The International Simutrans Forum


Author Topic: Pak128.Britain Experimental 0.83 feedback  (Read 1896 times)

0 Members and 1 Guest are viewing this topic.

Offline jk271

  • *
  • Posts: 292
  • Languages: CZ, EN, DE
Pak128.Britain Experimental 0.83 feedback
« on: May 27, 2012, 09:20:55 PM »
I was asked for user feedback for pak 128 Britain Experimental 0.83.
So I have started new game in 1825 ...

What I appreciate on pak 128 Britain (and exp too) are rivers. Mostly the scale. Having creaks, small rivers, rivers and very big rivers is nice and realistic. I am looking forward to see program to place rivers more realisticaly. (It is my wish to program change.)
I like factories with very different amount of production - from markets with single units (or tens of of units) per month to steel mill with demand of several hunderds of tons of coal and iron ore. I admire huge time range - from late 18th century to current date - for which is pak Britain (experimental and standard) playable (as a contrary to pak128 being playable since 1920 or pak64 being playable since 1840).

On the other hand there are some parts that needs some changes:

Part 1 - graphics

(*) There is a lack of tree diversity (at least for me). I do not know tree species in forest of Britain. Currect state can be realistic as I am not able to compare it with reality. I have found 2 species of tree there: brown (oak) and green (spruce).
I do not think that this pak has too dark buildings now, but the amount of brown trees makes it to seem dark. The brown trees does not need any colour change. Two or three new (green) species need to be created.

(*) Some graphical bugs to be fixed
    Puffing Billy - Vehicle alignment needs some improvement. Vehicle is correcrly on the track but shifted forward. It is significant in corners.
    LivestockDropper - Moved/shifted forward when is going from nord to south and from west to east. Other directions seem to have correct allignment.
    SDR-Locomotion - Shifted/moved in west-to-east, south-to-north and southwest-to-northeast direction images
    LMR-Rocket - shifted in west-to-east, east-to-west, north-to-south and south-to-north direction images

    cartMailNew - shifted out of the road

Part 2 -
I have read discussion about plans to change (decrease) meters_per_tile value from 250. I support it, it is an excleent suggestion.

I suggest to increase station_coverage from 3 to 4.

Enable road toll by default.

Part 3 - roads
This is weird for me - cheepest road has better parameters than most expencive one:
dirt_road - 80t
macadam_road - 35t
cobblestone_road - 8t

Road gate is missing in 1834 so there is no way for blocking public vehicles from enty to private roads.

Offline jk271

  • *
  • Posts: 292
  • Languages: CZ, EN, DE
Re: Pak128.Britain Experimental 0.83 feedback
« Reply #1 on: May 27, 2012, 09:25:01 PM »
part 4 - input storage capacities 

Contrary to pak64 (both - standard and experimental) input storage capacity in pak 128 Britain experimental does not need huge changes. It is inherited from standard pak 128 Britain.

I have created a table showing a distance for which supplying with raw materials work without "leaps".
First column is factory object ID.
Second column is input raw material needed by factory.
Third column contains time in game months for that factory has storage of input raw material. Value in braces is time in "game hours".
There is a distance, that can be travelled by vehicle to deliver raw materials without "leaps" in the column 4.

Using 4 files per km (250 meters per tile) factories does not generally need to increase storage capacity or decrease productivity. Generally (and with 250m/tile) to do nothing with input capacity is an acceptable solution.
On the other hand factories in that 19th century had storage for greater time than 3 or 4 hours. So slight increase of an input capacity (up 2 times) makes it more realistic.

I have read about intention to decrease meters_per_tile value. If meters_per_tile was 125, I would suggest to increase input capacity by 2 or 4 or decrease productivity by 2 or 4. Both cause the same effect - increase the distance for that can be delivered input raw materials without "leaps" as with meters_per_tile=125 the distance is decreased.

factory            raw material    storage month (hours in game)    distance without leaps in tiles (km at 250m/time)

Bakery1750        flour        2     (6.4  hours)    256 tiles (64km) at 10 km/h   
Bakery1840        flour        1.5   (4.8 hours)
Bakery1910        flour        1.33  (4.26 hours)   ?? less than following one ?? --> increase input capacity
Bakery1945        flour        1.5   (4.8 hours)
Bookshop1860        Bucher        0.53  (1.7 hours) ?? less than following one ?? --> increase input capacity
Bookshop1920        Bucher        0.83  (2.66 hours) ?? less than following one ?? --> increase input capacity
Bookshop1950        Bucher        2.5   (8.0  hours)
Bookshop1970        Bucher        2.0   (6.4 hours)
Brewery1750        Bretter        8.333 (26.66 hours)
Brewery1750        grain        0.636 (2.03 hours)
Brickworks1820        clay        1     (3.2  hours)
BuildersYard1800    cement        1.2   (3.85 hours)
BuildersYard1800    Bretter        1.22  (3.92 hours)
BuildersYard1800    bricks        1.22  (3.92 hours)
BuildersYard1800    stone        1.269 (4.06 hours)
Butchery1820        meat        0.5   (1.6 hours)
Cannery1810        meat        1.04  (3.33 hours)    133 tiles (33km) at 10 km/h
Cannery1810        fruit        1     (3.2  hours)
CementWorks1800        Stone        1.2   (3.86 hours)
CementWorks1800        clay        5.88  (18.82 hours)
ChinaShop1750        china        2     (6.4 hours)
ClothesShop1750        textile        2.25  (6.98 hours)
CoalYard1750        coal        1.667 (5.33 hours)
Diary1750        milk        0.6   (1.92 hours)
FishPort1750        fresh fish    0.5   (1.6  hours)    77  tiles (19km) at 12 km/h
Fishmongers1820        fish        1.0   (3.2  hours)    153 tiles (38km) at 12 km/h
GrainFarm1750                                - two times overtransported (yield per field is too (2 time) high **1
GrainMill1750        grain        0.544 (1.74 hours)    125  (31)    OK
Greengrocers1820    canned_food    1.5   (4.8 hours)
Greengrocers1820    fruit        1     (3.2 hours)
Grocer1810        vegetable    1     (3.2 hours)
Market1750        fruit        1.25  (4.0 hours)    160 tiles (40km) at 10km/h
Market1750        vegetables    1.25  (4.0 hours)    160 tiles (40km) at 10km/h
Market1750        meat        0.6   (1.92 hours)    77 tiles (19km) at 10km/h
Market1750        fish        2.5   (8.0 hours)
Pub1750            beer        1.343 (4.3 hours)
Pub1750            cider        0.6   (1.9 hours)
Slaughterhouse1750    livestock    0.583 (1.86 hours)    45 tiles (11km) at 6km/h   
SteelMill1790        coal        1.429 (3.6 hours)    182 tiles (46km) at 10km/h   
SteelMill1790        iron ore    1.429            182 tiles (46km) at 10km/h   

**1 I used months in game as a scale (1280 tiles/month)

Not all factories are included in the table. It will take much more time to complete it.
« Last Edit: May 27, 2012, 09:38:07 PM by jk271 »

Offline jamespetts gb

  • Simutrans-Extended project coordinator
  • Moderator
  • *
  • Posts: 19345
  • Cake baker
    • Bridgewater-Brunel
  • Languages: EN
Re: Pak128.Britain Experimental 0.83 feedback
« Reply #2 on: May 27, 2012, 09:44:34 PM »
rThank you very much for your detailed feedback - that is most heplful! I am glad that my work in calibrating the factories for Pak128.Britain paid off (that was one of my contributions to the Standard pakset). May I ask - what do you mean by "leaps" here when discussing the factory calibration? I am not sure that I fully understand that. Also - perhaos I am being dim, but why does one need to adjust factory produciton or storage capacity because of a change in the meters per tile value? I don't think that this affects factory production, does it?

Some of the issues that you report (such as lack of tree diversity and the graphical glitch on one of the station extension buildings) are inherited from Standard, and should perhaps be posted in that forum (especially if there is a graphical bug).Some of the graphics issues that you report (the misalignment of the mail cart, for instance) have been fixed ready for the next release (thanks in this instance to Giuseppe), and I think that I have fixed some of the early locomotive graphics, too.

As to road tolls, may I ask why you suggest that this be enabled by default? Traditionally in the UK, users have not been charged (at least, from the mid 19th century onwards) for use of roads. The calibration of road weights is an interesting observation. Do you have any thoughts as to how you would recalibrate them? Bear in mind that maximum speed is an important consideratiion alongside weight.

The suggestion of increasing station coverage from 3 to 4 in conjunction with the  change to 125m/tile seems sensible, and is probably a more or less realisitc distance (500m), allowing for duly intricate 'bus services, which should be more of the more fun aspects of a larger scale.

The rivers, incidentally, is more of an incident of the Experimental code than of the pakset- I think that it was Inkelyad who modified the rivers code in Experimental to produce this effect.

Thank you again for the feedback - much appreciated!

Offline jk271

  • *
  • Posts: 292
  • Languages: CZ, EN, DE
Re: Pak128.Britain Experimental 0.83 feedback
« Reply #3 on: May 27, 2012, 11:03:46 PM »
I have used "leaps" for following behavior of industry chain:
Consumer has small storage capacity, so transporter is unable to deliver raw materials before consumer runs out of it.

Example with iron ore mine and steelmill:
1) Steel mill has full storage, iron ore mine stops production
2) Amount of iron ore in steelmill's storage reaches limit so iron ore mine resumes to deliver iron ore. A ship goes on voyage from iron ore mine to steel mill.
3) Steel mill has no iron ore. It stops production.
4) Ship reached steel mill.
5) Another ships reached steel mill. There is a plenty of iron ore in the steel mill's storage. Iron ore stops  production.
6) goto: 2;

So steel mill's storage jumps between "too much raw materials" and "no raw materials".

... why does one need to adjust factory production or storage capacity because of a change in the meters per tile value?

Raw materials can be delivered to factory continuously and without leaps up to some "fixed distance". Such a distance is counted in tiles.

For example bakery1750 has such distance 256 tiles (with vehicles driving at 10km/h). Using 250 m/tile the distance is 64km. So if there is a grain mill nearer than 64km to bakery, flour can be transported without leaps. For 125 meters per tile the distance is still 256 tiles but it is 32km. So input capacity needs to be doubled for 125m/tile to provide same distace in km. Same effect can be reached by decreasing productivity.

So if the productivity (generally speaking) is considered too high it can be lowered instead of increasing input storage capacity. Mostly because of "15000 limit" on input capacity. Pak 128.Britain Experimental does not need decrease of productivity of factories. Furthermore there is no space for decrease because of program constrains.

Thank you for explication of way toll in Britain. I considered it because of railway in case you enable other company to use your's rayway network. But it is impossible without extension of current simutrans code.

Weight limits: I took it as a typo mistake. I suggest to swap dirty road and cobble stone weigth limits.

I do not think, that experimental has worse rivers placement than standard. Some years ago I was working on patch making valleys for rives when new game is being created. It worked. Unfortunately it was extremely slow.

Edit: I have discovered some inconsistences today:
woodenPlatformtrainstop costs 150, more than StationStoneBuilding (125)
brickViaductEarly has weight limit 999t but brickViaduct has 136t
« Last Edit: May 28, 2012, 09:03:03 PM by jk271 »