Page 117 of 154
Re: Cargo Distribution
Posted: 02 Jan 2012 17:55
by Wasila
I've been doing some work on improving the network, and have set up several new road and rail routes to improve flow. However, my attempts to increase capacity on certain lines has hit a snag - although I have auto replace set up on several lines I simply don't have the money for the new trains and carriages I need. Profits did improve for a while, but are back down, and I believe that turning infra maintenance off is what has saved me so far.
The new express to Birmingham does seem to have had the desired effect, although again due to lack of trains it is deeply overloaded.
Most often I start the game and try to connect two or three reasonably sized cities by train, then add bus/tram services to drastically increase transported passengers, then continue to connect more and more cities in a similar way.
Looking at my network, aren't most cities on the important lines surely too far away from each otherto be connected by RV? That is, after all, why they are connected by train. How exactly would this apply on my network?
That requires a sophisticated railway system though which can handle trains overtaking each other and isn't trivial to get fool proof

I do use express services, although as you say they can't really overtake except at stations.
Edit: I can't load your game here at the moment, but I might be able to take a look a bit later. How many passengers do you have waiting at the bigger stops? And what are your station ratings like? Any station drop below 50? If the station rating drops, you'll get less passengers (or goods in general) to transport, so keeping it up is rather important...
Waterloo has the most, at 7000 passengers (the station in the south of London). Most want to travel to the next London terminus (by tram - I have been laying this on thick to attempt to deal with this). Again taking Maidstone as an example, there are 4,800 passengers, most of whom still want to take the train towards London. My highest cargo rating is mediocre at 40%

This shows you the extent of my problem, although I don't think getting less passengers to transport is the real problem here...
Newer save attached.
Much appreciated,
Wasila
Re: Cargo Distribution
Posted: 02 Jan 2012 18:36
by JGR
Wasila wrote:Waterloo has the most, at 7000 passengers (the station in the south of London). Most want to travel to the next London terminus (by tram - I have been laying this on thick to attempt to deal with this). Again taking Maidstone as an example, there are 4,800 passengers, most of whom still want to take the train towards London. My highest cargo rating is mediocre at 40%

This shows you the extent of my problem, although I don't think getting less passengers to transport is the real problem here...
Newer save attached.
Much appreciated,
Wasila
I've compiled the latest cargodist from git, and got your game to load.
I don't seem to have the right version of the BR set, so trains are broken, but otherwise I can see what is going on.
Profits are mainly down as you've been buying trains, it will soon go back up as new connections get added to the mesh.
In general, single-track lines are best avoided.
Especially on long and busy routes like London-Birmingham.
I'd also suggest using longer/more capacious stations and trains. Carriages are cheaper than locomotives, and where the money actually comes from at the end of the day.
I would focus less on geographic expansion right now, and try to improve the network efficiency/profitability.
As an aside, your stations' catchment covers very little of the towns they serve. Though frankly there's no truly elegant way to deal with that.
Re: Cargo Distribution
Posted: 02 Jan 2012 19:07
by Wasila
Profits are mainly down as you've been buying trains, it will soon go back up as new connections get added to the mesh.
I've been looking at Operating Profit, according to the graph, not the final number in the finances, and operating profit fell steadily until the last three quarters, in 1960, registered barely any profit at all, and despite reducing town growth a lot of routes seem to be getting worse. The increased losses may be because the trains between Birmingham and London have been losing so much, despite being great profit makers previously - the reason I didn't introduce express trains before was because the journeys take so long (over a year?), and this may be a cause.
In general, single-track lines are best avoided.
Especially on long and busy routes like London-Birmingham.
NuTracks and infra maintenance had forced me to avoid shelling out too much on track, and I carried this over to the mainlines, perhaps mistakenly. I have spent some money on widening the track, but I haven't as of yet seen any effect.
I'd also suggest using longer/more capacious stations and trains. Carriages are cheaper than locomotives, and where the money actually comes from at the end of the day.
I'll start focusing on this, since autorenew isn't working. We'll see how it helps...
As an aside, your stations' catchment covers very little of the towns they serve. Though frankly there's no truly elegant way to deal with that.
Are we speaking of cities or smaller towns? How does one improve catchment in smaller towns without demolishing right through them?
Re: Cargo Distribution
Posted: 02 Jan 2012 22:30
by JGR
Wasila wrote:I've been looking at Operating Profit, according to the graph, not the final number in the finances, and operating profit fell steadily until the last three quarters, in 1960, registered barely any profit at all, and despite reducing town growth a lot of routes seem to be getting worse. The increased losses may be because the trains between Birmingham and London have been losing so much, despite being great profit makers previously - the reason I didn't introduce express trains before was because the journeys take so long (over a year?), and this may be a cause.
NuTracks and infra maintenance had forced me to avoid shelling out too much on track, and I carried this over to the mainlines, perhaps mistakenly. I have spent some money on widening the track, but I haven't as of yet seen any effect.
I'll start focusing on this, since autorenew isn't working. We'll see how it helps...
Are we speaking of cities or smaller towns? How does one improve catchment in smaller towns without demolishing right through them?
Express trains should not lose money after the first return trip or two. Have you checked how much cargo they're carrying?
Part of the rationale for dual track is so that journeys do not take so long.
Reducing town growth will not do anything, most of the towns there are already very large.
Both really, but cities more obviously. The usual strategy is by using bus feeders, by abusing the distant-join feature or by using metros/local trains.
I would not worry about that now though.
Re: Cargo Distribution
Posted: 03 Jan 2012 13:22
by Creat
JGR wrote:As an aside, your stations' catchment covers very little of the towns they serve. Though frankly there's no truly elegant way to deal with that.
With cargodist there is an excellent way to deal with this: use buses. Have a bus stop connected to the station and fill the town up with stops until most of it is covered. Then run a single bus (for very small towns) or two (with one going clockwise, the other anti-clockwise through the stops). This will bring the passengers the town generates to the station and distribute those arriving there to the right houses (or whatever their destination may be).
If you don't mind the unrealistic nature you can also (for smaller towns) use bus stops themselves and all connect them to the station, their catchment area is then basically added to it. I personally don't like this as it basically abuses a game mechanic, but you might feel differently.
I still didn't have a chance to look at the savegame, but if the problem you have is because most passengers get stuck at bottleneck connections at some point of their journey, introducing even more will just make it worse.
JGR wrote:Express trains should not lose money after the first return trip or two. Have you checked how much cargo they're carrying?
Part of the rationale for dual track is so that journeys do not take so long.
Reducing town growth will not do anything, most of the towns there are already very large.
Both really, but cities more obviously. The usual strategy is by using bus feeders, by abusing the distant-join feature or by using metros/local trains.
I would not worry about that now though.
Keep in mind that they might run empty on the very first trip, as any connection is initialized the moment the train stops in the station, therefore there can't be any passengers for this train just yet. I also remember there was some discussion about routes looking at orders instead of just where trains at a station want to go, but don't know if that ever changed this behavior. Also if I remember correctly it takes a very LONG time for passengers which are stuck at some station to pick a different (new) route. So mainly new passengers will take that express. I might also be wrong here, I haven't been playing much lately and might be a bit out of touch with the mechanics there

Re: Cargo Distribution
Posted: 03 Jan 2012 17:16
by Eddi
are you aware that the compilefarm fails on linux and mac builds? see
http://farm.openttd.org/browse/OTTD3PT-CD/latest for details.
Re: Cargo Distribution
Posted: 03 Jan 2012 18:33
by fonso
Thanks for the hint. I should probably restructure my repository. It's not the best idea to have the compile farm immediately build anything I write. There should be a "stable" and a "dev" branch ...
Re: Cargo Distribution
Posted: 03 Jan 2012 21:21
by Wasila
With cargodist there is an excellent way to deal with this: use buses. Have a bus stop connected to the station and fill the town up with stops until most of it is covered. Then run a single bus (for very small towns) or two (with one going clockwise, the other anti-clockwise through the stops). This will bring the passengers the town generates to the station and distribute those arriving there to the right houses (or whatever their destination may be).
If you don't mind the unrealistic nature you can also (for smaller towns) use bus stops themselves and all connect them to the station, their catchment area is then basically added to it. I personally don't like this as it basically abuses a game mechanic, but you might feel differently.
Interesting technique - I have only used such a system with the largest of cities before.
Keep in mind that they might run empty on the very first trip, as any connection is initialized the moment the train stops in the station, therefore there can't be any passengers for this train just yet. I also remember there was some discussion about routes looking at orders instead of just where trains at a station want to go, but don't know if that ever changed this behavior. Also if I remember correctly it takes a very LONG time for passengers which are stuck at some station to pick a different (new) route. So mainly new passengers will take that express. I might also be wrong here, I haven't been playing much lately and might be a bit out of touch with the mechanics there

They have been making money now, and the network as a whole seems to be back in profit territory - I have been turning a nice little sum over in recent months. Thanks for your time and patience so far (and to JGR as well); hopefully I'll be able to build more intelligent networks in the future. I may still have difficulty in this particular map, though, as despite a decade of investment in the network it still seems to be struggling to move all the passengers...
Re: Cargo Distribution
Posted: 04 Jan 2012 17:57
by JGR
Creat wrote:With cargodist there is an excellent way to deal with this: use buses. Have a bus stop connected to the station and fill the town up with stops until most of it is covered. Then run a single bus (for very small towns) or two (with one going clockwise, the other anti-clockwise through the stops). This will bring the passengers the town generates to the station and distribute those arriving there to the right houses (or whatever their destination may be).
If you don't mind the unrealistic nature you can also (for smaller towns) use bus stops themselves and all connect them to the station, their catchment area is then basically added to it. I personally don't like this as it basically abuses a game mechanic, but you might feel differently.
The issue is that for any reasonably sized town, you typically need a whole fleet of buses to even try to meet demand, and that tends to result in major congestion problems at the bus stops.
Generally my strategy is to use distant-join to limit the stations needed to cover most of the town. Anything more than 6-8 or so tends to become unmanageable. That and extending the catchment area of the main station(s) more than the feeder stations.
Re: Cargo Distribution
Posted: 04 Jan 2012 18:36
by Creat
JGR wrote:Creat wrote:With cargodist there is an excellent way to deal with this: use buses. Have a bus stop connected to the station and fill the town up with stops until most of it is covered. Then run a single bus (for very small towns) or two (with one going clockwise, the other anti-clockwise through the stops). This will bring the passengers the town generates to the station and distribute those arriving there to the right houses (or whatever their destination may be).
If you don't mind the unrealistic nature you can also (for smaller towns) use bus stops themselves and all connect them to the station, their catchment area is then basically added to it. I personally don't like this as it basically abuses a game mechanic, but you might feel differently.
The issue is that for any reasonably sized town, you typically need a whole fleet of buses to even try to meet demand, and that tends to result in major congestion problems at the bus stops.
Generally my strategy is to use distant-join to limit the stations needed to cover most of the town. Anything more than 6-8 or so tends to become unmanageable. That and extending the catchment area of the main station(s) more than the feeder stations.
The main problem is, that to this day there isn't any usable way with OpenTTD to keep vehicles separated. After a while they will ALWAYS drive around on one route as a single lump. Which comes from the simple fact, that the second bus usually has less passengers to load and departs more quickly, even more so for the third and so on. You might mention timetables, but that becomes a headache too as you not only need to initialize them and the adjust them slightly to allow for vehicles to be able to catch up after a break down, but adding stops to a route is even worse. Also you need to manually space them out (for example by sending them out of the depot with the correct distances). The newest attempt seems to be
this solution, I'm trying it out over the next few days I guess and can then report how well it works. Seems simple enough, which is promising. The
last approach was confused by conditional maintenance (depot) orders and stopped working with those altogether, which rendered the entire patch useless to me. This has an even worst effect with medium or long distance trains, as later ones tend to run mostly empty and consequently the station rating drops drops drastically.
But enough OT (and ranting), back to the issue at hand... I usually use one of two systems: either circular bus routes which are served in both directions (so at least 2, usually 4-8 buses) or trams with a similar layout and the added possibility to include neighboring (small-ish) towns with the same network. The later approach tends to be more complex as the routes are not just circular (though usually they are the basis of the whole thing) but will include shortcut-routes (express) and those connecting other towns to the same primary railway station (possibly implicitly creating an inner-city express route). Even with the default setting for which amount of % is payed out for transfer routes I rarely ever have a negative income on any route or vehicle. Also profit is easily in the range of millions of Euros within the first couple of years.
One side note for this though: you need to pick your train and especially RV sets carefully. Especially the later one needs to have big enough passenger numbers to allow them in inner-city routes. I personally prefer the German Road Vehicles Set 0.23, but this has a few (but small) drawbacks. For example it occasionally happens that during a short period there are no vehicles at all available which can be refitted to carry tourists. This can be a problem with things like ECS Vectors (which I always use) and outdated vehicles. You can't add to the fleet during those times and your chains may collapse or at least run at sub-optimal levels. But basically it provides great buses and trams, even though it's far from finished, and works very well...
Re: Cargo Distribution
Posted: 04 Jan 2012 20:40
by Antonio1984
Hi..I'm trying to add the 32bpp graphics to this binary...I've read the read file, but without result...I am able to add these 32bpp models to other binaries, simply adding the tar files to the "data" folder...but in this case doesn't work..I've also tried to create a newgrf folder, a baseset folder...but nothing...any suggest??
thank you!
Re: Cargo Distribution
Posted: 04 Jan 2012 21:06
by JGR
The implementation of auto timetable separation which is in Chill's patchpack seems to work well enough for bus routes, after enough initialisation delay anyway.
As for RV capacities, my rule of thumb is that capacities of less than 100 passengers are not worth bothering with. I can't remember the name of the set I use. Will check when I get back. Also loading speeds are a very significant factor.
As for trams, I've not found them to be any better than buses, then again I might need to try more sets.
I've not used ECS so can't comment on that.
Re: Cargo Distribution
Posted: 06 Jan 2012 10:46
by donchatryit
Quick CargoDist related question. I have recently installed the latest version of CargoDist and now I can't open any of my previous savegames, even those that were saved with the cargodist version of Open TTD 1.1.0. Any ideas how to solve this and what is causing it? I can post a screenie, if it would be useful.
Re: Cargo Distribution
Posted: 06 Jan 2012 13:17
by Eddi
CargoDist does not provide savegame compatibility.
Re: Cargo Distribution
Posted: 06 Jan 2012 14:40
by Creat
JGR wrote:The implementation of auto timetable separation which is in Chill's patchpack seems to work well enough for bus routes, after enough initialisation delay anyway.
As for RV capacities, my rule of thumb is that capacities of less than 100 passengers are not worth bothering with. I can't remember the name of the set I use. Will check when I get back. Also loading speeds are a very significant factor.
As for trams, I've not found them to be any better than buses, then again I might need to try more sets.
I've not used ECS so can't comment on that.
Unless Chill has modified that patch it is the one that can't handle conditional maintenance orders. But I agree, it works fine otherwise. The one I'm using now (link in post above) seems to work
mostly, but occasionally I have to click on 'auto fill' again, might be because a route takes longer with more vehicles as waiting at stops is more likely.
The German Road Vehicle set has different types of buses. There are city line, regional and coach. The city line buses have quick loading times and reasonable capacities (around 100 in the 60s). The trams include double wagons which carry 260 passengers (in the early 70s I think). With this set you need to start with buses (they are pricey enough) and can later switch to trams (150k € each or so I think), which aren't affordable at the beginning.
I forgot to mention one very big difference between running local transport and extending the catchment area via bus stops and the like: Town growth is tied to Passenger deliveries. Not the number of passengers delivered, but actual drop-offs. So you can imagine the difference of using local lines (a dozen stops or so per minute) vs. an extended catchment area (generally only long-distance trains stopping every now and then). I can even give you numbers, since I didn't run a local line in one of my cities in my last game (but it was completely covered by the station).
The city without bus routes started with around 1600 residents. It reached about 2500 after 6 years or so. A city that started with roughly 2500 residents and had local bus lines was at close to 10000 after the same amount of game time. I think the bigger starting cities are the ones that also grow twice as fast (per default setting), so this does reduce the effective difference a bit. But considering I didn't even extend the original 5-stop-line to the newly grown areas in the now much bigger city (either by adding bus lines or extending the existing one).
Quite frankly I think I'll have to find a way to reduce town growth (maybe at least turn of that x2 bonus if possible), since it often spirals so far out of control that I spend most of my time just managing the bigger cities and adjusting the network around those passenger increases. Adding new stuff is after all much more fun than just adjusting existing stuff
One last thing: if you don't have that already you should set "city growth" to 3x3 grid in the advanced options (instead of 'random'), otherwise creating any sort of local transport becomes an unmanageable nightmare...
Edit: completely forgot to ask, did you check which road vehicle set you are using? And is it good?
Re: Cargo Distribution
Posted: 06 Jan 2012 17:07
by JGR
If you place conditional depot orders in such a way as to minimise the effect on the timetable, the algorithm does not seem to be thrown out too much by servicing.
Most of my RV routes don't have significant bunching problems.
I'm using the Ikarus set v5.3 and eGRVTS v1.0.
(I also have HEQS v1.1.0 in this game, but barely use it).
I tend to use "better roads" for new towns, which works well enough, though I sometimes "improve" the layout manually.
I should probably have reduced town growth before, as I now have a lot of rather large cities to deal with.
I've since reduced town growth to slow.
I tried using the town growth depends on supply patch for a short while, but the map I'm using has almost no industry (and I CBA to try changing that now), so town growth dependant on goods is not much use, though in principle I agree with it.
I spend almost all of my time adjusting existing stuff, as I run a passenger service of some kind to every town on the map already. It can be just as fun as building new stuff ^^.
Re: Cargo Distribution
Posted: 06 Jan 2012 18:23
by fonso
JGR, Creat, Wasila: It's nice to hear about those playing styles and strategies of yours. However, I think it's a bit off-topic in this thread as this thread is more about Cargodist development. I made a new thread for discussing playing strategies with Cargodist or related patches:
http://www.tt-forums.net/viewtopic.php?f=29&t=58091. Could you please move the discussion there?
Re: Cargo Distribution
Posted: 06 Jan 2012 20:13
by Wasila
fonso wrote:JGR, Creat, Wasila: It's nice to hear about those playing styles and strategies of yours. However, I think it's a bit off-topic in this thread as this thread is more about Cargodist development. I made a new thread for discussing playing strategies with Cargodist or related patches:
http://www.tt-forums.net/viewtopic.php?f=29&t=58091. Could you please move the discussion there?
Heh heh. Fair point - I'm done at least ghough (profit's way up, thank you!

) so you could get a mod to move what we already posted if you want.
Re: Cargo Distribution
Posted: 06 Jan 2012 22:59
by Dag
I have the newest versin of cargodist - 4.1.2012.
Looks like freight is handeled by cargodist even if set to asymetric.
i could recreate this in a game without any grfs. I just send wood to 2 sawmills and cargodist kicked in-
I like cargodist very much. Transporting passengers and mail without just makes no fun. Other cargoes I like to ditribute myself. So hopefully the switch to asymetrc works again in the next build
Re: Cargo Distribution
Posted: 06 Jan 2012 23:18
by Eddi
you need to switch to "manual" to turn cargodist off.