better to count "number of blocks" linked to the salvage computer rather than blocks per grouping.

 

[FunnyBunny14]

Power usage per block

linear increase per block

Yes, but that is still an exponential increase in the whole. Say the base value is 1000 e/s. If you had 500 blocks in your array (I assume the computer isn't counted) the power usage would be 500 x (1000 + 0.002 x 500) = 1 million e/s. Yeah, that isn't a gentle increase. That is salvage murder.

I still think that this would be better:
power for array = base power + (X amount) x ( (amount of modules in array) - 1 )

 

[Nebulon-B_Frigate_FTW]

Personally, I never liked how you cant have beam outputs adjacent or they count as one group. It's fine for weapons, but it makes salvagers annoying to build, which is part of why single-block groups are thing now.
I think salvagers should work like this: Total blocks/sec of salvage scaling linearly with number of salvagers, then that is split between each designated output. You could make a super-salvage-needle or a big flat salvage beam array. Single blocks would be the same as non-single, so it should fix the issue.

 

[FunnyBunny14]

Personally, I never liked how you cant have beam outputs adjacent or they count as one group. It's fine for weapons, but it makes salvagers annoying to build, which is part of why single-block groups are thing now.
I think salvagers should work like this: Total blocks/sec of salvage scaling linearly with number of salvagers, then that is split between each designated output. You could make a super-salvage-needle or a big flat salvage beam array. Single blocks would be the same as non-single, so it should fix the issue.

Only if you SHIFT + V does it count as a whole group. You can jsut use two different computers with logic to have a compact automated array.

 

[Valiant70]

Only if you SHIFT + V does it count as a whole group. You can jsut use two different computers with logic to have a compact automated array.

Apparently I wrote that when I was too sleepy to math. XD The idea is that extremely large salvage arrays need to be balanced by power usage, while small ones are not heavily restricted. This also needs to take into account the possibility of someone making a ton of little bitty groups in attempt to exploit it, and ignore the fact that the large numbers of modules are separate groupings.

I'm not going to write any formulae tonight as I'm half-asleep again... oh joy.

 

[FunnyBunny14]

Small arrays should be balanced by their speed, or rather the lack of. Big arrays should also indeed need more power, but not too much.

 

[Valiant70]

Small arrays should be balanced by their speed,mor rather the lack of. Big arrays should also indeed need more power, but not too much.

That was exactly my point. It seems reasonable, though, for large arrays of salvage modules to be somewhat less efficient. I don't mean ten blocks versus one. I mean ten blocks versus a planetary-scale miner.

 

[FunnyBunny14]

Maybe the power usage will get changed in the next update. Calbiri is working on a new value for salvage, maybe he'll change the way power works for salvage too.

 

[Valiant70]

That's interesting. Has he said anything besides "a new value?"

 

[FunnyBunny14]

He didn't say anything. I found it with the help of a diff program. *Going to one of many threads were I posted it*
Ah, here it is:

<SalvageDamageNeededPerBlock>0</SalvageDamageNeededPerBlock>
<!-- 0 will salvage on tick, otherwise the damage has to be met on the object to salvage one block -->
<SalvageDamagePerHit>100</SalvageDamagePerHit>
<!-- salvage damage done per tick to fill up SalvageDamageNeededPerBlock -->

It would appear that in the future salvage will need to do a certain amount of 'damage' to blocks (I'm guessing it will be the block hp + armor) to completely salvage that block. This would mean that all the slave effects would work in their current state. And like I said, maybe that will meant that the power usage for bigger arrays will, well, be bigger.