[0.17.7] Pumps substantially slower than 0.16.x, but no changelog entry -- intentional or bug?

[dewiniaid]

This was mentioned on the Discord earlier, so I tested it.

A test map for fluid wagon unloading speed (full fluid wagon to pump directly to storage tank, with the storage tanks being voided via pump to infinity pipe after the train leaves) yielded the following results in 0.17.7:

• 380 ticks for a train to leave the station when unloading with 1 pump.
• 202 ticks for 2 pumps
• 152 ticks for 3 pumps

A similar test map in 0.16.x yielded the following:

• 206 ticks for a train to leave the station when unloading with 1 pump.
• 132 ticks for 2 pumps
• 109 ticks for 3 pumps

There is nothing in any 0.17 changelogs I can find regarding pump speeds and whatnot, so I'm posting this here in case it is an unintended change.

[BlueTemplar]

I see that some lines have been removed in fluid.lua :

Code: Select all

    pressure_to_speed_ratio = 0.4,
    flow_to_energy_ratio = 0.59,
    -- fuel_value = "8MJ",

These are for 0.16.51 oil, but all fluids have them set the same in vanilla. (Except for the commented out fuel value...)

I'm guessing that these variables
(which if I'm not mistaken account for wave speed and viscosity)
have been removed in the transition from 0.16.51 to 0.17, until the time that the fluid algorithm is fully updated,
and whatever has replaced them is not exactly the same ?

(Pumps might have been affected, because even though they don't use fluid mechanics on input, they still do on output ?)

[BlueTemplar]

Also, like in 0.16, steam still has this property :

Code: Select all

    gas_temperature = 15,

I wonder what's up with that ?

(Also, a quirk, like in 0.16, the default_temperature for water and steam is 15, but for the other fluids 25...)

[GotLag]

(Also, a quirk, like in 0.16, the default_temperature for water and steam is 15, but for the other fluids 25...)

This is what enables the long-time feature of engines destroying other fluids. Their higher-than-background temperature and minimal heat capacity mean there's a tiny amount of energy inherent to them so a steam engine will consume them to generate a negligible amount of power.

[Dominik]

I see that some lines have been removed in fluid.lua :

Code: Select all

    pressure_to_speed_ratio = 0.4,
    flow_to_energy_ratio = 0.59,
    -- fuel_value = "8MJ",

These are for 0.16.51 oil, but all fluids have them set the same in vanilla. (Except for the commented out fuel value...)

I'm guessing that these variables
(which if I'm not mistaken account for wave speed and viscosity)
have been removed in the transition from 0.16.51 to 0.17, until the time that the fluid algorithm is fully updated,
and whatever has replaced them is not exactly the same ?

(Pumps might have been affected, because even though they don't use fluid mechanics on input, they still do on output ?)

These values remain the same, they have only moved from the prototype into code. So this is not the reason.

[Dominik]

I have done a lot of work on the pumps that was related to the fluid mixing mechanics. Right now I am not aware of anything that should cause difference in throughput. It is probably a result of interaction of this new code and the old fluid algorithm, because the code was written with the new algo in place and with the old it works somehow differently. Nerfing the pumps was planned anyway (is postponed) as their pumping speed is ridiculous and is a mistake. So do not consider this a bug and it will be further addressed together with the fluid algo.

[BlueTemplar]

These values remain the same, they have only moved from the prototype into code. So this is not the reason.

Temporarily I guess, while the kinks with the updated algorithm are worked out ?
(Feels like modders should still have access to them...)

[Dominik]

These values remain the same, they have only moved from the prototype into code. So this is not the reason.

Temporarily I guess, while the kinks with the updated algorithm are worked out ?
(Feels like modders should still have access to them...)

The new algorithm used different fluid properties. We will see how it will come out in the end.

[Erythion]

These values remain the same, they have only moved from the prototype into code. So this is not the reason.

That said, can modded fluids still use these properties in the prototypes in .17.X, or are these variables entirely removed from prototype declarations?

[Dominik]

They have been removed as a part of optimisation, not accessible at this point.

[djedeleste]

Would it be possible to get an indication of what the target throughput of pumps will be ? (roughly at +/-1k, or similar to current or something ?)
Would lessen rework if we can start planning around the target already.

NB: i hereby promise to not complain if the end implementation ends up differing wildly :p

[Dominik]

I think that 3k is reasonable. But more people will have to agree to a number.

[Kingdud]

Dominik, could we get the current pump as a 'tier 2' pump (make it require space science for all I care) just to make mega-basing not a huge headache? 3k/s makes sense for pre-rocket factories, but not for post-rocket megabases. I know mods exist for that problem, I just live in a world where stuff you guys make is wayyyy more UPS efficient than what mods offer.

[Ultros]

Dominik, could we get the current pump as a 'tier 2' pump (make it require space science for all I care) just to make mega-basing not a huge headache? 3k/s makes sense for pre-rocket factories, but not for post-rocket megabases. I know mods exist for that problem, I just live in a world where stuff you guys make is wayyyy more UPS efficient than what mods offer.

Agreed on this, I've had to resort to waterfilled chained pumps (technically a vanilla-compatible feature) in order to get megabase pipes to not be so taxing on the CPU. Even then, the waterfilled pumps put out lower fluid (~6k/s vs ~12k/s possible in 0.16, going by the nuclear blueprints for 0.16 assuming that level of water flow and bottlenecking on water).

[Dominik]

Making a pipeline of pumps is definitely not the intended way it should work. It is supposed to be another logistical challenge. The limit is the pipe throughput, the pump max should always be above it. Of course if the limit is the performance, that is also not right, but it simply can't be infinite
If a mod only changes the constants, such as the pipe sizes, it does not affect performance. So mods should be the way to go if you want to do kinky stuff to increase performance.

[Dominik]

another report on this viewtopic.php?f=7&t=69045

[Bilka]

So what exactly is the current speed of the pump?

[Kingdud]

So what exactly is the current speed of the pump?

The upper bound seems to be 7133 units/s. It's not consistent, the pumps can easily end up deliver less than that (even between a full tank and an empty tank). If you download 'New Oil Layout - bugreport' from viewtopic.php?f=7&t=69045 and play with it, you can replace the fluid void with a pump then a fluid void and see how it changes the output. The two pumps going into the central tank will work at ~5500 unit/s, but if you place a pump extracting from the central tank then going into the fluid void, you'll see ~7133 unit/s on that single pump's throughput.

[Ultros]

So what exactly is the current speed of the pump?

I tested this based on a blueprint of a pump-bus nuclear reactor design from 0.16 (which used to do the full 12k water/s). The sum of the steam output after reaching equilibrium is around 6.8k per second going by the sum of the steam consumption from the turbines.

[MasterBuilder]

As a baseline comparison, moving barrelled fluid on a full blue belt is 2250/s (45*50). Making pumps 3k seems reasonably close with a nice advantage (750) over belts.