FIRS Industry Replacement Set - Development

[aantono]

Not sure if this is a bug or what's going on, but for some reason, I have a train trying load Eng. Sup. from a feeder station, and the station shows 140 crates available, but the train is refusing to load them (it only loaded 1% = 1 crate). Any idea why that happenend? New deliveries of ENSP to the station get loaded onto the train just fine, but the earlier ones for some reason don't.

[Wahazar]

Is there a way to stop FIRS from creating gas stations next to streetcar tracks without road? It creates gas stations next to my industrial tram track and it is weird to pump gas into electric trams:)

Maybe build restriction to the city area could help a bit. Without this, in case of interurban roads, we can observe long chains of petrol stations
(unless 1 industry per city is allowed).
Additionally, some game script expect petrol delivery to the city to determine its growth.

2048x2048 maps are a bad feature.

I have completely opposite feelings convinced by multiplayer game, however it is complete off-topic here

[andythenorth]

The location check could search for house tiles, which has a high chance of locating the petrol station next to or near to a road.

I like the idea of remote gas stations on lonely highways though

[Wahazar]

I like the idea of remote gas stations on lonely highways though

Issue is, that they are not lonely - there is lot of them, and not only on highways:

gass.png

(979.33 KiB) Downloaded 15 times

Of course, forcing 1 industry per town should decrease such behaviour.

[Eddi]

2048x2048 maps are a bad feature.

designing a newgrf with a too narrow sense of play styles in mind is a BAD FEATURE.

[andythenorth]

designing a newgrf with a too narrow sense of play styles in mind is a BAD FEATURE.

Well we (you could design the algorithm, I'll put it in FIRS) could try and tweak payment curves for large maps. The industry clustering already adapts to map size...

I think it's a tricky design issue.
- optimising for the assumption of long distances might cause problems over shorter distances, which still exist (and may even be the common case) in a 2048x2048 map
- testing is a PITA

Alternative approach: try and enforce a maximum distance between ENSP / FMSP supply sources, and industries that consume them. This is also tricky, because
- many industries end up inter-dependent on locations
- it would be easy to imagine this resulting in a mostly empty map, with all industries co-located in one region
- it may cause long game start times as the map generator tries to place industries (low map gen times is an explicit FIRS goal)

[Eddi]

... or you could just make a parameter: optimal delivery distance for supplies: {short|medium|long|extreme}

[Wahazar]

... or you could just make a parameter: optimal delivery distance for supplies: {short|medium|long|extreme}

I don't think that such detailed parameters is really needed, in my opinion just on/off switch is enough: original or flattened curve.
If the newgrf can determine map size, the best would be off/automatic/on, with automatic as default.
The threshold can be if sum of MapSizeX+MapSizeX exceeds 2500
(1024x1024 is still playable with standard curve, except early vehicles, whereas I tested 2048x512 and found very difficult to use ships along map edge).

Currently I'm testing following modifiers for penalty_lowerbound and single_penalty_length:

Code: Select all

#define MODIFY_CDAYS1(increase, number)  number * (1 + 1 * increase)
#define MOFIFY_CDAYS2(increase, number)  number * (1 + 2 * increase)

and 20% of base price reduction.
(The exception are perishable goods, for which I'm using MODIFY_CDAYS1 for both penalty_lowerbound and single_penalty_length,
and bulk, for which MODIFY_CDAYS2 is applied for penalty_lowerbound).

Of course, such setting is not very friendly for local urban transport, but it is balanced by increased income for higher distances,
especially if you have complex cargodist network - I noticed 10% higher income from long distance passenger trains, local trains/buses are hardly to estimate because they are used for transfer - even if they have 20% less income from local cargo, they get higher amount from long distance transit.

There is still smaller income from coal load transported 2000 tiles by ship compared to train hauling same cargo 1000 tiles away,
which is bizarre, however such discrepancy is slighter and acceptable (ships does not need infrastructure).

[damerell]

Alternative approach: try and enforce a maximum distance between ENSP / FMSP supply sources, and industries that consume them. This is also tricky,

... also, it seems odd to address the "This coal is from China, it must be very valuable!" problem in OTTD with one particular kind of goods in one particular newgrf.

Likewise, I think the root of the problem here is the inability to account correctly for profit on feeder services. Fixing that in OTTD is a hard problem, but that's the place to fix it, not by fiddling with FIRS...

(That said, as noted above, if you're using UKRS(2) engines like the 9F or 0-8-0 to move goods around, you're likely to lose money; and there's a fine selection of mixed traffic engines that normally don't see much use.)

[andythenorth]

... also, it seems odd to address the "This coal is from China, it must be very valuable!" problem in OTTD with one particular kind of goods in one particular newgrf.

Well usually the problem players have is 'why do goods from further away earn more?' issue (it's a non-issue really).

But in this case the aim is to make specific cargos from further away earn even more, to support moving small amounts of them across the map with inefficient networks, or vehicles in grfs that are known to have high running costs.

What makes you think any of this is a bad idea?

[Eddi]

... or you could just make a parameter: optimal delivery distance for supplies: {short|medium|long|extreme}

I don't think that such detailed parameters is really needed, in my opinion just on/off switch is enough: original or flattened curve.
If the newgrf can determine map size, the best would be off/automatic/on, with automatic as default.

but that is not entirely accurate. the average distance you have to transport supplies not only depends on map size, but also on industry density (and number of different industries, and industry ratios), which may make "automatic" more complicated than it needs to be.

[aantono]

In FIRS, for Smithy Forge it says the cargo production is 8 crates for 8 tons of metal delivered. If I deliver 8 tons of metal, I don't seem to be getting 8 crates of ENSP (I have train configured to only load ENSP). Does the Forge automatically splits the delivered metal and only produces some ENSP, even though FMSP (farm supplies) are not being picked up by anything loading at the stations around?

[PikkaBird]

... or you could just make a parameter: optimal delivery distance for supplies: {short|medium|long|extreme}

NewGRF authors abdicating responsibility for gameplay is also a BAD FEATURE. I'm familiar with FIRS and have read this thread and I still have no idea what "optimal delivery distance for supplies" is supposed to mean, so how is a new player supposed to figure out how to set a NewGRF parameter straight out of the box?

We have this joke that the perfect train set would be one where every property of every vehicle is a parameter. Complete customisability which can suit any play style, sounds great, huh?

[andythenorth]

In FIRS, for Smithy Forge it says the cargo production is 8 crates for 8 tons of metal delivered. If I deliver 8 tons of metal, I don't seem to be getting 8 crates of ENSP

Where two cargos are produced, the industry splits output to each cargo 50:50. It makes no attempt to determine if one or both cargos are being collected.

[aantono]

In FIRS, for Smithy Forge it says the cargo production is 8 crates for 8 tons of metal delivered. If I deliver 8 tons of metal, I don't seem to be getting 8 crates of ENSP

Where two cargos are produced, the industry splits output to each cargo 50:50. It makes no attempt to determine if one or both cargos are being collected.

So if I were to intend to use the Forge just to produce ENSP, I should not expect 8 crates for 8 tons, but rather 4 crates for 8 tons. Is that right?

[andythenorth]

So if I were to intend to use the Forge just to produce ENSP, I should not expect 8 crates for 8 tons, but rather 4 crates for 8 tons. Is that right?

Yes, absolutely correct.

[Eddi]

so how is a new player supposed to figure out how to set a NewGRF parameter straight out of the box?

by a sensible description in the description field of the parameter. of course.

We have this joke that the perfect train set would be one where every property of every vehicle is a parameter. Complete customisability which can suit any play style, sounds great, huh?

and how is this generically named global tweak a "adjust every single property of every individual vehicle" switch?

surely there is a balance between no customizablility and full customizability. but it's not a death argument against adding any kind of parameter at all.

[PikkaBird]

surely there is a balance between no customizablility and full customizability. but it's not a death argument against adding any kind of parameter at all.

There is, but saying "this is too hard to balance, I'll just make it a parameter and let the player figure it out" rarely leads to good set design. Having adjustable scales of costs, production, capacities etc can also make it hard for authors of other, interacting, newgrfs to balance their set against yours.

[whatnames]

Hello!

New FIRS user here, got a dumb question if anyone could help. I'm trying to get some stone from a Dredging Site in the middle of a water are, but I don't seem to have any ships that can be refitted to transport Stone (or Clay). I checked for more ships or FIRS content but I couldn't tell what I should get. The rest of the networks are working ok, btw.

Here is a screenie of my NewGRF settings (I tried to include as much info as necessary on the image):



Thanks for any help on this! I'm really loving FIRS

[planetmaker]

You want FISH ship set or its successor SQUID ate FISH. Same author as FIRS, same awesome.