Factory game mechanics - Simu vs OTTD

[VS]

I just experienced it again. Simutrans cargo transport feels very static, compared to OpenTTD. I think I can finally tell why...

The fact that OpenTTD factories convert goods instantly and always do so as x=y (never a+b=c) means that individual raw material productions can change freely, without unbalancing anything. No destinations help this, too, but are presumably not a requirement. On the other hand, complicated cargo handling and interdependencies (as in pak128) seem to kill the dynamics completely. If I need 20% of X, 75% of Y and 5% of Z, increase in production of Z does not help anything.

It's such a classic example of "less is more" that it's actually funny. How to fix this? I have no concrete idea... But I hope it is evident that productivity increases need more complicated mechanism than just random +N% events. (I haven't really tried what happens with the current increase capabilities, though!)

[Sarlock]

Is the fact that factories have specific connections to other factories a game engine requirement?  Or a gameplay design choice?  I had always wondered why iron ore from the mine could only go to specific steel mills and not others - presumably iron ore should be deliverable to any factory that accepts that raw material as a feedstock (and has spare capacity).  Passengers clearly have a set destination in mind, but each piece of iron ore doesn't really care where it gets sent to...

Unspecific destinations for goods would make the act of balancing far easier as the aggregate demand for each good could be totalled and a factory build choice determined from that number (whether to build a new middle or end consumer or whether to build another raw material supplier).

I also think that the connectivity of factory chains is one of the biggest difficulties for new players to understand to get goods moving around their map.

[isidoro]

Well, the answer is simple.  It all depends on the model.

a) A factory that produces something that is to be accepted by any factory requiring it is more like a salesman traveling around with his merchandise, offering it to shops, industries, etc.

b) The simutrans model centers of transportation logistics.  You have a company that have already sold some merchandise to a factory, shop, etc. and that merchandise has to be delivered.

In fact, the model followed by ST is more transportation-logistics centered...  I certainly like it more than OTTD's.  The same applies to what VS comments.  It seems more realistic to me that to make bread you need flour, yeast and water in certain proportions.  You can't make bread out of water alone.  Besides, you can always build a pakset with only one -> one formulas.

[VS]

Destinations can be switched off globally for all industries, or not. It's not really a good idea, as contracts are a challenge on its own. I'm not saying that fixed proportions are bad, either.

What I want to say is that having everything static is just a bit weird, once you see that increases in production can offer new challenges.

I think it is not impossible or hard to get a somewhat dynamic economy even in Simutrans... just not done yet

[Ters]

I think it's realistic and preferable that factories need all inputs in certain ratios to make a product, and that the factories, not the transport companies, decide who buys from who. Some factories could have optional inputs in order to make a more luxurious version of the basic output, like cakes instead of bread for a bakery, but I'm not sure it would affect the gameplay significantly.

If something is too static compared to real life, it's the fact that factories stay true to their contracts forever. In some cases, they might in real life, but I think that for the most parts, contracts get renegotiated every few years. However, I suspect it would be very unpopular for some players if they had to constantly change their network because factories change suppliers and customers.

[sdog]

I think it is not impossible or hard to get a somewhat dynamic economy even in Simutrans... just not done yet

factories shutting down: implemented in experimental; not desired in simutrans as a lot of people want exactly that static world.

[prissi]

Factories in TTD will also produce more the better the station connected is served. You will find farmes supplying entire coutries if served well enough. Thus not only production is instantaneous but also supply is more or less limitless.

Interestingsly the most though after GRF sets (ECR and british industries addon) are not catering the "OpenTTD" default model. Especially the latter requires certain amounts and does not allow overstocking.

[Roads]

VS, if I am understanding you correctly, I'm afraid I have to disagree.  I don't find anything static about the transportation of goods.  In fact just the opposite is true as it requires a lot of micro management just to keep vehicles from getting stuck.  The thing it took me awhile to realize is that the amounts factories need of each raw product to produce the final product will vary over time.

An example of what WILL happen is you have a factory that needs 3 raw materials, each in various amounts.  You setup your routes and vehicles to carry the amounts needed and eventually get those "stabilized" so that the factory is producing just the amount you have trucks to pick up.  At this point don't get comfortable.  If you do it won't be long before you get the dreaded "vehicle XXX is stuck!"

[Ters]

It's been long since I had any non-deliberately stuck vehicles.

[el_slapper]

It's been long since I had any non-deliberately stuck vehicles.

Me 12 hours..... I guess building a complex network of trucks for a complex industry - in my case sand + coal = glass + grain = beer to supermarket - is a tricky thing. My main mistake is to "spare" by having one single depot for each place. Therefore, I did have a lot of problems at the Beer factory & the glass factory, with trucks not loading blocking the next one.

Next time I'll connect such a complex industry, I'll use trains. This will FORCE me to have a proper design, with proper flow management.

That being said, I've been punished for my own stupidity. Real life is complex, & just setting up a line of trucks hoping everything would go right was foolish from me.

Current dynamism(passenger-driven) is good, maybe not enough. That being said, I'm sure the TTD solution is worse. It dumbs down the game to "set up line, forget line, replace vehicles when blue".

[Ters]

When trucks are set to wait for load, I always have one loading bay for each vehicle with a choose sign in front, plus another loading bay (usually outside the choose sign) for trucks that don't wait for a load. I have also found that I need to take care inside the choose sign area, or city cars might ruin it by somehow getting stuck (unless that has been changed in the last year or less). With trains and airplanes I can be a bit more flexible as there are no city cars, while ships offer no challenge at all when it comes to harbor layout.

[Combuijs]

City cars still get stuck in such situations, one of the reasons I hardly play with city cars at all. I use more or less the same layout. City cars turning around (180 degrees) at a choose signal would be a nice solution...

[Ters]

In my experience, city cars only get stuck in loading bay areas if they arrive at an intersection and all exits, except their point of entry, lead to an occupied loading bay. If they arrive at an occupied loading bay by a straight or curved road, they will turn around. So I ensure that at all intersection either lead to another intersection or a straight/curved road, and haven't had a problem in maybe over a year.

[Combuijs]

Yes, I have noticed my case is different. I don't always work with choose signs.

I have one loading bay and a straight road of 3 tiles leading to it. The fourth tile is an intersection with the main road. I have four trucks loading from that bay. In the most extreme case there will be one truck in the loading bay waiting for goods and three trucks waiting on the three tiles before it. No problem there, cars on the main road can pass. However, when a city car nestles in between the trucks it won't turn around, it just occupies a tile. The fourth truck then blocks the main road which leads to a complete and thorough block. I know, I should use a private sign at the start of the road (but there is not one in pak64...) to prevent these kind of things, but that is too tiresome.

So I just play without city cars. They cause too many problems for only a small gain in game fun.

[Ters]

I seem to remember the AI building stops like that. I turned off the AI.

[prissi]

Actually, truck should not do this any more (blocking intersections) since some time.

[Combuijs]

Actually, truck should not do this any more (blocking intersections) since some time.

That is right, it now blocks the tile before the intersection! Which in my case is as bad as it was, but it might help in other cases...

[Ters]

Another thing that might be too static is the fact that factories don't evolve. Pak64 has old and modern style farms, mills and sawmills, but the old ones never go out of business. A windmill is still an integral part of some of my grain-based chains in 2020, and several others are parts of chains I don't serve.

[prissi]

Maybe in 2100 or so windmills will be again en vogue ... The problem with replacing with other factories is size and production. While the latter can change, the size is difficult; especially when there are lots of stations and streets around.

[Fabio]

As factories shutting down is deemed unwanted in Simutrans Standard (unlike Exp), we could at least implement updating them.
In dat files you can specify to which new factory it should upgrade, and Pakset maintainers have the responsibility to ensure upgraded factories have the same tiles and layout of the old ones, and same fields as well. They might have different stats and new products, though, provided they supply AT LEAST same products as the old one and have the same minimum production. In case these conditions are not matched for a Pakset error, the game simply won't replace them.

[Zeno]

While the latter can change, the size is difficult; especially when there are lots of stations and streets around.

I think there should be an easy solution for that. Old and new industry versions have same size = upgradeable; different size = no upgrade possible.

PS: Ooops, Fabio was faster! >:(

[Ters]

For farms and sawmills, the size is the same, so that could be a requirement, but it might not be ideal for windmill to modern mill. Simutrans already deals with growing buildings that might need to relocate (town halls), but having factories do the same will only be acceptable if other disruptive changes in industries are accepted. (Though town halls moving might be a quite disruptive to passenger networks today. They generate a lot of passengers and mail.)

[Fabio]

Well, in this case it could be set that if there is no room for upgrading to a bigger and newer factory, just don't upgrade it
In congested parts of the map it would lead to some heritage business not modernized, but heck, doesn't it happen also in real world?

[rsdworker]

in real world - some do expand

[Fabio]

in real world - some do expand

I explain better what I meant:
When a factory is retired, it becomes eligible to be upgraded.
At that point, if it has same size OR it can expand in the given place (e.g. bulldozing some houses or expanding in the empty tiles), it upgrades & expand. If expanding would require relocation, it doesn't upgrade. (And this behavior could even be enabled/disabled in simuconf.tab).

I also believe not all factories should upgrade in the same time, but only 1-3 (max) every season change (4 times a year).

[Sarlock]

There would have to be a balance between upgrading old factories and keeping some around for historical reasons...

I did a tour of a state of the art manufacturing facility some years ago and all of these high tech manufacturing machines were located in a 100+ year old building.  It was very impressive.  From the outside the building looked like an old textile mill that had been long forgotten.

I agree that if factories do upgrade, they should retain the same footprint.

[Fabio]

Factories (but maybe other objects as well) could have 3 dates (instead of 2): intro, retire and obsolete.
They are built between their intro and their retire dates, but they get replaced only after the obsolete date. Between retire and obsolete they won't be built but will not be replaced.
If you set a manufacturing plant to be retired in 1930 but obsolete in 1990, after 1930 only newer model plants will be built, but older won't be replaced until 1990. Obviously some factories might be retired but never obsolete.