Cargo & Industry fixes for 1.06

[RulerofRails]

Nice. This is useful.

How come we missed editing the Crystals production start date till now.

Actually I think that there isn't that much demand for Machinery. Sure there is a demand at a lot of buildings that take it, but the rate is super low. Machinery is hard to produce, so can only be a worthwhile endeavor if produced in bulk (if you focus on it). The rate is so low that little real consumption takes place, it doesn't rot fast either, so the typical thing is for a mass scale re-hauling around the map -- whether you intentionally take advantage of that or not.

This was where the idea came to add demand at Mines, Logging Camps etc.. And to get some production boost too. I had the concept for this worked out with Blackhawk. Akin to Fertilizer at farms, but a little more tricky thanks to these resources often running in lower demand environments (houses provide map wide background demand and therefore typically decent prices for farms) and Machinery is higher priced than Fertilizer. Our solution was to use two chains: one production and the other for the actual conversion. This lets real consumption take place + giving a boost effect. In super low price environments we could potentially see these resources take a little more loss (production/profit), but it seemed manageable to me.

A conceptual (it's in TM) use for Machinery was as a boost in a conversion industry. An example 3 Logs + 0.2 Machinery = 6 Lumber. However, there is a single value that caps all conversion entries (this is the problem with putting multiple conversions into a warehouse/port). Unless someone works out a super low resource map style that doesn't use much micro, my testing didn't show that this could be helpful on the typical map, and it's definitely not a boost. The "boost" possible (seen at the default farms) is achieved because it combines a Production and a Conversion entry.

On the topic of Machinery try this from a long time ago: http://hawkdawg.com/rrt/rrt3/hp/h-c.htm


PS.
Actually, I'm pretty sure that the express values in the bca files work. One evidence would be that Ned added express demands to a lot of buildings in TM. Production events work for express too (even by territory, proof in some of Oilcan's maps like Canyon Lands, or Zambu Email).

[Gumboots]

Nice. This is useful.

I'm thinking of adding some more to it, so we can see demands over time in a form that makes sense to the eye.

How come we missed editing the Crystals production start date till now.

Dunno. Maybe we're just a bit thick.

Actually I think that there isn't that much demand for Machinery. Sure there is a demand at a lot of buildings that take it, but the rate is super low. Machinery is hard to produce, so can only be a worthwhile endeavor if produced in bulk (if you focus on it). The rate is so low that little real consumption takes place, it doesn't rot fast either, so the typical thing is for a mass scale re-hauling around the map -- whether you intentionally take advantage of that or not.

Ok, but the rate of demand from each house is pretty low for a lot of cargoes, but the number of houses makes consumption pretty good. Goods is an example. Demanded at only 0.05 loads/year for small houses, and not demanded anywhere except houses, but it's usually a profitable cargo with a pretty stable price.

OTOH, Cheese is a cargo that can tend to crash easily. It's demanded at 0.02 loads/year for small houses, which are usually the most common ones IME. It also has a much faster rot time than Goods, which would tend to increase consumption and stabilise price, so the difference between a 0.02 demand and a 0.05 demand obviously has a dramatic effect on price stability. This sort of thing is worth thinking about when setting demands.

This was where the idea came to add demand at Mines, Logging Camps etc.. And to get some production boost too. I had the concept for this worked out with Blackhawk. Akin to Fertilizer at farms, but a little more tricky thanks to these resources often running in lower demand environments (houses provide map wide background demand and therefore typically decent prices for farms) and Machinery is higher priced than Fertilizer. Our solution was to use two chains: one production and the other for the actual conversion. This lets real consumption take place + giving a boost effect. In super low price environments we could potentially see these resources take a little more loss (production/profit), but it seemed manageable to me.

Yes, I can see that forcing higher priced cargoes into the production chain could adversely affect profits. Having some of them as demands only would allow more flexibility. And you'd want to be careful about which industries got a production boost, and by how much.

PS.
Actually, I'm pretty sure that the express values in the bca files work. One evidence would be that Ned added express demands to a lot of buildings in TM. Production events work for express too (even by territory, proof in some of Oilcan's maps like Canyon Lands, or Zambu Email).

Yes, I know production events work for express. My Latvia map has them. But the Barracks, for example, has no Troops mentioned in the .bca and the Hotel has no pax mentioned, so obviously some express production is locked into the .exe. This makes sense, because we know express prices are locked into the .exe too, and are not affected by events or by .bca edits.

[RulerofRails]

Um. Passengers, mail, and troops supply/demand have special values in the footer of all buildings.

From Pjay's notes

Code: Select all

102 (-067)	:     4 : float	: passengers : (commercial 0.5, house 1, hotel museum 2, church cinema deptstore 3, stadium 5)
106 (-063)      :     4	: float	: passengers : (commercial 0.5, house 1, hotel museum 2, church cinema deptstore 3, stadium 5)
110 (-059)      :     4	: float	: mail : (commercial 2, house 1)
114 (-055)      :     4	: float	: mail : (commercial 2, house 1)
118 (-051)	:     4	: float	: troops : (barracks 1)
122 (-047)	:     4	: float	: troops : (barracks 1)

I have never done a serious attempt to adjust these but I think the 1st is production and the 2nd demand.


I highlighted the values for the Hotel bca.

Hotel pax values.jpg

Is it funny that we might be surprised when something works as it should?

[Gumboots]

Ah, I was thinking they'd be in an insane location like, y'know, the production/supply chains section of the file. Silly me.

I wouldn't bother adjusting them for actual gameplay. AFAICT the balance there is one of the things they got right, although I suppose we should test an adjustment just so we can see how (or if) it works.

BTW, I'm thinking at the moment of splitting the Furnace into two industries as a sort of "private alpha 1.07" test thingy. I can then test play any 1.06 maps by editing them to suit. They should still open in 1.06 as long as cargoes stay the same and no industries are removed. IIRC we have 16 industry slots still available, so no problem with creating a few more. It's only removal that could cause instant CTD.

Which is why I'm thinking the cargo chain should be left alone. If we change that to ditch some 1.06 cargoes and substitute others, then I expect that any maps that call the defunct cargoes simply would not open in "1.07", and of course you wouldn't be able to edit them to "1.07" in 1.06 because it would lack the necessary cargoes. Catch 22.

There's also no room to add extra cargoes, so no chance of editing by creating a test installation with several extra ones (some of which would be redundant during play). Unless I'm missing something, you would have to directly edit the .gmp hex to update to "1.07", and that is a rather daunting prospect.

Edit: Come to think of it, we can always implement cargo changes by simply renaming things in RT3.lng and changing a couple of icons, but I'd prefer to avoid that since it means building in inconsistencies on the back end.

Edit again (brain farts rule): I suppose there would be a workaround, in that you could open a 1.06 map in 1.06, then edit out any of the changed cargoes, then save the thing before opening it in "1.07", and then editing the required cargoes back in. Sounds pretty tedious, and I'm not sure I'd be up for it.

[Gumboots]

Ok, updated the sheet with the correct express cargo supply and demand.

[Gumboots]

Hey question, for anyone who knows for sure.

In terms of industries and cargoes, what determines if a map will load without CTD?

I know 1.05 maps load in 1.06, since they don't have any conflicts with missing stuff (it's just missing). OTOH, changing an enabled industry will definitely make a map CTD.

What about industries that aren't enabled? Are they ignored, or not? For example, if you disable all 1.06 industries and cargoes, would a 1.06 map then load in a "1.07"?

I'm basically wondering how much we can change before existing 1.06 maps become impossible to work with, and what process would have to be gone through to update them to work with any new patch. If it involves hunting through .gmp hex then I think it's effectively a no-go.

[RulerofRails]

For some basic concept testing: edits in the RT3.lng would probably be useful. You can also test any single input conversion chain in a warehouse, although remember that the running costs per unit produced for warehouses are a little lower than for real industries (based on the build cost and warehouses are cheap).

Where are you thinking to go with this? Are you thinking about polishing up the 1.06 cargoes with balancing measures and smaller adjustments, or go full-scale with new/re-purposed cargoes etc.?

In terms of what will load, I don't know for sure. But I will speculate anyway. I just tried to take a 1.06 map: Africa & the Middle East, change the prefix, then load it in 1.05. It loaded fine. Even without disabling any buildings etc.. A port recipe set to demand Crystals before now demands Alcohol (first cargo in the list). So maybe there is some forgiveness in the system?

I know that a manually placed building can cause a missing file CTD if the game doesn't find files for it. That will definitely cause a crash.

[Gumboots]

Where are you thinking to go with this?

Round in circles, followed by disappearing up my own butt. It is, after all, RT3.

Are you thinking about polishing up the 1.06 cargoes with balancing measures and smaller adjustments, or go full-scale with new/re-purposed cargoes etc.?

I think there's general agreement that we don't much like 1.06 as it stands. The cargoes and industries weren't thought through properly. OTOH, I think that's its big selling point is the extra event logic that 1.06 has available. That's really good stuff, and is really the only thing I want 1.06 for. But there are some extras, like custom newspapers and in-game-buildable docks, which potentially add interest. Locos are slowly getting fixed, so are slowly becoming more usable. So the basics are there for something good, especially if we base things on an .exe which has had track counts and price islands and whatever else fixed.

The big issue is cargoes. We're tightly limited on those (IMO industries have adequate scope) and they have to cover a timeframe from 1830 to at least 2030, and everyone wants to argue endlessly about which ones we should have. That makes me think we should only fix stuff that is definitely screwed.

The less we change, the easier things will be. Balancing measures are fine. They shouldn't break anything, so we can implement as many of those as we like, but I'm inclined to keep cargo changes minimal. Every time we create something new, we're also creating a new testing cycle. So I think the best way to approach it is to start by listing "Stuff that is definitely screwed", and then ruthlessly go for the easiest and sanest solution to make it not screwed.

Taking the obvious one, I'd say we go ahead and split the Furnace into two industries. I'd be fine with an ore smelter being called "Furnace" for game purposes. A smelter is a furnace anyway, so there's no real need to bork anything by changing the name. If you want to use a 1.06 map, just tick off an extra industry (Cement Works, or whatever it turns out to be). Should be easy and trouble-free. Result should be two reliable production chains, which is what is wanted.

In terms of what will load, I don't know for sure. But I will speculate anyway. I just tried to take a 1.06 map: Africa & the Middle East, change the prefix, then load it in 1.05. It loaded fine. Even without disabling any buildings etc.. A port recipe set to demand Crystals before now demands Alcohol (first cargo in the list). So maybe there is some forgiveness in the system?

I know that a manually placed building can cause a missing file CTD if the game doesn't find files for it. That will definitely cause a crash.

Cool. That's handy info to start with. So perhaps the only "updating" issue is placed buildings. That would be an easy thing to update for. Although if ports are wanting Alcohol instead of Crystals you have to wonder what happens at the other end of the list.

Edit; Oh hang on a minute. If industries are being selected by the .gmp on the basis of their position in the list, that is going to overturn half the map if a couple more are added in odd spots. You'd still be able to correct it in the editor if you knew what to expect, but it'd be a fair bit more tedious.

I vaguely remember some discussion between Ned and Stoker about the industry. IIRC they had different opinions on how it works.

Found it: viewtopic.php?f=24&t=3641&p=36398#p36398

This makes me think that if we need to add an extra industry (like a Cement Works) then it may be best to deliberately give it a prefix that shunts it to the end of the industry list. That way it wouldn't upset the existing order, so updating existing maps should be far less trouble. A cement works could be classified as x_Cement_Works, or something like that.

Your PM the other week about RT3.lng edits to change loco names mentioned that it doesn't affect the event availability list but changes the name elsewhere. If the same applies to cargoes and industries, it may be possible to use a prefixed industry name in the editor list but have it show "normally" during gameplay.

[Gumboots]

Hey here's an idea (just to really screw things up). We have the extra 30 potential cargoes hidden inside 1.06:

Code: Select all

~4459Beer
~4464Candidates
~4466China
~4468Containers
~4471Detergents
~4472Deuterium
~4477Energy
~4478Fish
~4479Food
~4481Glass
~4482Gravel
~4495Money
~4497Newspaper
~4507Paint
~4508Perfume
~4509Potash
~4510Pottery
~4511Prisoners
~4515Salt
~4516Sand
~4518Spaceships
~4520Syrup
~4523Tea
~4525Tin
~4529Tobacco
~4530Tools
~4538Valuables
~4541Wine
~4542Wire

Now if I understand things correctly, the game won't handle having more than 52 cargoes enabled in any scenario. Personally I think this is enough, because once it gets over 45 cargoes in use the cargo display gets annoyingly cluttered anyway. Things start overlapping and going out of position, and it all becomes a bit of a mess. I prefer to not even use all of 1.06's 51 cargoes at the same time, and think that if you can't design a scenario with 40 to 45 cargoes as a maximum then you should probably rethink things.

Still, the game engine handles having 51 or (maybe) 52 cargoes enabled. So, what does "enabled" actually mean? Does it mean "ticked on in the industry list, and assigned to one or more cities/regions"? Or does it mean "has a .cty sitting in CargoTypes and is called by one or more buildings in BuildingTypes"?

If it's the former, then we could have the whole lot set up with .cty and .bca files if we wanted to, providing nobody ever tried to use them all at once (and you can bet some kid will try it). The only catch would be that any .bca that referenced a cargo would be calling if that building was enabled anywhere, so you'd have to give some thought to production chains so that enough stuff could really be disabled.

So, how does it work? If you have (hypothetically) 80 cargo .cty files alive and kicking in CargoTypes, and 80 cargoes called by various buildings in BuildingTypes, would you still be able to construct a viable working scenario by only selectively enabling the relevant industries in whatever cities and regions (keeping the total under 52), or would that all turn to crap?

---

Edit: Have just been doing a lot of reading up. First thing is it looks like there is a hard-coded limit of 60 cargoes:

Currently, GMP files store event condition types using a four-byte integer. "YTD company <cargo> hauled" conditions, for example, occupy the codes from 000000FDh through 00000139h. This allocation scheme is what limits the game to 60 different cargoes, of which 41 are currently defined.

Presumably, if you had 80 .cty files sitting around in CargoTypes, the last 20 of them would simply drop off the end of the list and you'd never see them in the editor. Which is handy to know before getting carried away.

It's also rather handy in that it ties in well with the CargoIcons_*.dds that does the icons on the sides of cargo cars. Back in that same thread I see that, at the time, people didn't know how these were assigned. That explains why 1.06 cargoes never called them properly, but I figured it out when I made my bugfix patch for them.

Anyway, CargoIcons_*.dds has 64 slots available on it, and we can easily use them all if we really want to. 52 of them are taken up by 1.06, but that includes four slots that are taken by Mail, Passengers, Any Express and Troops. These cargoes don't actually use cargo icons in the game, because the cars are so distinctive anyway so don't need icons, but with those four icons there on the .dds there are 60 slots vacant. This just happens to tie in nicely with the 60 available slots for cargoes in the .gmp coding, so that's all good to go.

What appears to define the 52 cargo limit in practice is this:

Something to watch out for: cargo limit may be lower than the 60 that the event system suggested. The structure representing the economic record for each square of the map is 471 bytes long; there's an array of floating-point numbers representing cargo quantities available 259 bytes into it. If each floating-point entry is 4 bytes, you can see there might be a bit of trouble past the 52nd cargo or so.

So there's the real limit: the game won't allow space for more than 52 cargoes when tallying up activity on any economic cell. That's a hard limit that you can't get around. However, it implies (would need to test) that it would be possible to have up to 60 cargoes defined in CargoTypes, providing that no more than 52 of them were enabled at any one time. IOW, at least 8 of them would need to be unchecked in the industry list, and not assigned to any regions or cities(ie: slider percentage set to zero everywhere).

[RulerofRails]

Try and see. Here's some info on the "52 limit": viewtopic.php?p=2256#p2256.

I don't know how a cargo is "enabled" behind the scenes, but If you didn't notice in the Western Geatland thread I recently mentioned finding Milo's info on what makes a cargo available in an actual game, in the context of placed vs. seeded ports/warehouses, but there is probably all the info you need there: viewtopic.php?p=3210#p3210

The majority case is that a cargo is made available when whatever industry produces it is enabled in the list.

When I said the Port was demanding Alcohol when it used to demand Crystals, all that was happening is that the 1.05 game defaulted to Alcohol (the 1st item) when it couldn't find Crystals. I wouldn't assume this is indication of a numbering problem.

PS. Don't really understand it, but I found a quote about the flatbed cars here by Milo. Not sure if it's anything?

The mystery byte at offset 20h is a 'visual appearance' modifier to the car. 1-8 apply to Flatbed cars, while 9-11 are Open_Hopper cars. This is how the Weapons flatbed shows up, for example. The default for cargos without a 'visual appearance' modifier is just to paint the resource icon on the outside of the railcar.