News:

Simutrans Sites
Know our official sites. Find tools and resources for Simutrans.

Thoughts on shunting, adding&removing cars, etc...

Started by Fabio, November 09, 2012, 05:42:22 PM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

Fabio

These features have been discussed extensively in the past, but I can't find where.

I want to quote this interesting contribution:

Quote from: Roads on November 09, 2012, 04:37:16 PM
One thing I wanted to say and forgot about Railroad Tycoon.  It bothers me having a train that is loaded only one way.  Maybe it is because for the better part of my life some of my family and many of their friends were truckers.  Their one big complaint was not diesel prices or being away from home, sleep deprivation, etc., it was having to dead head home for the weekend!  That was by far their biggest expense on a weekly basis.  I can only think trains would be similar.  Railroad Tycoon (only the original) had switching stations that allowed you to change cars, add cars, remove cars,  wait forloads, etc.  You could create some really complex operations.  It is unrealistic to think every little factory would have a switching station with an unlimited number and variety of cars though so some compromise would be needed there.  Perhaps only switching stations in the larger cities.  This should allow for only short dead head trips and actually even an added element - moving empty cars from one location to another.

Being unrealistic is not the number one reason I dislike having trains, trucks, boats being loaded only one way.  It is simply not interesting having a train
pick up coal at a mine and delivering it to a power plant.  There has to be some complexity to a route to make it interesting.




If we find older threads, we can merge them with this one.

Roads

This from Kieron in another thread:

QuoteThe lack of management required with waggons is also one of the aspects I dislike about Railroad Tycoon. I can understand the reasoning behind this simplified gameplay - with there being more to do in the wider economy in RRT having to design detailed track layouts and manage waggons could overload the player. However I prefer the aspects Simutrans focuses on.

No doubt many more people than Kieron do not like this because it inevitably produces situations so complex that at times you want to just throw your hands up and quit.  For that reason if this is implemented it certainly needs to be an option with the default off.  Here is the thing.  Difficulty and complexity is what IMHO, keeps a game alive.  At one time Civilization was the top selling game.  At the time, devs said they could make it much more complex with trade, etc., but that it would overwhelm many players.  So what happened?  They kept it simple and eventually people got bored and looked for other games.  A very similar thing happened with Everquest.  The original game was hideously difficult.  People would wait at the entrance to Sebilis for hours just to get a "camp."  I have done so myself.  Then EQ2 was released.  Corpse retrieval was removed, everything was made easier and the game faded into obscurity.

I don't believe people lose interest in games because of difficulty.  They lose interest when the get bored.  And again IMHO, a game is like most things in life.  It either evolves or it dies.

Ters

Is it really realistic for anything but passenger, mail and box/container cars to run loaded both ways? In any case, the same cars must go both way, unlike in the Railroad Tycoons I've played.

While it would satisfy the suppressed model railroad builder in me to have locomotives take turns at different consists (or whatever the right English word is) and to have the actual cars switched between trains at freight hubs rather than having the goods reloaded, such things would not work in Simutrans today. It would require freight hubs of several hundred tiles in size, and more than ten such in a map, each many hundreds, even thousands, tiles apart. Such big maps won't work. This is the hypothetical, supercomputer powered Simutrans 3000.

ӔO

you would also need a way to run trains both ways on a shunting line without it jamming.
My Sketchup open project sources
various projects rolled up: http://dl.dropbox.com/u/17111233/Roll_up.rar

Colour safe chart:

greenling

Those idea it Super.
I Like it.
I live on a Railwayroute there drives trains The parts and connect.
Than must i not call train in a depot to exchange wagons.
Edit That idea it s for Simutrans exp also useful.
Opening hours 20:00 - 23:00
(In Night from friday on saturday and saturday on sunday it possibly that i be keep longer in Forum.)
I am The Assistant from Pakfilearcheologist!
Working on a big Problem!

ӔO

if anyone has ever played train simulator 2012... some of those rail yards are not possible under simutrans.
My Sketchup open project sources
various projects rolled up: http://dl.dropbox.com/u/17111233/Roll_up.rar

Colour safe chart:

Vladki

Quote from: Ters on November 09, 2012, 07:47:59 PM
Is it really realistic for anything but passenger, mail and box/container cars to run loaded both ways? In any case, the same cars must go both way, unlike in the Railroad Tycoons I've played.

Ters is right. The cars have to return somehow to be loaded again. And if they are used for special cargo - wood, steel, cement, milk, then they have to return empty. Only passenger, mail, piece and maybe bulk goods have a chance to run loaded in both directions. And IMHO real cargo trains return empty even for bulk goods. Think of a big coal trains for power plants.

When I played Railroad Tycoon, I wondered, how the cars magically disappear when unloaded, and different new cars appear for return. I think Simutrans and OpenTTD does it the right way.

VS

Reading the last few posts, I think an observation could be made: If the law of mass preservation holds true in the game, it is easiest to have no switching.

One could imagine that cars return empty in different trains, but where its the advantage of that? Price? And you must return as many cars as are coming. Perhaps moving 20 empty cars at half the speed of loaded 10 cars is more economical, taking advantage of some non-linearity. But then you have imbalance in engine counts! :D Assuming you could instruct two trains to merge with only one engine running at 50% speed, it would work out, but this is sooo convoluted... Balancing such a thing would be a hell, both for authors and players.

My projects... Tools for messing with Simutrans graphics. Graphic archive - templates and some other stuff for painters. Development logs for most recent information on what is going on. And of course pak128!

ӔO

It's not unusual seeing empty wagons being moved around to their next pickup destination.

I think the biggest hurdle would be making such a system user friendly. You really must have a good understanding of how freight rail yards operate to use this system. Seeing as setting up a rail is a bit complicated already, I think it wouldn't help the learning curve either.


but if someone were to code it... I do have some ideas on how it can be done.
My Sketchup open project sources
various projects rolled up: http://dl.dropbox.com/u/17111233/Roll_up.rar

Colour safe chart:

Fabio

There could be easier scenarios and simpler implementations of switching.
Think of this case:
A--B--C
where A and B are suppliers and X is the consumer of both.
You could start the line at A with engine + 5 cars. At B the train couples with another 5 cars and goes to unload at C, then back to B where it uncouples 5 cars and finally back to A.
This layout would give savings compared to a 10 cars train all along and would also cost less and clutter less the lines list than 2 distinct AC and BC lines of 5 cats each.

Vladki

Nice example Fabio. I'd like to have such situation possible. Now I usually do it so that there are two trains: A-B (5 cars) and B-C (10 cars), and cargo is transferred at station B. But the drawback is I have to buy two engines and 5 extra cars, and one more platform. And running with full length train which is not fully loaded generates a financial loss.

Fabio

Additionally, I hate to have too many lines (or any scheduleless train).
Adding/removing cars could also be used to have RoRo ferries and train shuttles (they could pick up the transported trucks hiding them), or helper engines for mountainous or unelectrified sections.

Roads

Some of the thoughts here seem to be way over thinking what needs to be done.  In my games now when I'm into micro management, I have switching stations.  I use the depot.  When a train unloads, I have it go to the depot.  When I get the message that the train has entered the depot, I change the consist (yes Ters, that's the right word) to whatever I want it to pick up next, change the route to one I've previously set up, click start and the train is off again.  Rinse, repeat.

Fabio

My Gosh! Do you REALLY do that? You have a boatload of patience... I could endure it not even once!
I totally abhor micromanagement, I'm rather the build & forget guy.

sdog

@ӔO  fabio's suggestion wouldn't need to actually simulate the shunting process, but the effect.
e.g. an entity where consists A and B drives into and consist X and Y leave where all parts of X, x_i ∈ A ∪ B were in either A or B.

this could be extended to a yard storing vehicles, which would be something like a scriptable depot.


There are nice examples where railroad tycoon like behaviour was used in real life. Barges transporting goods downstream often were scrapped and the timber sold instead of pulling them up-river. Typically the higher one gets on a stream the cheaper lumber and workforce is too.



Quote from: Fabio on November 10, 2012, 12:28:39 AM
My Gosh! Do you REALLY do that? You have a boatload of patience... I could endure it not even once!
I totally abhor micromanagement, I'm rather the build & forget guy.

Still it was you who suggested shunting ... :-)

ӔO

@sdog, ah gotcha. Thanks for clarifying.

I'll try and draw up my idea for such a thing, but basically it involves numbered 'units' of train wagons that locomotives can pickup and drop off.
My Sketchup open project sources
various projects rolled up: http://dl.dropbox.com/u/17111233/Roll_up.rar

Colour safe chart:

Fabio

Quote from: sdog on November 10, 2012, 12:30:49 AM
Still it was you who suggested shunting ... :-)

Indeed I suggested it, but as a scripting or whatever so that once I got it right I can trust it's working for years.
I can spend quite some time building a line and adjusting it, but then I want it to work without me needing to take care of it every moment...

Just imagine: I usually freeplay, hence when I need to update a vehicle, I prefer to withdraw it (or even sell it on the place) and buy immediately a new one rather than waiting for it to reach the depot and so on...

Roads

Yes Fabio, I really do that occasionally.  The thing is I hardly ever apply the same method to every situation.  If a situation develops that is interesting to micro manage then I'll do it.

greenling

#18
I have view a video there the Lokomtion change for Electric on diesel.
http://www.youtube.com/watch?v=sJgoDfTHW9w
Edit: I Have those simutrans spezial exe that can.
Only i must me Remember how i those Simutrans Spezial exe have be park.
Opening hours 20:00 - 23:00
(In Night from friday on saturday and saturday on sunday it possibly that i be keep longer in Forum.)
I am The Assistant from Pakfilearcheologist!
Working on a big Problem!

sdog

the music at 0:40 ... sounds like Kanno ... good memories ...

thanks for the video greenling.

Ters

The Japanese must either wield some secret magic, or the video is faked. There are several things in that video that's not in any Simutrans I've seen here in the Western World.

kierongreen

There is an experimental Japanese version of simutrans (different from the one on this forum). I don't keep up to date with the features so maybe it is real?

greenling

Opening hours 20:00 - 23:00
(In Night from friday on saturday and saturday on sunday it possibly that i be keep longer in Forum.)
I am The Assistant from Pakfilearcheologist!
Working on a big Problem!

Sarlock

IMHO shunting, etc, wouldn't add much to the game... a bit of realism perhaps, the economic advantage would be potentially offset by slightly higher operating costs since in the end most paksets want a fairly standard level of return for a well run operation.

A lot of freight lines have to return empty cars to their origin, basically anything specialized (chemicals, grain, coal/minerals, oil/fuel, etc).  Being able to run both ways fully loaded is a rare event, even being half full on the return trip is a nice bonus usually.  Most areas that produce raw materials do not consume much on the flip side... trains generally run one way loaded and empty back.  Container cars are really the only possibility for freight that can run both ways semi-loaded.  But even then these cars are generally half full on part of the journey as there is always a mismatch between industrial goods consumption and production for any given delivery area.  And these types of cars are generally delivered to intermodal yards not end consumers.

The same occurs in long haul trucking, as Roads alluded to.  It's a real blessing to get return freight on a long haul journey... you're usually making at least part of the journey back empty.
Current projects: Pak128 Trees, blender graphics


Roads

Sarlock, what you are saying is true if you think about freight only in terms of one destination and return.  For switching stations or trucking for that matter to work and be worthwhile and interesting, you have to think in terms of multiple destinations.  Instead of having say, five short routes, you have one long route and along that route you have other trains that intersect (for lack of a better word) which pick up part of the goods and carry them to different destinations, etc., etc.

A well thought out and designed system could easily replace 50 short routes with 7 to 10 long ones.  And I can tell you, the reward for developing such a system is enormous to see it in action.  Because you actually created something rather than just clicking on A to B and hitting start.

Fabio

Quote from: Roads on November 10, 2012, 10:46:03 AM
A well thought out and designed system could easily replace 50 short routes with 7 to 10 long ones.  And I can tell you, the reward for developing such a system is enormous to see it in action.  Because you actually created something rather than just clicking on A to B and hitting start.

This is something I completely agree with and I want to underline. :thumbsup:

Ters

Quote from: Roads on November 10, 2012, 10:46:03 AM
Instead of having say, five short routes, you have one long route and along that route you have other trains that intersect (for lack of a better word) which pick up part of the goods and carry them to different destinations, etc., etc.

They've mostly stopped doing that over here, apparently because it's too costly. Almost everything that isn't intermodal is point-to-point. I know that at least one lumber train picks up more wagons along the way (but only at a single place, doubling it's length and with a change of locomotive). And a car importer probably shunts car wagons comming across the border into different trains for distribution to the major cities.

wlindley

Drive across Interstate 10 the nearly 4,000 km from Los Angeles to Jacksonville, Florida, and you will see on the parallel Union Pacific Railroad, about half the trains are "General Merchandise" freight trains.  These trains are often a mile long, and composed of every type of railcar imaginable -- intermodal container, tank containers, piggyback trailer-on-flatcar, covered hoppers full of grain or dog-food, open hoppers with coal or sand, gondolas with scrap, flatcars with bulldozers, boxcars with almost anything, and more.  Each of those trains represents hundreds of hours of switching (shunting) and the trains are often broken apart in sections, and new sections added along the way.  It is a process similar to how your blood flows from the heart through the arteries and splits into millions of capillaries, then reassembles back into veins back to the heart... a journey of infinite mixing and complexity ... if we could simulate even a bit of this, it would be quite fun.

prissi

That train get less shunted can be also seen that there were thoughts about EMU/DMU for freights, liek the japane JRF200 or the german Cargotrain (which are in pak64 and pak64.japan) But both were not too sucessful. One of the reason was low motor power for the german, and the fact that there were so many left over slow and powerful engines from german reunification and smaller lines going out of buisness.

Ters

Though I wrote that trains a mostly point-to-point today here, and perhaps in Western Europe in general, that is something that has changed over time, and Simutrans spans both time and space. Still, I strongly suspect that it would require a massive redesign of Simutrans' vehicle handling and routing to get this working properly. That the Japanese can change the locomotive might not be of too much help. I know that they used to do that on border crossings between Norway and Sweden before, and they might still do that down on the continent, especially where the electrification is different. In that way, it could have been fun for multiplayer games, especially if zones of operation was implemented. A player 1 could hire another player (player 2) to pull player 1's consist through player 2's zone using player 2's locomotive. Otherwise, it would just be eye candy that complicates the game both on the inside and outside.

Markohs

#31
Passenger trains were often composed in Spain.

There was a train I took from Barcelona here to Andalusia, you had to be extra careful to pick the seat you paid for in your bill, because when it reached Alcazar de San Juan, near andalusia, it began splitting in branches, the electric head was replaced by diesel ones and the waggons into different convios, one going direction to Jaen/Granada/Malaga and the other headed to south-west, Seville and Cadiz, more or less, I can't recall the details. On the other direction, convois were merged incrementally and it all ran together to Barcelona. That was sometimes a problem because if one branch was late, the whole train had to wait for the missing part. But it used to work quite smooth.

I think they don't use to do this in Spain any longer for passenger trains.

EDIT: Info about this train (now replaced by other services, coudn't find nothing in english)

http://www.ferropedia.es/wiki/Arco_Garc%C3%ADa_Lorca

Roads


QuoteIt is a process similar to how your blood flows from the heart through the arteries and splits into millions of capillaries, then reassembles back into veins back to the heart... a journey of infinite mixing and complexity ... if we could simulate even a bit of this, it would be quite fun.

Yes, Wlindley, that is exactly what I'm talking about.  What would make this even more awesome is that in Simutrans you can already interface with trucks and boats for pick ups and drop offs.


This is undoubtedly vastly over simplifying but as far as what needs to be done, all you really need to do is record the keystrokes in a depot, give it a name and then add it to your route.

Ters

Quote from: Roads on November 10, 2012, 05:45:02 PM
This is undoubtedly vastly over simplifying but as far as what needs to be done, all you really need to do is record the keystrokes in a depot, give it a name and then add it to your route.

Having this happen in depots would ruin the fun in it for me. I want to see it. The trains would also need some synchronizing mechanism. That is perhaps the biggest hurdle when it comes to more complex train behaviour: the lack of a fixed predictable schedule.

Fabio

But as there is already wait for load option, there could "easily" be a wait for cars to be added one.
This would require additional platforms or platform-like shunting yards, but then trains could wait indefinitely.
I believe there is a fair point in between overly complex realism and the lack of this feature at all.
Some basic shunting (and roll-on-roll-off ferrying) should IMHO be simulated in a transport simulator, we need to cherrypick which aspects could be actually implementable.