The International Simutrans Forum

 

Author Topic: [Experimental release] Pak128.Britain-Ex 0.6  (Read 18132 times)

0 Members and 1 Guest are viewing this topic.

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18679
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
[Experimental release] Pak128.Britain-Ex 0.6
« on: April 10, 2010, 08:26:53 PM »
A new version of the preview and testing version of the Simutrans-Experimental version of Pak128.Britain is now available: download it here. This is a complete pakset and should be installed in its own directory. It is only compatible with Simutrans-Experimental (version 7.2 and higher) - the normal Pak128.Britain should be used with Simutrans-Standard.

This is a preliminary version meant for testing and evaluation, although it should balance better with Simutrans-Experimental than the standard Pak128.Britain. Not all of the Simutrans-Experimental features are implemented yet (although I am planning to implement them all in due course).

Here is a list of the features supported by this release of Pak128.Britain-Ex:

  • Comfort
  • Loading times
  • Catering
  • Upgrading (partial)
  • Way constraints
  • Reversing
  • Tilting trains
  • Overcrowding (only on a few vehicles so far - work in progress)
  • Industry obsolescence
  • Industry upgrading
  • Weight limits
  • Tractive effort

See here for a more detailed description of the work outstanding and so far completed, and also for information on how to get involved in contributing to the project and obtaining the sources.

Changes from 0.5 include:

  • CHANGE: Add comments explaining the effect of the new comfort settings in simuconf.tab
  • FIX: Diesel barges did not have the appropriate way constraints set
  • FIX: River boats can now use rivers.
  • CHANGE: Power corrected for boats to avoid using gear.
  • FIX: Fields absent with recent versions of Makeobj
  • FIX: Improved graphic for 1880 office building
  • FIX: Incorrect price for Class 47 locomotive
  • CHANGE: Updated graphics for the Class 4MT Tank, Mk I RBR and Mk III TRB
  • CHANGE: The substation graphic is now used as a functional substation and not as a city building
  • ADD: Combined electrification types
  • FIX: Missing translation texts for BR Mk3a vehicles
  • FIX: Missing translation text for BR Class 153 with trolley service
  • CHANGE: More accurate graphics for BR Mk. I and Mk. III catering vehicles

Any feedback on this pakset variant (especially relating to the gameplay balance) would be very much welcome.

Offline neroden

  • Devotees (Inactive)
  • *
  • Posts: 831
  • Nathanael Nerode
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #1 on: April 10, 2010, 11:35:50 PM »
Any feedback on this pakset variant (especially relating to the gameplay balance) would be very much welcome.

I can already tell you that in 0.5 it was still too hard to make money in the early years (pre-1900).  I haven't actually gotten a game past 1900.

Don't change too much yet though.... now that the code base is stabilized I'm going to go back and dig out my balancing spreadsheet.  I wonder if I can make an automated program to extract the vehicle information from the .dats into a spreadsheet, or vice versa -- maybe something in Perl.  It's particularly obnoxious entering every single new vehicle by hand into the spreadsheet.  Since OpenOffice .ods spreadsheets are actually a relatively straightforward form of XML (zipped) this might be easier than it looks....

There is one thing I can tell you immediately though, which *does* need to be improved and isn't purely balancing.  In the pre-1900 era, the passenger road transport methods get faster and faster.  Putting the 6 kph mail cart on the same roads slows the horse omnibuses down as they get stuck behind it.  Unfortunately, the faster mail coaches become obsolete with no replacements until significantly later, so there's a period in the 1860s and thereabouts when it is simply not reasonable to carry local mail.

This affects Pak.Britain standard too, but it's much more important for Pak.Britain experimental, because the slowed-down passenger omnibuses *make less money* in Experimental because their average speed is slower.
It's important to have local-mail-delivery options which can keep up with the local-passenger-delivery options. 

Sorry if this is all fixed in 0.6, but I only just figured this out so I thought I'd mention it....

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18679
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #2 on: April 10, 2010, 11:48:35 PM »
Neroden,

thank you very much for your feedback: it is much appreciated. Those things have not changed from 0.5 to 0.6 - I shall have to look into that. Which specific mail vehicle becomes obsolete with no replacement until later? Do you mean the mail stage coach? That was not intended for local mail transport, but for inter-urban mail transport, which is overtaken by railways in the 1840s, so it is intentional that it is retired with no replacement until later. The idea with the mail cart is that it is simply used to collect mail from various places around the town and take it to a railway station, then take the mail from the railway station to various places around the town. Perhaps it should be made a little faster - say 12 - 15kph? Or do you know of what was done historically to transport mail in such circumstances?

Offline neroden

  • Devotees (Inactive)
  • *
  • Posts: 831
  • Nathanael Nerode
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #3 on: April 11, 2010, 12:50:44 AM »
Neroden,

thank you very much for your feedback: it is much appreciated. Those things have not changed from 0.5 to 0.6 - I shall have to look into that. Which specific mail vehicle becomes obsolete with no replacement until later? Do you mean the mail stage coach? That was not intended for local mail transport, but for inter-urban mail transport, which is overtaken by railways in the 1840s, so it is intentional that it is retired with no replacement until later. The idea with the mail cart is that it is simply used to collect mail from various places around the town and take it to a railway station, then take the mail from the railway station to various places around the town. Perhaps it should be made a little faster - say 12 - 15kph?

That's reasonable.

Quote
Or do you know of what was done historically to transport mail in such circumstances?

Historically, people walked to the post office to drop the mail off and pick it up.  In other words, the post office would have had a much larger catchment area prior to the establishment of pillar boxes (in 1853).

In the US, 1863 was the beginning of free household delivery within cities.  I'm not entirely sure what they used, but I'm guessing it was a guy walking around with a bag -- who doesn't interfere with the road traffic at all  :-)  I couldn't find the UK date in five minutes.

I don't think simutrans is set up to model either of these situations properly, and it would probably be too much work to do so in the short run (though if you ever do implement 'walking to the station', then you'd probably have the infrastructure for it).

So in the interests of making the game work as a game, the cart should run at 15kph, which is the top horse speed.

Or, wait, I have a better idea: the horse omnibuses and the hackney cab could be given a (very) small mail capacity, maybe 1 or 2.  That's actually a cute and sort-of-realistic solution -- people *did* hand letters to cab drivers to take to the station, I've read about it in Victorian novels.

But would the man on the bike still be useful?  I guess he's still useful for after 1918, as there are still small amounts of mail to be collected in many places, and the big mail trucks can't be run profitable.  He would also be faster, but would be slowed down by all those omnibuses on the road (so better make him break even at 15 mph).

---

One more "balancing" issue before we get to pricing: in 0.5 the "light traction engine" is apparently not actually powerful enough to haul, at any speed above 0, any of the Eagle Trailers if they're full -- and at the time of introduction these appear to be the only trailers it is legally allowed to haul.  :-P  This combination seems undesirable as it makes the entire set of vehicles unusable (since you can't *prevent* a cart from loading up to full capacity).

---
Um, one more thought.   :P  In experimental  there are some problems with the wooden road bridges.  There's WoodTrestleRoad (999t weight limit), which is probably from a merge from standard; there's Wooden Road Trestle (Macadam), with brick artwork, and Wooden Road Trestle (cobblestone), with brick artwork.

In fact, as far as I can tell, wooden road bridges came in "just wood" and "tarmacked wood".  As far as I know, you do not put cobblestones on a wood bridge substructure; the extra weight hurts and the surface is no better than wood.    You also don't usually put macadam on it, because the wood surface actually gives a better ride than macadam.  (Tarmac is a better surface, so wooden bridges do get tarmacked.)

Also, the initial "Wooden Road Trestle (Macadam)" is used to build the initial intercity roads, but it won't carry a stagecoach.  I suggest:
- eliminating Wooden Road Trestle (Macadam) and Wooden Road Trestle (Cobblestone)
- creating a Wooden Road Trestle (Wood Deck) with a weight limit of 3t (so it can carry the stagecoach) and a speed limit of 15, using the wood trestle artwork from Standard.  Having to slow down for bridges was *classic*.
- Keeping the Wooden Road Trestle (tarmac) and using the wooden bridge artwork until you can get some artwork of a paved bridge.
« Last Edit: April 11, 2010, 01:31:32 AM by neroden »

Offline Dutchman on Rails

  • *
  • Posts: 239
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #4 on: April 11, 2010, 12:05:08 PM »
I'm sorry to have to report that at the moment of writing the experimental.simutrans.us site with the download links for the latest Simutrans-Ex, Config files and Pak128.Britain-Ex shows connection problems. It appears to be a recurring problem that solves itself after the first two days following each release, so I hope better luck tonight or tomorrow.

Regarding the balancing, I know it's difficult. I generally play starting in 1750, and need some changes in the .tab files to help me along as well as a radically different management style from Simutrans standard. However, having figured out these things, I no longer have much trouble with the balance. In fact, I find it challenging and giving me a feel of the difficulties involved in that time.

This is not the first time I've read that people find the early game in Simutrans-Ex overly challenging. Would it help if - once I find a version that is enough trouble-free, I write about my experiences in a detailed story?

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18679
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #5 on: April 11, 2010, 12:22:06 PM »
Neroden and Dutchman,

thank you both very much for your feedback - it is much appreciated. As to the mail - what I shall do for the time being is have a new mail cart introduced in 1825 with a top speed of 12kph, and a very slightly higher purchase cost, but otherwise identical to the normal mail cart. I shall also increase the top speed of the man on a bicycle from 20kph to 22kph.

As to the bridges - I have evidently made an error integrating the wooden trestle road bridge from Standard, especially if the graphics are all wrong for the wooden road bridges with weight limits. Your suggestion does seem sensible, however, so I shall look into following that.

As to Dutchman's comments on the balancing - I'd be very interested indeed in any information as to how the necessary management style differs between Standard and Experimental, and also in any saved games in which you have deployed this updated management style. I should further be very interested in the details of what you have had to change in relation to the .tab files. Having the information now rather than later would be most beneficial, as I can see then how the game works in practice and take that into account in future development.

However, if you want to post a story in the "arena" forum going through your experiences chronologically with screenshots and saved game downloads, that would be excellent, too. It would be very useful for me to see the evolution of a game in action, and also might help to generate interest in Simutrans-Experimental. I am hoping that version 7.3 should be reasonably stable and well-balanced, so do have a go and see whether you can set that up. I should be intrigued to see it!

Offline AvG

  • *
  • Posts: 308
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #6 on: April 11, 2010, 02:22:41 PM »
In Exp-DRC I already use for "ages" the possibilty of carrying mail by a horse/wagon-combination.
1 bag of mail per horse. So a 4-team span can carry 4 bags. It's in the .dat-file of the horses.
Be carefull with the speed of those horses. DRC uses max-speed 10 km/h.
I am old (72) but have the advantage of having seen in reality horses at work.
At normal speed you could as a pedestrian easily speed up with a towing horse. (so 10 km/h is already quite high)
And I will never apply for your postman-job if I have to do 20-25 km/h with my loaded bike. (a modern postman-bike can carry 55 kg mail)
BTW: DRC uses also roadstops with a mailbox.(intended for 1900-1930)
AvG

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18679
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #7 on: April 11, 2010, 02:36:43 PM »
AvG,

thank you for the interesting information and thoughts! As for the bicycle - I put that as the maximum speed, not necessarily the actual speed: the actual speed is determined by a combination of the power and the weight, so, if those things are set up correctly, it should give a fairly realistic speed.

The reason that I gave the speed as I did was because I ride a bicycle to work quite often, and usually do between 15-20 miles per hour even with a bit of a load, although that'll be less than the postman, I suppose. From what I understand, mail bags used to be a lot lighter in the old days than they are now.

Offline neroden

  • Devotees (Inactive)
  • *
  • Posts: 831
  • Nathanael Nerode
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #8 on: April 11, 2010, 03:48:01 PM »
Footnote on the light traction engine.  It actually will haul an Eagle Trailer at 4km/h. 

However, it still claims its maximum speed is 0 km/h....  Hmm?

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18679
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #9 on: April 11, 2010, 03:54:56 PM »
I think that 4kph is a hard-coded minimum speed to avoid deadlocks...

Offline neroden

  • Devotees (Inactive)
  • *
  • Posts: 831
  • Nathanael Nerode
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #10 on: April 11, 2010, 09:12:22 PM »
Got one more for you: water transport is still messed up.  There are no passenger or mail ferries which can run in "small waterways" in the pre-1900 period.  There should probably be at least one....

Also, at the moment "Stream" is non-navigable (good), "Small River" is apparently not classed as a waterway (?!?), "River" is a "Waterway" and a "Small Waterway", and "Wide River" is a "Waterway".  Canals are "Waterway" and "Small Waterway".

I think "Small River" should be a "Waterway" and a "Small Waterway", while "River" and "Wide River" should be "Waterway".  As it is there are really not very many waterways which aren't small on any given map, even with the river settings tuned to generate as many Wide Rivers as possible.  And there are enough obstructive streams without making Small River useless.  Plus which it looks like the current state of Small River is an accident (it has an appropriate speed limit and weight limit and is "navigable", but it doesn't have the "Waterway" character...

Looking at it, there are some problems remaining with the ships.  Only the horses and horse-drawn barges have "MAY USE: Small Waterway MUST USE: Waterway".  The wind ships, all of them, have *NO* modifiers (which means they may use both the ocean and waterways, but not small waterways). 

This looks like a pak generation error?  (One you thought you'd fixed?)  The source files make it seem like they ought to have different settings....  the big ships should be ocean-only, the medium ships ocean and waterway, but not small waterway, and the smallest ones waterway and small waterway, but not ocean.  But that's not what's showing up in-game.

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18679
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #11 on: April 11, 2010, 09:21:04 PM »
Neroden,

thank you for your feedback. The boats in Pak128.Britain are work in progress, so I'd expect some things (ferries, etc.) to be missing. Steamships and diesel/oil powered ships are incomplete.

As to the way constraints - I shall have to look into that, I think...

Offline Dutchman on Rails

  • *
  • Posts: 239
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #12 on: April 12, 2010, 07:09:12 AM »
Hi James,

Re: Game balance and playback/tutorial.

Well, so far previous games have been plagued by such issues I've reported earlier and were subsequently fixed as livestock drovers moving at 'light speed', not being able to place signals (knocking the railways out of the game), industries having unlimited production, industries springing up too fast etc. For me, to be able to write a decent play report stretching out over decades of play, it has to be fun playing... ;)

However, I can give some pointers here. I will make a separate topic in the 'discussion' section as not to pollute this thread further.

Offline sdog

  • Devotee
  • *
  • Posts: 2039
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #13 on: April 15, 2010, 07:40:26 AM »
some observations when using Pak128.Britain-Ex 0.6 with exp 7.3:

- catenary for normal tram tracks in 1945 can not be built
- blackpool balloon and brush have absurdley high loading times. unless they were complete construction failures, trams should load quicker than buses.
- heavy tram and normal tram track look the same, and can easily be confused.
- scammel mechanical horse is obsolete in the same month when scammel scarab is introduced, making upgrade difficult.
- morris eigh mail van doesn't make a profit, when loaded 95% on both routes, and going mostly over 80km/h roads.

- it is really cheap to bulldoze houses. it is often cheaper to doze a hole block and use the lower/raise land tool than using the very expensive artificial slope tool to level one field.

Offline neroden

  • Devotees (Inactive)
  • *
  • Posts: 831
  • Nathanael Nerode
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #14 on: April 17, 2010, 03:56:26 PM »
James, I made changes corresponding to my proposed road-bridge changes in my fork of the pakbritain-experimental git repository. :-)

They seem good, but I don't know how to write specifications for upgrades so that people with the 'old' bridges can load their games.  (WoodTrestleRoad->wooden_road_bridge_wood, wooden_road_bridge_dirt->wooden_road_bridge_wood,wooden_road_bridge_cobbles->wooden_road_bridge_wood).

EDIT: I also made the obvious change of making the "small river" into a "small waterway".  Changes to the others require more thought.
« Last Edit: April 17, 2010, 04:14:25 PM by neroden »

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18679
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #15 on: April 17, 2010, 04:47:12 PM »
Sdog,

thank you for your feedback. The problem with tram catenary is actually a problem in the code that I have now found and fixed (relating to way constraints). The two different types of tram tracks share graphics because nobody has yet had time to produce different graphics for them; I am unsure how they would actually look different in any event. Do you have any suggestions as to how to differentiate?

As to the trams - I didn't know that trams have lower loading time than 'buses - is there a particular reason for this? The Balloon trams were actually quite successful: indeed, they are still in service (I rode on one last Monday). The issue about the obsolescence dates is dealt with in another thread.

Finally, the Morris 8: Pak128.Britain is not fully balanced yet: I am awaiting some hopefully quite detailed data on running costs, etc. However, I have noticed that the Morris Eight mail van has a running cost inconsistent with other Morris Eight vans, and have corrected that ready for the next version.

Neroden,

thank you for your work - I was planning to do that myself when I got around to it. As to specifying upgrades from previous version of the pakset, use compat.tab: the format should be apparent from looking at what is there already.

Offline sdog

  • Devotee
  • *
  • Posts: 2039
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #16 on: April 23, 2010, 06:50:50 PM »
james,

i've tried a passenger game starting in 1945, as i mentioned earlier. here are the savegames. it's far from being profitable:
http//dl.dropbox.com/u/1876190/EX%205.sve
http//dl.dropbox.com/u/1876190/EX%206.sve

i tried to link the two largest cities (the ones with tram lines) with a train line, and tried to increase the population until it became profitable. i still couldn't get close to profitability even when increasing both to over 40k population. I used the second cheapes tracks, rated at 140 km/h and swindon twinsets and also the heavies steam locomotive still able to run on it.

the second savegame also has the problem of being in the replacement dilemma.

one of my bigger problems is the not very dense urban sprawl, requiring a lot of stops and lines.

btw, more than half full loaded Balloon trams can't move on standard tram rais, they are heavier than 35 tons.

i know it's not supposed to be well balanced at this time, but i still give you information, maybe you can use it someway.
« Last Edit: April 24, 2010, 03:39:42 PM by sdog »

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18679
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #17 on: April 24, 2010, 12:57:50 AM »
SDog,

thank you very much for your work and analysis. As you correctly identify, much of the problem appears to be attributable to the urban sprawl, which, in turn, is related to a lack of variety of high density buildings in the 1920s-1950s era. I know that there has recently been a great deal of work to extend the number of buildings in Victorian times, but it seems that more work is necessary in later times, too. It is well-known, in real life, that it is very difficult for public transportation to be profitable with very low density developments.

Looking at your railway line, if all that you can fill is one single carriage train on a single track line operating with only one other such train and one-tile platforms, you really don't need a train at all (at least, not in 1945) - a 'bus will do perfectly well for that route. It is notable that the routes on which you make a loss are those with a high free capacity : transported ratio: on those where the green graph (transported) is above the red graph (free capacity), you make a profit; but those are a minority of routes.

One thing that I am working on is allowing older buildings to be built when new towns are generated (as opposed to when towns grow naturally): this might alleviate the problem a little, but, really, Pak128.Britain needs more high density buildings in the inter-war era.

In the meantime, you could always try starting in either 1955 or 1905 to see whether that makes any difference. Thank you again for your work on this!

Offline sdog

  • Devotee
  • *
  • Posts: 2039
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #18 on: April 24, 2010, 02:17:51 AM »
oh, i thought those new, denser buildings would also be built until the modern ones in 1950 are built. but they also become obsolete. One of the reasons i stared in 1946 was to see the new buildings :-)

My earlier experience was that with buses on longer routes not enough passengers will use a route, the travel times just get to high. This also one of the reasons there is a tram backbone of the trafic system, a bus only system lost too many passengers. Bunching of buses alleviated or perhaps even caused this. Something that also changed is the sheer number of passengers is lower now, before i had problems to cope with the large number of short distance travelers, that required very high capacity transport systems.

The railway line extension to Leicester is likely not necessary however. But it is cheaper in inf. maint. than a tram line. building it was also rather inexpensive, as removing roads and houses is quite cheap in the urban sprawl.

Related to this is also another observation: Since the artifical slope tools are extremely expensive (5k$) in comparison to removeing city buildings, it's often better to buldoze a quarter and use slope up or down (1k$), or to bulldoze a whole corridor through a city to avoid a slope. The second has also the advantage of considerable savings at stops (the connecting stops between forward and back way can be saved.)

You could always use one of the victorian buildings as placeholder for this time. Your plan however would also create more realistic cities.

Perhaps, you can solve it by doing the city creation in several growth steps, each at a different set time. example: going first 100 years back, grow city to 1/4, jump 25 years forward grow to 1/2 etc. This would also cause the city to have an old core and newer surrounding area.

Regarding your earlier question on how to make the different tram tracks less confusable: The main issue is as i see it, the almost identical icon. Just putting an H for heavy in the icon would already help a lot. When the bug with the catenary is solved, it will hardly be used, except where it really is needed. So player will know where they built it. My difficulty with it came mostly from accidentially choosing it (it's the first icon) and later had to check the whole track for a few slow patches.

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18679
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #19 on: April 24, 2010, 12:50:47 PM »
Sdog,

thank you for your feedback. I have taken your suggestion with the tram tracks, and modified the icon (and made the heavyweight tracks themselves slightly lighter, although the effect is barely noticeable).

I did have some trouble, however, implementing in the code a system to allow the construction of older buildings when a town is built: for some reason, only one type of older building was ever built: a particularly small 17th/18th century house with a very low passenger/mail level. Given that I have quite a number of other things that need doing on the code, and that Prissi has indicated in the past a desire to implement the desired feature into Simutrans-Standard at some point, I shall leave this one for the time being, on the basis that it's probably quicker just to make the new buildings for that era.

I noticed the low number of passengers overall: this is partly as a result of the lower density buildings, and perhaps partly a result of the passenger factor setting: I shall try reverting to the default of 16 and see whether that helps somewhat. However, I did, using a debugger, change the level in your game to 28, and the passenger numbers were still quite low: the number of private car trips were fairly low, as were the number of "too slow", which suggests that the building density issue is the real problem.

In relation to the journey time tolerances, I shall change them in the next release of Pak128.Britain-Ex to:

Code: [Select]
min_local_tolerance = 35
max_local_tolerance = 90
min_midrange_tolerance = 60
max_midrange_tolerance = 240
min_longdistance_tolerance = 150
max_longdistance_tolerance = 1080

to see whether this balances better: they may have been to stringent before. I shall also allow a slightly higher proportion of passengers to be either mid-range or long distance compared to local (40:38:23 local:midrange:long-distance as opposed to 43:37:20), and allow mid-range journeys to be slightly longer.

As to the cost of removing buildings: in Simutrans-Experimental 8.0, on which I am currently working, the cost of removing stations will be based on the cost of the stations themselves, rather than the generic "cost_multiply_remove_haus" setting. This means that the "cost_multiply_remove_haus" can be increased without making it too expensive to bulldoze one's own stations, resulting in a greater differential in cost between bulldozing one's own buildings and bulldozing city buildings. In the next release of Pak128.Britain-Ex, the cost of removing city buildings will be increased substantially.

Perhaps you could have a go starting a new game in about 1910, with the altered journey time tolerance settings listed above, and see how you get on? Thank you again very much for your investigations here: they are extremely helpful.

marcosrcc

  • Guest
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #20 on: April 24, 2010, 02:27:22 PM »
The link to download the pak is broken at this time :(

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18679
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #21 on: April 24, 2010, 02:52:27 PM »
Macrosrcc,

hmm - it seems to be working for me. Have another go...?

Offline sdog

  • Devotee
  • *
  • Posts: 2039
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #22 on: April 24, 2010, 11:12:41 PM »
just created a couple of maps in 1910, setting 80 cities at a 1000x500ish map. median pop 6k.
the trend was that most cities had only high density buildings, even the wee villages.

one extreme example of a large town:
http://dl.dropbox.com/u/1876190/Screenshot-1.png

maybe the streets couldn't expand, and it thus got stuck. however in the south-eas there would be a free street to expand along.

next time i'll try a flatter map, to see how city generation then will look like.

edit.:
this might be more experimental than pak related, sometimes a bit difficult to find the correct thread for it.

marcosrcc

  • Guest
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #23 on: April 25, 2010, 12:01:47 AM »
Macrosrcc,

hmm - it seems to be working for me. Have another go...?

Not here :(Not Found
The requested URL /Pak128.Britain-Ex.zip was not found on this server.
Apache/2.2.9 (Debian) PHP/5.2.6-1+lenny8 with Suhosin-Patch mod_python/3.3.1 Python/2.5.2 mod_perl/2.0.4 Perl/v5.10.0 Server at experimental.simutrans.us Port 80

The config files also show the same error, but the binaries are fine.

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18679
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #24 on: April 25, 2010, 12:03:41 AM »
Sdog,

this is actually a known issue, and equally affects Standard. Recently, Knightly posted a patch to fix it (for Standard), but it was rejected because it slowed down map generation too much.

Edit

Macrosrcc,

I have just tried it again with this link: http://experimental.simutrans.us/Pak128.Britain-Ex.zip, and it still works fine for me.

marcosrcc

  • Guest
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #25 on: April 25, 2010, 12:14:55 AM »
That's weird. I have no trouble downloadind anything else other than the pak and the config files.

I tried a couple web browsers and even wget. No luck. 404 not found.

Is there any mirror? A torrent?

Cheers!

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18679
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #26 on: April 25, 2010, 12:28:52 AM »
That is really very odd indeed. I'm not aware of any mirrors or torrents. Has anyone else had any trouble?

Offline Lmallet

  • Devotee
  • *
  • Posts: 727
  • 128 on the brain...
  • Languages: FR, EN
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #27 on: April 25, 2010, 01:43:32 AM »
Macrosrcc,

I have just tried it again with this link: http://experimental.simutrans.us/Pak128.Britain-Ex.zip, and it still works fine for me.
This link does not work for me.

EDIT:  seems to have a mind of its own, works now.
« Last Edit: April 25, 2010, 05:02:48 AM by Lmallet »

Offline romik

  • *
  • Posts: 61
  • Good Games Simutrans
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #28 on: April 25, 2010, 11:55:29 AM »
. . .

Edit

Macrosrcc,

I have just tried it again with this link: http://experimental.simutrans.us/Pak128.Britain-Ex.zip, and it still works fine for me.
Error (Webová stránka "experimental.simutrans.us" nebyla nalezena)
The International Simutrans Forum > Language boards > CS - Česky (Czech) > Poradna  (Moderator: VS) > Topic: problem http://forum.simutrans.com/index.php?topic=494.msg46724#msg46724
« Last Edit: April 25, 2010, 12:04:34 PM by romik »

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18679
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #29 on: April 25, 2010, 12:30:58 PM »
Very, very odd. It's still working for me. Very idiosyncratic. For all those who are having trouble, I can only recommend trying again regularly...

Offline sdog

  • Devotee
  • *
  • Posts: 2039
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #30 on: April 26, 2010, 05:53:31 AM »
Just tried a game starting in 1910. Well, just is actually yesterday evening, since i couldn't stop.

encountered two minor bugs:
- heavy tram tracks on wooden trestle bridge images are not alligned properly, trains cross without problem though. (near Ripon in the savegame)
- petrol electric omnibus in 1910 can't move when loaded.

economics at this time works very well. provided enough income at small sclaes (connecting nearby 10k cities) but stayed rather challengeing also with a 70k city. Only after having saved enough to link both population centres, maybe 100 km appart, i make quite good money (about 10k/month). In standard with pak 128 this would be a game breaking cash-cow route.

This rail link was only possible to me after the schools engine became available in the 30s. It is fast enough, but stays under 65t. Heavier tracks were to expensive for me to build.

Before i tried a ferry link, it drew hardly any passengers however. I was running 3 clyde steamers. Both had an average load of less than 10 passengers, but were still profitable. Their max load btw is 450 passengers.

The ships took way too long, the maximum travel time of the passengers must have kicked in. It was also noticeable that reducing the number of vehicles on a route can sometimes also reduce the average load of the vehicles. This was very apparent with those boats. One ship could not attract any passengers, two could attract some, but the line was not profitable. With 3 or 4 ships the line became profitable.

The same effect happened with buses and trams. The typical simutrans way of just culling the line whe it's non profitable does not always work. In some cases adding another vehicle will even trigger the route from being completely empty to be fully loaded.

http://dl.dropbox.com/u/1876190/EX11.sve

Offline neroden

  • Devotees (Inactive)
  • *
  • Posts: 831
  • Nathanael Nerode
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #31 on: April 26, 2010, 05:03:10 PM »
- heavy tram tracks on wooden trestle bridge images are not alligned properly, trains cross without problem though. (near Ripon in the savegame)
This is a new report, thank you :)

Quote
- petrol electric omnibus in 1910 can't move when loaded.
Part of the mass report I made in another topic ("ships too powerless"); there are a lot of these "0km" vehicles.  :(  They have been my major nemesis in running a long game.
Quote
The ships took way too long, the maximum travel time of the passengers must have kicked in. It was also noticeable that reducing the number of vehicles on a route can sometimes also reduce the average load of the vehicles. This was very apparent with those boats. One ship could not attract any passengers, two could attract some, but the line was not profitable. With 3 or 4 ships the line became profitable.

The same effect happened with buses and trams. The typical simutrans way of just culling the line whe it's non profitable does not always work. In some cases adding another vehicle will even trigger the route from being completely empty to be fully loaded.
This is actually excellent and realistic -- this is the effect of waiting time.  A second vehicle cuts the passengers' waiting time in half, making the route much more attractive.  The time tolerances did need to be adjusted a bit because certain logical ship routes simply attracted 0 passengers -- I think they're going to be user-adjustable in 8.0.

I've also found that adding more vehicles to a line can *reduce* the number of passengers, if it causes congestion at an intersection or grade crossing, because the travel times increase due to congestion.  This provides a natural incentive for grade separation and careful rerouting, but only on high-volume lines, which is kind of cool. :)

Offline sdog

  • Devotee
  • *
  • Posts: 2039
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #32 on: April 26, 2010, 05:15:37 PM »
Yes it's great, it seems that some additions to the model increased it's complexity greatly. Suddenly a lot of real world effects appear nobody has actually modelled, and i guess not anticipated.

We have two targets for optimization in a route now, passenger numbers and vehicle utilization. With a limiting factor of congestion. Compared to only one in standard, it's quite an advantage.

Now it also makes sense to have ring lines around cities, with two rings rotating opposite ways. In standard passengers only use one line and travel the whole ring to get one station back. simuexperimental-citizens are smarter. Travel times are reduced considerably. I really noticed an impact in my profits when adding the second ring.

Offline neroden

  • Devotees (Inactive)
  • *
  • Posts: 831
  • Nathanael Nerode
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #33 on: April 26, 2010, 06:52:38 PM »
Well, James, as of the latest (-devel) changes, we now have a different city problem: when I start in 1800 with fairly small cities, I get *nothing but* low density buildings in *every* city, making it impossible to run any passenger lines at all.

I used to get mostly high density with a few low density around the edges.

Somewhere in between the two extremes would be fine.
I'm guessing this is due to the city growth changes in 7bd1947ea90be82674c586691f999dec5e799f3c

Rather than fixing this by changing the city construction model (again) I think perhaps it's best to change it by vastly increasing the renovation percentage.  The default in pak128.Britain-Ex is 12, which is WAAAAY too low.  I kicked it up to 33.

....AND, tested with 33, and the cities come out looking beautiful.  :)

Better yet, the "Divide renovation percentage by 3 when using quick_city_growth" means that this is equivalent to 11 for the "old" model, so it should continue to work the "old" way on the "old" model.

Offline neroden

  • Devotees (Inactive)
  • *
  • Posts: 831
  • Nathanael Nerode
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #34 on: April 26, 2010, 07:10:15 PM »
Another comment on balancing in -devel: the cost of demolishing houses used to be too low, but now it's too high!

The ultra-cheap, zero-passenger, zero-mail house, in 1800, *formerly* cost 250 to demolish; now it costs 6250 (!).  The bigger ones cost $25,000 or more!  Practically speaking, in the real world it usually *has* been cheaper to demolish small houses than to build viaducts over them, so this multiplier is simply far too high.  The minimum-level house should cost around 1000-2000 to demolish.

Also, on an unrelated note, the "Buy House" option from standard seems to have vanished from experimental.  (Or is it the pak it's missing from?)

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18679
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #35 on: May 09, 2010, 05:02:40 PM »
Apologies for not having replied to this sooner: this is a fascinating saved game and accompanying analysis from SDog, and I did want time to reply properly. Firstly, I am very glad that some of the design aims of Simutrans-Experimental are finally percolating their way through to people's playing experiences: we have suburban rail, efficiency of network design, choosing the most suitable vehicles for local transportation, and an end to fun-stopping mega-profits all demonstrated with great effectiveness in this one saved game. It is very interesting to note the entirely realistic correlation between ridership and frequency, which is entirely absent in Standard: this is a most encouraging sign.

It is evident that some things still need refining, however. The Schools class locomotive's real-life weight is 68t, so some further investigation is needed into the correct balance of the weights/prices of railway track. It should be possible for a player to afford the heaviest, fastest track if he/she is frugal and has a large enough empire. Might I ask - was there a reason that you didn't use earlier, light locomotives such as the GER Claude Hamilton or the GWR Bulldog?

Another thing that will need some looking into, possibly, are the journey time tolerance settings. They might currently be a little too harsh: I have eased them for the next version, so we shall see how that turns out: likewise, I have increased the power of the petrol-electric omnibus.

It is evident that there needs to be some filling in of the timeline of rail vehicles at around the era in which you are playing, on which I have already started work. Your suburban trains were using the LMS 4F and the Metropolitan Dreadnought carriages, which is an... interesting combination. I have produced an LMS 3P tank locomotive (introduced: 1930) and plan to produce some LMS suburban carriages to go with it, and will also produce an equivalent set of suburban vehicles from about 20-30 years earlier.

Incidentally, I had already increased the renovation percentage for the -devel branch in my local files, which will be updated when the Pak128.Britain-Ex for 8.0 is released. Likewise, the cost of demolishing houses has been reworked to make it more realistic. (And "Buy House" is present in Pak128.Britain-Ex for me with 8.0 - so I am not sure what problem that Neroden is having).

Offline neroden

  • Devotees (Inactive)
  • *
  • Posts: 831
  • Nathanael Nerode
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #36 on: May 09, 2010, 08:39:29 PM »
(And "Buy House" is present in Pak128.Britain-Ex for me with 8.0 - so I am not sure what problem that Neroden is having).

It's there in your build, I just can't figure out how to get it when building from source.  I'm going to try to clean things up so that "anyone can build it".  I guess I can do this in parallel to the line ending fixes.

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 18679
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #37 on: May 09, 2010, 09:32:16 PM »
Hmm, that's odd.

Offline neroden

  • Devotees (Inactive)
  • *
  • Posts: 831
  • Nathanael Nerode
Re: [Experimental release] Pak128.Britain-Ex 0.6
« Reply #38 on: May 11, 2010, 06:06:29 AM »
Hmm, that's odd.

As noted elsewhere, I found the problem while rebranching and it's fixed in the rebranch.  (A change in standard got lost during previous attempts to sort out the multiple copies of files in the repo.)