The International Simutrans Forum

 

Author Topic: EXP11.0 passenger problems  (Read 7081 times)

0 Members and 1 Guest are viewing this topic.

Offline AvG

  • *
  • Posts: 308
EXP11.0 passenger problems
« on: August 18, 2013, 11:26:30 AM »
Running EXP11.0 Year 1820.


Started in 1820 with passenger-factor 15 (standard)
In previous versions this was allways a problem in the early years.
For several monthes it seemed OK; then suddenly I got lots of pax, so many that, with the existing means of pax-transport, it could not be handled anymore.
So reset passenger-factor to 10.
After some 3 monthes I checked again. The in-town lines seemed OK (but making losses)
Noticed 2 odd things: Pax want to go to destinations that has no pax-stop. (goods only; fishing-harbor and a textile-mill)
I included via an extra busstop the fishing harbor, but had to notice that pax indeed went to the harbor, but on arrival immediately wanted to go back.
In the area where I started the sim I have four towns.
They are situated like this:
                                                 C
              B              A
                                                       D
On a given moment I saw over 600 pax in A
Stagecoaches on A>B empty. Odd!!
Analysis of the +600 pax: 176 pax want to go to B, but have a routing via C. This is impossible, because there is only 1 route A>B via road.


Analysing more intercity-stops I see more of this type of problems.


Actions I am taking:- Remove ALL intercity-vehicles to get rid of all intercity-pax (including the faulty ones) in the area.
- Upgrade to EXP11.5. (checked but did nothing on the problem)
- After gradually putting back the intercity vehicles thoroughly check what's going on.


Anybody ideas?????


PS Uploading a sve-file is difficult (6.2MB) and the use of lots of DCR-paks.


AvG




 



Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18617
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: EXP11.0 passenger problems
« Reply #1 on: August 18, 2013, 12:07:45 PM »
Thank you for your report. I am not sure that I fully understand the details of your issue, so a few questions if I may. Firstly:

Quote
For several monthes it seemed OK; then suddenly I got lots of pax, so many that, with the existing means of pax-transport, it could not be handled anymore.

It is hard to know whether having so many passengers that you cannot transport them easily in 1820 is proper simulation without knowing all the parameters of the calculations: if you are using a non-standard pakset in particular, anything there might have altered the calibration (more or larger buildings, larger towns, etc.). I am afraid that it is impossible to determine whether this is correct behaviour or not without a saved game.

Quote
Pax want to go to destinations that has no pax-stop. (goods only; fishing-harbor and a textile-mill)

I am not sure that I fully understand this: where passengers want to go is not constrained by the transport network. Passengers want to go all over the place, and it is the player's job to transport them to where they already want to go. Can you be more specific about what you mean here and why you think that this is a bug?

Quote
I included via an extra busstop the fishing harbor, but had to notice that pax indeed went to the harbor, but on arrival immediately wanted to go back.

Can you be more specific here? If this is a bug, however, reproducing it will not be possible without a saved game.

Quote
In the area where I started the sim I have four towns.
They are situated like this:
                                                 C
              B              A
                                                       D
On a given moment I saw over 600 pax in A
Stagecoaches on A>B empty. Odd!!
Analysis of the +600 pax: 176 pax want to go to B, but have a routing via C. This is impossible, because there is only 1 route A>B via road.

I am afraid that it will not be possible to reproduce this issue without a saved game. Also, what do you mean that there is only one route A to B via road? Are there any A to C or B to C routes? Are there even stops in B?

Offline AvG

  • *
  • Posts: 308
Re: EXP11.0 passenger problems
« Reply #2 on: August 18, 2013, 04:19:37 PM »
Maybe it is a good idea to tell how I play at the moment.
From the Simutrans map collection I downed a map of Holland.
With help of PSP I enlarged the map to a scale that using in Simutrans one tile is 125m.
With help of Google-maps and Excel I locate the place of towns, waterways and put them on the map. As a matter of fact I am developing a scenario on the fly, starting in 1820 and at BpM25.
Wikipedia gives info on population at the time or at least for 2013. The latter has to be scaled back for 1820.
Placing the towns and increasing the population to the right level is done via the public player. Same for rivers and other in 1820 existing waterways.
At the start I got (accidently) a couple of industries (arable farm, cattle farm, sheep farm, slaughterhouse and a market.)
A gravelwinning, gravel crusher/trader, railsleeper-factory and a railconstructer were designed. A rail-tile may be constructed when 70tons of ballast(crushed gravel) and 16 tons of railsleepers are
transported to the construction area.
I use for starters the Britpaks, but in most cases I have to alter prices, maintenance, speed, looks etc. All prices are based on real price-levels at the time (1820). Lots of reading, calculating and aducated guesses.
Starting railtransport in 1820 is difficult because there are no signals. Why is the flagman gone?? Borrowed one from EXP8.x. It works.
I am aiming for the situation that realy existing industry in Holland can be simulated.
Connecting roads are realy bad in these days and are simulated by cart-roads. These cartroads you get when you drive with horse+wagon a few times over uncultivated land. Historic files mention that in wintertime the average speed of coaches dropped to 2 miles/hour. I have choosen for 5km/h on that type of road.
All changes have to fit in the Simutrans system.


to be continued!!!

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18617
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: EXP11.0 passenger problems
« Reply #3 on: August 18, 2013, 04:36:25 PM »
That is interesting. I am particularly interested in your research on prices - would you be able to put any information that you have on real world prices in this thread?

Offline Jando

  • *
  • Posts: 486
Re: EXP11.0 passenger problems
« Reply #4 on: August 18, 2013, 09:51:42 PM »
...
In the area where I started the sim I have four towns.
They are situated like this:
                                                 C
              B              A
                                                       D
On a given moment I saw over 600 pax in A
Stagecoaches on A>B empty. Odd!!
Analysis of the +600 pax: 176 pax want to go to B, but have a routing via C. This is impossible, because there is only 1 route A>B via road.
...

I'm not sure whether I understand you right, but on my maps I see the same on some routes (but with fewer passengers cause I likely have smaller towns). It goes like this.

There are 3 stops A, B and C, in my case almost in a straight line A-B-C. There are coaches going A-B, B-C and A-C. And I see passengers at A with destination B going via C (roughly 3 times the distance of the direct route A-B), and the direct coaches from A to B will travel empty.

In my case it's a side-effect of how waiting times are calculated in the code. Like James (hi, James!) told in another thread waiting times are averaged, but not actual waiting times - and passengers look at these calculated waiting times when they decide on a route. And in some cases the numbers tell that travelling via C is faster - although in fact it takes much longer. :)

Offline AvG

  • *
  • Posts: 308
Re: EXP11.0 passenger problems
« Reply #5 on: August 19, 2013, 11:27:01 AM »
Jando,
Your example is slightly different from my situation because, the pax that have choosen to go via C, will not come back in A.


In my situation the pax make a large detour and will end at the starting point A. Town B can only be reached via one road starting in A.


AvG

Offline AvG

  • *
  • Posts: 308
Re: EXP11.0 passenger problems
« Reply #6 on: August 19, 2013, 12:28:39 PM »
James,
First I wish to make it clear that the things I mentioned are not necessaraly bugs.
Within my approach I try to make the simulation as realistic as possible, but within the possibilities of Simutrans-EXP.
I realize that there are chances, when working constantly on limits, of unwanted behavior of the standard EXP.exe.


That said and still trying to solve my problems;
- The axle_load parameter. It seems that this is not implemented, because many roads and vehicles have no values in their dat-file. Is that correct and
is the conclusion that this parameter cannot influence the pax-behavior?
Here I experienced the following: -Set up a line in Amsterdam for transport of cooled food. 4 vehicles (horses+coolwagon). Total weight 9 tons/vehiclecombo.
They have transported each over 70 tons and then I get the message: Can't find a route. I did NOT change anything. Investigation.
Then I realized that the starting point of the line was within the canals of Amsterdam, so it was necessary to pass a bridge with a max load of 8 tons.
I made a new cooled-wagon with 1 ton cap less. Problem solved!!
But still the question how it is possible to make 14 runs without a problem????


Another odd thing:
- I use cart_roads for city connection (unless it is proven that there was a paved way in 1820)
- Most of these cart_roads are laid by the public_player when erecting the towns. (see previous post)
- Cart-roads are cheap to buy  (1 HCr/km) and cost nothing to maintain.
After maybe 2 game-monthes I saw the max speed on these roads was 15km. I used not my latest insight for cart-roads . Changed to max 6km/h
Some monthes later I changed again to max 5 km/h.
Now the problem: I checked I have only ONE card_road pak in the game.
On the map however I see 3 different results.
-All public player made roads still have their original value.
-All default player roads have changed to the latest situation.


Is this explanable and wanted??
AvG


Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18617
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: EXP11.0 passenger problems
« Reply #7 on: August 19, 2013, 05:04:48 PM »
Hello - thank you for your further feedback.

I am not sure that I fully understand your original question about making 14 runs without a problem - what exactly do you mean by this? As you may have discovered elsewhere, axle load for ways is distinct to the total weight for bridges. For a bridge, the total weight of the whole convoy (including load and all vehicles) is what counts, and axle load is disregarded. For all other ways, it is the axle load that counts. If no axle_load parameter is set in the .dat files, the axle load is extrapolated from the weight divided by the number of axles (set by an "axles=" parameter). If no "axles=" parameter is provided, then the number of axles is assumed to be 2.

As to the different weights that you see on the roads, this is because the weight of any given way tile is set when it is laid, and is not subsequently refreshed from the pakset values. This is necessary to take into account limits on way objects, for example. Re-laying the roads (using the CTRL key) should re-set these values.

Offline AvG

  • *
  • Posts: 308
Re: EXP11.0 passenger problems
« Reply #8 on: August 19, 2013, 07:26:24 PM »
What do I mean with those 14 runs??
Very simple: Every wagon could make 14 runs before the program noticed they were to heavy.


OK. The road from A to B is a cart_road with a capacity of 10 tons. The stagecoach, also with full-load will never exceed that.


I understand the not refreshing of the cart-roads. I will replace them manually.


Thanks a lot James.


AvG

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18617
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: EXP11.0 passenger problems
« Reply #9 on: August 19, 2013, 07:28:08 PM »
Without being able to see this in action in an actual saved game, I can only guess what might be responsible: because, for bridges, convoys take into account their load, it might be that the load was heavier on the 14th run than on the 13 previous runs, and that the additional weight of the load took the whole convoy above the limit.

Offline AvG

  • *
  • Posts: 308
Re: EXP11.0 passenger problems
« Reply #10 on: August 20, 2013, 03:26:19 PM »
James,
Maybe you are right. The problem is probably solved.


There is however still a passenger problem. Maybe I have a direction where to look at.
This is what I did after the 600 pax problem.
I removed all the wrong cart_roads and replaced them with the right 5 km/h.
I also removed all intercity connections, so all intercitypax is gone.
I still use passenger_factor 10


This all done started again with 1 intercity coach per line.
In a month I have again in a 12K-city (Breda) over 100 pax at the townhall.
Enough to start analising.
Again an odd behavior.
6 pax with a destination in an other town (Tilburg) should go via Townhall. They don't. They have a different stop as destinatin and pass Townhall.
Luckily I have Breda divided in several sections. The S (south) section has a line with stopnames like S1, S2, etc. If a line has at the start or at the end
the same route and stops as an adjacent line you see stop-names like S3/W6.
At the moment the pax-problems seem to occur at these combined stops.
I rearranged the lines now that all start- and ending-points are the TownHall.
I still followed the 6 faulty pax for Tilburg and in the new line-system they immediately got the right destination and via stop (TownHall)
I will go on experimenting and keep you informed.
AvG


BTW: Maybe passenger-factor 10 is still to high.

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18617
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: EXP11.0 passenger problems
« Reply #11 on: August 20, 2013, 04:58:34 PM »
Remember, the passenger factor of 15/16 is intended to be used for very large games with lots of towns, when a good proportion of the passengers will want to go to towns so far away that the journey will be too long for most of them. Smaller maps will see a higher number of passengers available for player transport much earlier.

As to the specific routing issues that you describe, it is vary hard for me to follow precisely what is happening just by that description. Why should the passengers with the destination in Tilburg go via the stop near the town hall? If there are a great many passengers waiting at the town hall (if that is your connecting stop), the wait time there might have become very high.

Offline AvG

  • *
  • Posts: 308
Re: EXP11.0 passenger problems
« Reply #12 on: August 22, 2013, 08:15:01 PM »
I have done what I said in my previous post.
- All combined stops in the cities are changed in such a way that they all serve only one line.
- All lines have the same central starting point, the Townhall.


Pax with destination SE6, can decide to take line S,  leave the transport at S5 and walk the last part (0,5 km) to the destination stop SE6? You see that happen. They take a wrong line, make a large detour and leave at stops wich are not theirdestination stops. For both lines waiting-times seems extremely long because they are shown as real time; they have to be divided because of BmP25 by ~14. After that division the difference
of the waiting times of both lines is marginal. Is this behavior meant this way? If yes, perfect!!

You also see that on a line with stops numbered W1 to W7, pax have as destination W4 via W1.  (Starting at the townhall) This is already strange
because you allways pass W1 on the way to W4
They enter the bus, stay on the bus till W1 and go out the bus at that stop, instead of W4. W4 is the farest point of the line and a walk of +20 tiles.
This has nothing to do with waiting times, different pak, or whatever.


This pax-behavior needs investigation or explanation.
AvG

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18617
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: EXP11.0 passenger problems
« Reply #13 on: August 22, 2013, 09:21:24 PM »
It is very difficult for me to see everything in context to have an understanding of what is happening without a saved game: with a description, I can only guess. Firstly, the walking: passengers can in principle walk long distances to and from stops (the stop radius is now much larger), and will factor in walking time. If this is the behaviour that you intend to describe, then that is indeed intended.

As to the W1 and W4, this is not necessarily unrelated to waiting time, since waiting time is stored by destination, rather than by convoy/line. If the waiting time becomes excessive to one destination, passengers will find indirect routes faster.

Offline AvG

  • *
  • Posts: 308
Re: EXP11.0 passenger problems
« Reply #14 on: August 24, 2013, 10:33:47 AM »
James,
the W1-W4 problem cannot explained via waiting times or alternative possibilities. There is only ONE possibility: the W-line. In my example the pax enter the bus and leave on W1, where the bus is going via
W2 and W3 (without a large detour) to W4. You realy expect them to stay inside till W4.


However, I decided to put my DCR-scenario aside after playing the year 1820.
A couple of conclusions:
- I have a severe passenger problem. Passengers with destination city to the West are leaving town to cities in the East.
- Towngrowth is to fast. Way to fast. Tilburg has grown in 1820 over 30%, where I am aiming for 1,7%. I increased the grow-factors 10-fold. Have no idea if that's enough. They are however limited to 9999.
- Also the size of Tilburg is growing to fast. Especialy along connection roads. In East-West direction the size is at the end of 1820 already 7,5 km. In 2013 the real value is 10 km. I do not know if it is possible to give the growth of the town-centre more priority.

I will make a restart with a smaller map, make a copy of the problem area, and based on normal Pak-Britain try to reproduce my problems. That way I can make .sve-files for you.
For this purpose I will use BpM22
AvG


 

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18617
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: EXP11.0 passenger problems
« Reply #15 on: August 24, 2013, 11:23:35 AM »
Thank you - a saved game would be very helpful. As to town growth, Neroden found and fixed on his branch a while ago a bug relating to this. This fix should be incorporated into the next major release of Simutrans-Experimental (12.0).

I can look into the other issues in more detail with a saved game.

Offline AvG

  • *
  • Posts: 308
Re: EXP11.0 passenger problems
« Reply #16 on: August 27, 2013, 10:59:36 AM »
Made a new map of size : 639 x 447 and copied the towns of the original map.
Installed Britpak 0.9.0 and palyed it for one year on BpM22.


No pax-problems at all.


I realised when doing this I made afundamental error. I changed 3 variables at once (BmP, mapsize, Pak), wich is , when tracing a proces-error wrong.


I think however that BmP and Pak will have no influence. My pak-changes are limited in general to itemprices and speed.


My original mapsize is 2075 x 2423. Is that size considered to be medium, large, very large???


Due to the nature of my problems, pax going the opposite way compared to their destination, I could imagine some problems in formulas using these size and empowered to 2 or 3 (do not know these
formulas). If my map is quite large their might be NOT a lot of experience available.
Just for info: In one town 332 pax are waiting of wich 324 are heading the wrong direction!!!!


Does it make sence to start a new trial based on the larger map (needs lots of time and motivation)
AvG

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18617
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: EXP11.0 passenger problems
« Reply #17 on: August 27, 2013, 11:17:31 AM »
The size and shape of the map should not have any effect on this issue: the pakset might, however, as different configuration parameters may produce different results (although it is difficult to see what is producing the results that you are getting). I do not think that there is a need to start with a larger map, but you might want to examine closely all of the various settings that you have in your pakset to check for things which are outside the norm.

Offline AvG

  • *
  • Posts: 308
Re: EXP11.0 passenger problems
« Reply #18 on: September 05, 2013, 09:16:49 PM »
James,
I still have the passenger-problem.
What have I done since my last entry:
- Tried to copy my original problem- Hollandmap (EXP11.5 + Brit 090+DCR-paks) and played with the standard settings except BpM25 and stopcoverage 5 instead of 6. (So no DCR-influences)
  I also reduced some costs (monthly) in order not to get broke.
  After 1 year I stopped, because I saw no oddities.


So I decided to make a restart of my Holland-map (EXP11.5 + Brit 090 +DRC-paks)
Again stop-coverage 5, BpM25, and also max growfactors.
After one and a half month I have some 40 towns/cities. Everytime, when a new town has been made via the public player, and is connected by me and
provided with the needed infrastructure, I let the public player erect a new town. I choose this way, because setting up a complete map with all towns, roads
and waterways is an extremely boring job.
The DCR-paks are different from the originals, but only in looks, price, speed, capacity.
I use from the start onwards a combo-stop. (passengers + mail)
My horses on passengerlines are able to carry a small amount of mail.
Roads are in principe the same, but again different capacities, speed, etc)
Transport-fees are very small (based on 1820).
I add a couple of goods to reflect the 1820 Dutch economy.
All these things I do already for years without any problems.


So the new Holland-scenario is running now for 6 game-weeks.
Several 10K-towns are divided in 4 parts, where every part has its own passenger/mail-line and all 4 are connected at the TownHall. Occasionally they use
a same stop. (like Rotterdam E1 / N4)
Via the stoplist I check regularly the status of the top-10 crowded stops.


I see Eindhoven W5 has quite suddenly 250 passengers. That's a lot for a 10K-town. (After 6 weeks and no special buildings!!!)
Check: they all want to go via the TownHall to a lot a different directions. Quite normal, but a lot!!
A couple of game-hours later it shows over 300.
I started the line with a 3-persons horse-cart, added already a 6-persons Hackney, but against such numbers the normal vehicles are not sufficient.
For a situation like that I have designed a special vehicle, with a very high capacity (20/8 in 1820!!!), very low comfort and speed. I called it a Crowd-car, but
it is more like cattle-transport.
I bought 5 of them, made a schedule  W5-TownHall and tried to solve the problem this way.
Remember they all showed: Via TownHall.
Followed what happened. The Crowd-cars pick up each 28 passengers, transport them to the TownHall and are unloaded.
Seems allright, untill I discover that the number of passengers stays allmost the same. Not exactly, because the rest of the infrastucture is funtioning.
I think  that most of these passengers upon arrivel at the TownHall run back to W5. (distance 0,88 km)


So I uncheck: Allow routing_on_foot.
Result: Passengers won't leave the bus anymore.


The only way is destroying the fully-loaded vehicles untill W5 is empty.


A further inspection shows more odd behavior like: Standing at N6, wanting to go to W2 via E6/N1. Going via TownHall is a lot easier and shorter.


These experiences lead to some questions:
- Am I the only user of the latest developments? And the only one with this kind of problems? (Mapsize 2075X2432 and 200-year scenario)
- Is it likely that my DCR-stuff can cause all this?
- The W5-problem looks like a faulty passenger-generation. Is this all tested?
- Would it be wise to go back to EXP-10 and see if I also have the problems there?
AvG


 

Offline Junna

  • Devotee
  • *
  • Posts: 1081
Re: EXP11.0 passenger problems
« Reply #19 on: September 05, 2013, 09:33:02 PM »
Apologies, but you were a bit unclear, are you using 11.5 or the latest?

You should perhaps also upload your pak and save-game so it can be observed/recreated?

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18617
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: EXP11.0 passenger problems
« Reply #20 on: September 05, 2013, 09:54:52 PM »
AvG,

thank you for your report and explanation. I am afraid, however, without a saved game in which the problem can be recreated (or at least a precise description of steps sufficient for me to recreate it), I have no way of telling whether what is happening is a bug, and, if it is, fixing it, or even being able to give a sensible answer to your question as to whether the problem could be caused by your paksets. Would you be able to upload a saved game and pakset directory for me to see whether this can be recreated?

Offline AvG

  • *
  • Posts: 308
Re: EXP11.0 passenger problems
« Reply #21 on: September 05, 2013, 10:16:56 PM »
James, Junna
I will try tomorrow to make a copy of the files you ask for.
Total will be + 6 MB. Is it possible to upload via this forum??
AvG

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18617
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: EXP11.0 passenger problems
« Reply #22 on: September 05, 2013, 10:26:18 PM »
I suggest using http://files.simutrans-germany.com for uploading saved games and paksets.

Offline AvG

  • *
  • Posts: 308
Re: EXP11.0 passenger problems
« Reply #23 on: September 06, 2013, 02:38:39 PM »
Hi James, Junna,


The files are available on the mentioned German site.


Please do not look to serious at the images. They are under design/construction.


The .sve starts on 28 jan
I had a new check.
Let it run until 18 feb and look at stop Eindhoven W5. Pax varying between 0 and 450, sometimes in splitsecond.
Increase line-cap with one Hackney.
Increase    "         with 5 crowd-cars.


I am very interested in your findings.


I used EXP11.5 and just added the paks from the zip-file on Brit-090.


AvG

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18617
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: EXP11.0 passenger problems
« Reply #24 on: September 06, 2013, 06:22:26 PM »
AvG,

thank you - that is helpful. Would you mind posting the links to the files that you have uploaded here? It is quite hard to find individual files from the rather long list on that site if links are not given.

I shall investigate this when I get a moment.

Offline AvG

  • *
  • Posts: 308
Re: EXP11.0 passenger problems
« Reply #25 on: September 06, 2013, 06:38:24 PM »
James,
The filenames are:


Holland 1820-jan.sve
James-Paks.zip


AvG

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18617
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: EXP11.0 passenger problems
« Reply #26 on: September 06, 2013, 06:39:25 PM »
Thank you - will investigate when time allows!

Offline AvG

  • *
  • Posts: 308
Re: EXP11.0 passenger problems
« Reply #27 on: September 09, 2013, 10:22:24 AM »
James,
Can you, while investigating the passenger-problem, keep an eye on mail-generation. In one of your entries you write that the number would increase
due to the different mail-size (1 i.s.o. 25 kg). I have the impression that the amount of mail in EXP11.x is a lot less than in EXP10.


An observation in the uploaded sve-file:
Mid feb 1820. Tilburg TownHall 220 pax for Waalwijk. Ordered 10 coaches to solve that matter.
At hte moment the last coach was leaving the depot ALL the 220 pax changed destination to Tilburg S1/W1 ????
Hope you can find something.
AvG

Offline Junna

  • Devotee
  • *
  • Posts: 1081
Re: EXP11.0 passenger problems
« Reply #28 on: September 09, 2013, 11:52:45 AM »
An observation in the uploaded sve-file:
Mid feb 1820. Tilburg TownHall 220 pax for Waalwijk. Ordered 10 coaches to solve that matter.
At hte moment the last coach was leaving the depot ALL the 220 pax changed destination to Tilburg S1/W1 ????
Hope you can find something.
AvG

The passengers will, upon being unsatisfied with the wait, change their destination to an alternative one. It can be a bit hard to manage with slow and low-capacity vehicles in particular.

Offline AvG

  • *
  • Posts: 308
Re: EXP11.0 passenger problems
« Reply #29 on: September 09, 2013, 02:05:36 PM »
I know that Junna. Switching however from destinations of 70+ km outside a town to one inside seems highly unrealistic.


I discussed numbers already earlier here.


In 1820 in real life there was no capacity problem. Travelling was only possible for the rich and an occasionally entrepeneur.


The rest had simly no time nor money to spend on travelling.


This can be easy simulated by starting with a lower passenger_factor, which will grow over the years.


What is your experience with mail-numbers?
AvG

Offline Sarlock

  • Devotee
  • *
  • Posts: 1340
  • Languages: EN
Re: EXP11.0 passenger problems
« Reply #30 on: September 09, 2013, 02:36:28 PM »
In 1820 in real life there was no capacity problem. Travelling was only possible for the rich and an occasionally entrepeneur.

Indeed, you either walked all of the short distances you needed to travel or, if you were lucky enough to own one, you rode your horse.

Offline AvG

  • *
  • Posts: 308
Re: EXP11.0 passenger problems
« Reply #31 on: September 19, 2013, 05:54:37 PM »
James,


Just for your information. Maybe it helps.


EXP10 version from EXP Complete, has also pax-problems. A bit similar to EXP11.


Was building the same Holland-map,  like the one I uploaded for the pax-problems.


Had only 2 towns, Tilburg and Breda.


Starting on 1 jan 1820 I had in 6 weeks over 1000 pax in both cities. Passenger-factor setting 8.


Adjusted the factor to 4. No visible influence. You only see smaller numbers on buildings for pax and mail.


The same for factors 3, 2, 1. Now all numbers are zero.


Still lots of both at the stops.


Only the setting of 0 gave relief.


After using a lot of extra capacity I managed to have empty busstops. (30 stagecoaches 20 crowdcars)


Then I could see that on 1 stop, after 20 crowdcars, there were 0 pax and mail. Just a few seconds later it was building up again, and was over 300 in 10 seconds.


AvG

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18617
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: EXP11.0 passenger problems
« Reply #32 on: September 19, 2013, 10:15:21 PM »
You might be interested to know that the version on which I am currently working (which will eventually be 12.x) will have a great many more options for the customisation of passenger settings.

Offline AvG

  • *
  • Posts: 308
Re: EXP11.0 passenger problems
« Reply #33 on: September 21, 2013, 10:55:36 AM »
James,
I am glad you are working on the pax-situation.
Coulf you identify the pax-problem in my uploaded sve-file?
Is it correct that it exists already in the EXP-Complete program?


I noticed in my files I use in the early passenger-stop a number of 35 pax. Is that wrong? Does not fit 2^n.


Will EXP-12 work with earlier EXP-maps?
Will EXP-12 generate more mail than EXP11?


Any idea when we can expect EXP12? (days, weeks, monthes) Please don't feel pressed!!!!


AvG

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18617
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: EXP11.0 passenger problems
« Reply #34 on: September 21, 2013, 01:33:19 PM »
I have not looked into the specific passenger issue in this saved game, as whatever it is is unlikely to be applicable or relevant in the next version, and it is more efficient to put my limited time towards the next version than looking in detail into this specific issue. If the new version solves the problem, then there is no further work that needs to be undertaken; if it does not, there is a high likelihood that that the fix to the new version to solve it will be very different to whatever fix would be necessary with this version.

The storage capacity of stops for passengers/mail/goods need not be determined by the level any more (and so need not be a factor of two): it is quite correct, therefore, to have stops with 35 passengers as a capacity.

Simutrans-Experimental 12.x will indeed work with earlier maps; as to the mail, the mail generation rate will be customisable separately from the passengers, so whether more mail is generated is likely to depend on pakset settings.

As to the timing, that is very hard to predict, as it is difficult to predict both how long that any given task will take (because it is hard to foresee problems, solving which often takes orders of magnitude longer than adding feature, and because it is difficult to predict my availability more than a few days into the future), but I am hoping to have at least a release candidate some time this year.