Very Deceptive! (how to handle no consists?)
Moderator: OpenTTD Developers
Very Deceptive! (how to handle no consists?)
You know I watch people playing this on YouTube, and it looks deceptively easy and intuitive.
I have played OTTD with AIs and just trying to be the big dog, and got run into the ground by 1985. Regular cargo.
I have played JGR with Simple City Builder and FIRS (early basic) and run into the ground with a more or less stalled city growth.
This looks easy, but it ain't!
Also, I now know why RT2/RT3 abstract signals and junctions (with just slow downs). When it gets to be more than a few lines and complex Web of cargo dependencies, it is very easy to design an elegant network which completely deadlocks!
---
Now, in RT1(D)/RT2/RT3 there were consists which could be changed at the stations. Also, most chains were linear or, at most, dual inputs ... I never say anything like FIRS EXTREME!!!
It seems to me without abstracted platforms and consists, that to avoid deadlocking a production station, there should be a platform for each input and a platform for each output. Thus, stalled outputs cannot block inputs.
Now, if I don't make the platforms have unique names by building independent stations, then how can I give orders such only one platform will be exclusively used?
Thanks.
PS: I realized another solution. Every train contains all inputs and outputs such that production can never starve. But this seem inefficient.
I have played OTTD with AIs and just trying to be the big dog, and got run into the ground by 1985. Regular cargo.
I have played JGR with Simple City Builder and FIRS (early basic) and run into the ground with a more or less stalled city growth.
This looks easy, but it ain't!
Also, I now know why RT2/RT3 abstract signals and junctions (with just slow downs). When it gets to be more than a few lines and complex Web of cargo dependencies, it is very easy to design an elegant network which completely deadlocks!
---
Now, in RT1(D)/RT2/RT3 there were consists which could be changed at the stations. Also, most chains were linear or, at most, dual inputs ... I never say anything like FIRS EXTREME!!!
It seems to me without abstracted platforms and consists, that to avoid deadlocking a production station, there should be a platform for each input and a platform for each output. Thus, stalled outputs cannot block inputs.
Now, if I don't make the platforms have unique names by building independent stations, then how can I give orders such only one platform will be exclusively used?
Thanks.
PS: I realized another solution. Every train contains all inputs and outputs such that production can never starve. But this seem inefficient.
Re: Very Deceptive! (how to handle no consists?)
Waypoints are a good way to direct traffic. In fact, that's kinda their point Just build them directly before the station.MarkShot wrote:Now, if I don't make the platforms have unique names by building independent stations, then how can I give orders such only one platform will be exclusively used?
Re: Very Deceptive! (how to handle no consists?)
Thanks will try.
- andythenorth
- Tycoon
- Posts: 5658
- Joined: 31 Mar 2007 14:23
- Location: Lost in Music
Re: Very Deceptive! (how to handle no consists?)
FIRS Extreme is hideous, and regrettable. It was directly inspired by Railroad Tycoon 3, but took a wrong turn.MarkShot wrote:I never say anything like FIRS EXTREME!!!
Steeltown in FIRS v3 is much better than Extreme, and will be better again in FIRS v4. It's creates very linear gameplay though, no freeform sandboxing
FIRS Industry Replacement Set (released) | HEQS Heavy Equipment Set (trucks, industrial trams and more) (finished)
Unsinkable Sam (ships) (preview released) | CHIPS Has Improved Players' Stations (finished)
Iron Horse ((trains) (released) | Termite (tracks for Iron Horse) (released) | Busy Bee (game script) (released)
Road Hog (road vehicles and trams) (released)
Unsinkable Sam (ships) (preview released) | CHIPS Has Improved Players' Stations (finished)
Iron Horse ((trains) (released) | Termite (tracks for Iron Horse) (released) | Busy Bee (game script) (released)
Road Hog (road vehicles and trams) (released)
Re: Very Deceptive! (how to handle no consists?)
So, do you mean like this?
- Attachments
-
- 20190414 220352 Waypoints.jpg
- (3.32 MiB) Not downloaded yet
Re: Very Deceptive! (how to handle no consists?)
I have done a proof of concept with waypoints and I believe it has the potential to prevent deadlocks based on industry dependencies.
Thanks!
Thanks!
Re: Very Deceptive! (how to handle no consists?)
Now the only problem I see is since the trains on the main two way line are not identical, you can only have at most one of each type of cargo unless each type of cargo was doubled tracked. In which case the segregation of tracks themselves would address platforms.
Re: Very Deceptive! (how to handle no consists?)
MarkShot wrote:So, do you mean like this?
Any train coming from lower right will not enter any patform, Block Signals. Use Path signals
- Captain Rand
- Traffic Manager
- Posts: 192
- Joined: 28 Jan 2012 07:35
Re: Very Deceptive! (how to handle no consists?)
No it isn't!!!!!! (!!!!)andythenorth wrote:...............FIRS Extreme is hideous...........
I love it, and am very grateful for it. It's one of my very favourite things in my games.
I think the game would now be unplayable for me without it.
Pete.
There's nothing like a deadline to hone the concentration.
Good manners cost nothing, but earn respect.
" 'Impossible' is not in our vocabulary." Jack Chrichton, Farscape
Good manners cost nothing, but earn respect.
" 'Impossible' is not in our vocabulary." Jack Chrichton, Farscape
Re: Very Deceptive! (how to handle no consists?)
I was only thinking of platforms, but the rails themselves represent severe problems.
The standard double rail between stations only works when the numbers trains is 1/2 the number of cargo platforms.
The alternative to the above is to double rail each platform.
Effectively, the tremendous elasticity of OTTD networks is only possible if the following special conditions hold.
* Trains on the network are identical.
* Trains on the network are not identical, but sequencing their arrivals and departures are irrelevant.
---
I now understand why RT1/2/3 abstracted the rail network so much. It wasn't laziness of the programmers, but the extreme complexity players would be forced to deal with. Realistic yes, but fun no.
The standard double rail between stations only works when the numbers trains is 1/2 the number of cargo platforms.
The alternative to the above is to double rail each platform.
Effectively, the tremendous elasticity of OTTD networks is only possible if the following special conditions hold.
* Trains on the network are identical.
* Trains on the network are not identical, but sequencing their arrivals and departures are irrelevant.
---
I now understand why RT1/2/3 abstracted the rail network so much. It wasn't laziness of the programmers, but the extreme complexity players would be forced to deal with. Realistic yes, but fun no.
Re: Very Deceptive! (how to handle no consists?)
I don't really see the issue with double tracks. They have under the usual circumstances of running trains with similar speeds on them (roughly) the capacity of however many trains fit on them. What tends to lead to problems are stations and junctions.
But of course you can separate your network into all isolated lines and thus circumvent the problem. But where's the fun in that ...
But of course you can separate your network into all isolated lines and thus circumvent the problem. But where's the fun in that ...
Re: Very Deceptive! (how to handle no consists?)
I have an enormous network on a map of 8192x8192, and to say, it works well. Uh... well, at least, trains never get lost and there's no traffic jams. And this, without having the same trains or even timetables. How ? Let's say I don't only use waypoints. Signals are on the rails on every tile, there are only flying junctions, some lines are 4-track and breakdowns are disabled.MarkShot wrote:I was only thinking of platforms, but the rails themselves represent severe problems.
The standard double rail between stations only works when the numbers trains is 1/2 the number of cargo platforms.
The alternative to the above is to double rail each platform.
Effectively, the tremendous elasticity of OTTD networks is only possible if the following special conditions hold.
* Trains on the network are identical.
* Trains on the network are not identical, but sequencing their arrivals and departures are irrelevant.
---
I now understand why RT1/2/3 abstracted the rail network so much. It wasn't laziness of the programmers, but the extreme complexity players would be forced to deal with. Realistic yes, but fun no.
Re: Very Deceptive! (how to handle no consists?)
But RT2's habit of sticking everything in a hypothetical siding is tedious and you end up with hyperjammed stations where traffic is high and no possibility to expand them.MarkShot wrote:Now the only problem I see is since the trains on the main two way line are not identical, you can only have at most one of each type of cargo unless each type of cargo was doubled tracked. In which case the segregation of tracks themselves would address platforms.
OTTD rail takes time to master, but it's enjoyable. For cargo consider Ro-Ro stations for high traffic and segregate in and out lines using waypoints or build separate station - this is what would happen at real steel mills etc anyway. Double track lines can take a lot of traffic, but just like the real railways can reach capacity. Go four track if running lots of trains of different speeds on the same alignment. Learn to use path and block signals and you can do a lot (no need for one on every square though, that's overkill). Take traffic for shorter distances by road and if near water put it on a boat.
Basically see it as a learning and discovery process...you won't learn everything in a day on OTTD, but you can get better by playing and building lines. I have yet to find a better rail transport management sim.
Re: Very Deceptive! (how to handle no consists?)
I forgot, it took me years to get good at OpenTTD! Now, after about 15 years... FIRS Extreme is too easy. The default industries become harder in the long run because they increase production in response to high station ratings. If you don't want your pick-up stations overflowing with cargo, you have to build very efficient networks.
Deadlocks are The Enemy. I usually use separate stations, sometimes waypoints. Note: by control-clicking, you can have separate stations adjacent to one another. You can make it look like one station when signs are hidden.
Multi-speed lines: Recently I learned you can set speed limits! It's in the timetable; orders->timetable, click a "travel for" line. You don't have to fill in any other parts of the timetable. It's possible to make passing places, but the one time I tried I found it took about 60 squares for a 118mph train to pass an 81mph train, (both trains 7 squares long,) so I just made double tracks. I make multiple tracks a lot, even when trains are identical. Yesterday, I tried priority merging into a single line, but there was just too much traffic; either the merging trains slowed the main-line trains, or they almost never got to merge, or both. Even using much more powerful engines in the merging line wasn't enough. (That's when I first used speed limits.) Perhaps it works with very short, even more heavily overpowered trains. Those trains now have a separate oubound line for each source industry, but still have a single return line. Return of empty trains is less of a problem because the lower weight means much better accelleration, and because delays cost less.
Deadlocks are The Enemy. I usually use separate stations, sometimes waypoints. Note: by control-clicking, you can have separate stations adjacent to one another. You can make it look like one station when signs are hidden.
Multi-speed lines: Recently I learned you can set speed limits! It's in the timetable; orders->timetable, click a "travel for" line. You don't have to fill in any other parts of the timetable. It's possible to make passing places, but the one time I tried I found it took about 60 squares for a 118mph train to pass an 81mph train, (both trains 7 squares long,) so I just made double tracks. I make multiple tracks a lot, even when trains are identical. Yesterday, I tried priority merging into a single line, but there was just too much traffic; either the merging trains slowed the main-line trains, or they almost never got to merge, or both. Even using much more powerful engines in the merging line wasn't enough. (That's when I first used speed limits.) Perhaps it works with very short, even more heavily overpowered trains. Those trains now have a separate oubound line for each source industry, but still have a single return line. Return of empty trains is less of a problem because the lower weight means much better accelleration, and because delays cost less.
Extreme network builder. screenshot thread
Re: Very Deceptive! (how to handle no consists?)
I am playing my first full 100 years game JGR 0.30.3 on a 2048x2048 map with 1 AI with FIRS Extreme.
Standard vehicles just to get the feel for the whole game and master the UI.
I think in the future. I will probably be using game scripts playing for more narrow goals.
On one hand due to understanding the full game mechanics, a human has big advantage of the AI. But on the other hand, even if you allow build with pause, debugging/testing/tuning networks take a lot of running time. You can only do maybe one thing for a year while the AI can 5 if not perfect but good enough.
Standard vehicles just to get the feel for the whole game and master the UI.
I think in the future. I will probably be using game scripts playing for more narrow goals.
On one hand due to understanding the full game mechanics, a human has big advantage of the AI. But on the other hand, even if you allow build with pause, debugging/testing/tuning networks take a lot of running time. You can only do maybe one thing for a year while the AI can 5 if not perfect but good enough.
Re: Very Deceptive! (how to handle no consists?)
Yeah, time goes too fast if you try to make the network good. I don't play with AI, but vehicles are introduced and expire before I get a chance to use them. I always play with "vehicles never expire," but still never use many because by the time I get to look at them, there are better ones. A day length patch would help with this, but perhaps not with AI?
Extreme network builder. screenshot thread
Re: Very Deceptive! (how to handle no consists?)
Increase the day length factor to at least 5, and if you don't think that's enough, get it to max. 13-15 (it gets too slow past this).MarkShot wrote:I am playing my first full 100 years game JGR 0.30.3 on a 2048x2048 map with 1 AI with FIRS Extreme.
Standard vehicles just to get the feel for the whole game and master the UI.
I think in the future. I will probably be using game scripts playing for more narrow goals.
On one hand due to understanding the full game mechanics, a human has big advantage of the AI. But on the other hand, even if you allow build with pause, debugging/testing/tuning networks take a lot of running time. You can only do maybe one thing for a year while the AI can 5 if not perfect but good enough.
This will slow down a your game pace, but not your vehicles, allowing you to take more time to tune them.
Re: Very Deceptive! (how to handle no consists?)
Then, a 100 year game is going to run for months?
I already have RTW1/RTW2/MTW2 (modded) where playing Rome is like a 6 month or more affair!
Thanks.
I already have RTW1/RTW2/MTW2 (modded) where playing Rome is like a 6 month or more affair!
Thanks.
- NekoMaster
- Tycoon
- Posts: 4001
- Joined: 16 Aug 2008 22:26
- Skype: neko-master
- Location: Oshawa, Ontario, CANADA
Re: Very Deceptive! (how to handle no consists?)
Extreme is the only Economy I play.andythenorth wrote:FIRS Extreme is hideous, and regrettable. It was directly inspired by Railroad Tycoon 3, but took a wrong turn.MarkShot wrote:I never say anything like FIRS EXTREME!!!
Steeltown in FIRS v3 is much better than Extreme, and will be better again in FIRS v4. It's creates very linear gameplay though, no freeform sandboxing
Proud Canadian
Nekomasters Projects! (Downloads available on BaNaNaS!) \(>^w^<)/
# NARS ADD-ON SET 2CC | 2cc Rapid Transit For Me! (2ccRTFM) | 2cc Wagons In NML (2ccWIN)
# NML Category System (Organize your GRFS!) <- TT-Forums Exclusive Download!
Nekomasters Projects! (Downloads available on BaNaNaS!) \(>^w^<)/
# NARS ADD-ON SET 2CC | 2cc Rapid Transit For Me! (2ccRTFM) | 2cc Wagons In NML (2ccWIN)
# NML Category System (Organize your GRFS!) <- TT-Forums Exclusive Download!
Re: Very Deceptive! (how to handle no consists?)
I must admit I did look at TT original recipe economy, but it is not worthy of the sophistication available in the rest of the game.
Who is online
Users browsing this forum: Google [Bot], jfs and 12 guests