News:

Simutrans Forum Archive
A complete record of the old Simutrans Forum.

General pak-britain v1 feedback

Started by AP, July 25, 2009, 05:57:15 PM

Previous topic - Next topic

0 Members and 2 Guests are viewing this topic.

AP

I've been keeping a notepad of random feedback whilst playing - just tidied my desk, so, in no particular order:

- The bridge length thing needs resolving (well in hand it seems)

- The lack of passenger barges, ferries and ships is a real irritant since that is a very good and realistic way to expand passenger networks early, especially since we now have navigable rivers. Suggest they be prioritised even if the rest of the canal vehicles aren't; I ended up importing the generic pak128 boats in the end.

- The absence of mail vehicles pre-combustion engine is a pain since mail-and-passenger work as one network; the dual income is part of what makes it so efficient.

- I can't quite find an advantage of horse-drawn trams over horse-drawn carriages. They are the same speed, practically same capacity, but can't pass each other on a single tram line (obviously), so actually seem worse. And if you're using horses, clearly congestion isn't a problem. So I can't quite work out why I should pay for them.

- Given the aforementioned trouble with trams (horses not being powerful enough to make them useful, I wonder if there is scope for a cable-street-railway of some kind. Thinking London & Blackwall (1840) evolving into something like the San Francisco system (1873), which could handle many more passengers at speed. Make it out like a special kind of electrification, say...

- Can horse-drawn carriages carry more than one type of freight  - passengers + mail? ditto ferries when they appear. It would seem more realisitic for some routes

- Is there any way to set up a "carries anything" ferry? Was thinking of something to substitute for a rail ferry, which is quite a common (historically) situation for extending a rail network along or across bodies of water. In some uk situations, key ferries were later replaced by tunnels (R. Severn) or bridges (Firth of Forth) once the technology allowed. But since simutrans can't "carry" one vehicle on another, a ferry that would move whatever it was given would be a close approximation. I believe the Severn Ferry, though authorised to be a train ferry, was actually just of the conventional kind, although I know of at least one other UK precedent for a ferry carrying rail carriages (Langstone-St Helens IOW), as well as continental ones.

- With the pak-britain factories, the 1x1 "shops", and possibly the farms too, could they be made like oil rigs, in that vehicles could call at them directly, rather than requiring freight yards at each one? Might be tidier...

That's it for the moment...

jamespetts

AP,

thank you very much for the feedback: it is most useful :-) A number of responses:


  • This will be resolved in 1.03.
  • The trouble is that we only have one person drawing most of the graphics. These are planned, but they may take some time. If anyone would like to help (or recruit others to help), this would be very, very welcome!
  • As above - road goods vehicles are next on the list of priorities, I am told.
  • Hmm, this will need looking into. In real life, horse drawn trams could carry more passengers, since the horses could pull a heavier weight on the smoother tracks. That was the reason for having them. Can anyone with historical knowledge of tram and horse omnibus capacities assist here as to the figures?
  • An interesting idea, but that might be some way off given the lack of graphics producing capacity. I am not sure what the official maintainers would make of this idea.
  • No single vehicle in Simutrans can carry more than one type of load. That is a limitation with Simutrans, not the pakset.
  • See above.
  • Only oil rigs have this capability in Simutrans - it is built into the code.

Thank you very much for testing Pak128.Britain and posting your feedback. It is most useful. Do let us know if you would like to contribute to the project (anything from graphics to research to balancing to posting screenshots would be appreciated).
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

AP

Another thing I've discovered:

Bridge pricing and terrain-alteration pricing don't seem to tie in with one another:

Given that simutrans is mostly flat terrain, most bridges being just one level above what they are spanning, it is almost always cheaper to alter the terrain to make embankments and cuttings ($-1500/tile) than make a viaduct (($-1900/tile plus maintenance)). This somewhat links in with the bridge thread, but I have just built a railway through some rather hilly terrain, and found myself building precisely one viaduct, which was a bit of a suprise. Perhaps this is intended, however.

jamespetts

I'm not sure exactly what The Hood's intention was with this - perhaps it has not been balanced fully. That would rather suggest that the cost of embankments/cuttings are far too low: in reality, certainly, they are very expensive. Any other thoughts on embankment height?
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

AP

Well now, there is a logic to low embankments being cheaper than bridges, but high embankments costing more (which they do if you build the sloping-sided ones rather than the brick-sided ones because of the larger base area - realistic).

I think the problem is twofold-

1 - Simutrans is rather flat, so it's quite rare to need tall embankments, so you don't often actually need bridges.

2 - The "lower single square tile" (LSST) tool is priced to allow it to "undercut" bridges - or at least it's too cheap  - it costs ($1500 +$1500) to raise a tile +2 with brick sides, which is essentially a bridge but without any maintenance - so well worth doing instead of the $1900 bridge tile. To do that in "sloping sided" embankments costs ($5000+$2500) - which is far more realistic. If a "cheap" style embankment cost $7500 per tile, the slimmer, urban-type brick sided embankment which the "LSST" tool creates ought to be more than that by a fair margin. I think the "raise single square tile" tool needs to be about $75000 a go, or better, for the +1 to be $1500 and the +2 move to be $15000.

mjhn

I agree that the single square tool should increase in cost for higher embankments (I would want to take it further with partial square rais or low from or to a slope being $1500, and full square from a flat tile to a flat tile being $4500 for the first level and a higher amount for the second).

jamespetts

The code limits what can be differentiated. There are only two settings available in simuconf.tab:

cost_alter_land=1250
cost_set_slope=1500

The first one is for the raise and lower land tools. The second one is for all artificial slope tools. Given those limitations, any suggestions on how they be priced?
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

kierongreen

My original idea was for terrain altering to cost a fortune - at least 10 times what it is at the moment. Hence price of bridges and tunnels. I really want to force people to consider windy terraine following tracks and roads as an option due to cost rather than just building straight flat lines (which is what I end up doing in pak128).

AP

Quotecost_alter_land=1250
cost_set_slope=1500
That's a pain! It can't distinguish between setting a "sloping" tile (ie flipping a gradient) and setting a "flat" tile (ie raising land)!

Still, I think in most situations you can do most everything with alter_land, with a couple of extra clicks, unless you're in tight urban situations (when you pay the premium...) So I suggest:


cost_alter_land=1250
cost_set_slope=10000

alter_land can acheive any height, the cost just goes up by the square because of the pyramid footprint involved. Which is roughly as it should be, and does indeed force circuitous routes to be taken to avoid levels changes.

It's set_slope that I think should be priced out of use for all but the most essential works (ie urban embankments - which only a rich railway builds anyway)

And then we need to remove the bridge length limits, of course, so they get more use (in hand I think).

jamespetts

Kieron,

ahh, thank you for the insight. 10x may be slightly too much (earthworks to remove a small piece of errant terrain should not be too pricy), but I see AP's point that the set_slope price could be higher because it is a more flexible tool.

How about:

cost_alter_land=2500
cost_set_slope=5000


? In reality, many railways were built with substantial embankments, and this ought not be excessively expensive.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

kierongreen

See how those work in practice maybe...

The Hood

I think bridges / tunnels should be cheaper when spanning a height difference of 2, but not for 1, otherwise embankments/cuttings should be cheaper.  But it's really a case of try it and see.  Feedback welcome for inclusion in the main releases...

AP

QuoteI think bridges / tunnels should be cheaper when spanning a height difference of 2, but not for 1, otherwise embankments/cuttings should be cheaper.

I agree.

Quotecost_alter_land=2500
cost_set_slope=5000
No, they're too close.

a +1 embankment requires 2 clicks with alter_land, so (alter_land< (brick bridge)/2) is the requirement.

Brick bridge is $1900/tile, so alter_land = $750 would work (embankment then being $1500/tile)

QuoteIn reality, many railways were built with substantial embankments, and this ought not be excessively expensive.
Yes, but almost always they were triangular in section (large base, narrow top), ie done with the alter_land tool.

if
alter_land is $750     ....    then
cost_set_slope=$5000     ....     would perhaps be okay. (that's x6.6 rather than x10)

Thoughts?



jamespetts

If the artificial slope was much, much more expensive than alter land, people would only ever use the artificial slope in the very limited circumstances in which the alter land tool could not be used. An excessive disparity between the two should be avoided: there is no grounding in reality for any such disparity.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

AP

Quotepeople would only ever use the artificial slope in the very limited circumstances in which the alter land tool could not be used
That was the effect I was going for. Not sure I understand why it is a problem.

Quotethere is no grounding in reality for any such disparity.
Sure there is. If you build a mound of earth, it has sloping sides. All you need is an army of labourers and some shovels.

To make a tower of earth with straight sides is far more involved. To do that takes complex engineering and requires retaining walls, careful design and structural work, foundations, etc. You would only do it where land were at a premium - or you would use an urban viaduct instead (with shops under the arches etc).

The key thing is that it must be cheaper to build a x2 high triangular embankment than a x2 high shear embankment.


A +2 high triangular embankment, once you have the starting point, costs $7500(=$5000+$2500) per tile at present (cost_alter_land=1250), so if cost_alter_land=750, becomes $4500/tile. So at a bare minumum cost_set_slope cannot be less than $2500 (giving $5000 per tile for 2+) and I would in fact argue that to preserve the "slimline" embankment for urban contexts rather than having it used everywhere, having it cost $10000 per 2+ sheer embankment tile (vs $4500 for 2+ triangular embankment tile) is a pretty fair balance.

jamespetts

AP,

that is a very helpful and well-reasoned response: thank you. I hadn't thought of it like that. Would it perhaps be better to have cost_alter_land a little higher, though, to discourage overly excessive earthworks?
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

kierongreen

$750 is in my opinion far too cheap a cost to alter land - though I see reasoning. Maybe bridge costs need to be put up more....

AP

#17
Quote$750 is in my opinion far too cheap a cost to alter land - though I see reasoning. Maybe bridge costs need to be put up more
Certainly that is an option.

I think the key thing is that the three variables are in the correct relative balance. Change the price of one and the others must alter to keep the relationship similar, so that the correct ones are the "best value" in the intended circumstances.

If the 750-1900-5000 is about the right relative balance*, then you can clearly just multiply all of them by the same multiplier, it will still balance the same.

*To summarise:
cost_alter_land=750, cost_set_slope=5000, cost for one tile of brick viaduct=1900      ...means...
+1 slope embankment=$1500
+1 viaduct = $1900
+1 shear embankment=$5000
+2 viaduct = $1900
+2 slope embankment = $4500
+2 shear embankment = $10000
(all per tile)

QuoteMaybe bridge costs need to be put up more
On which logic $1500<bridge tile<$4500, and you also have to remember that bridges incur maintenance so you can't set it too close to the $4500/tile mark or it's not worthwhile building them.

[can of worms]Of course, this then links in with the earlier discussion about varying bridge cost with height[/can of worms]

kierongreen

Or maybe it's ok for bridges to be cheaper to build than embankments - keeping in mind that they'll be more expensive in the long term. In which case there's no problem keeping bridge at $1900 and putting embankments up further...

Maragil

As yet anyway, the 2 bridges we have aren't brilliant :D

H./

jamespetts

Quote from: Maragil on July 30, 2009, 08:20:24 AM
As yet anyway, the 2 bridges we have aren't brilliant :D

The bridges themselves are good - we just need more of them :-)
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

wlindley

1. All the various passenger carriages are confusing for us Yankees who don't quite suss the terminology.  Perhaps a little overview of what all the acronyms mean -- I start my games in the the 1920's and presume these are for the various railroads?

2. In the 1920s there is a very limited number of houses and buildings available, so the cities look remarkably monotonic.

3. Am really waiting for mail trucks for a balanced transport system.

4. Overall the gameplay is good and I really love the "look" of the pak graphics.  And without the irritating foolishness of the regular pak128 descriptions.

The Hood

wlindley,

1. Sorry for the confusion - you are right though, they are the railway companies that built them.  If you want to go for historical realism, build LNER locos with LNER carriages (LNER = London and North Eastern Railway) for example.  It may be a good idea to produce a guide for the vehicles at some point though, as there are a lot of them.  Any offers? :)

2. Same problem in every era.  Unfortunately there are more pressing things...

3. ...Like this

4. Thanks!

sojo

I have the pakset tested, yesterday with a simutrans-nightly. Not with simutrans-experim.

It looks very very nice. Vehilces and stations are very good.

But I have seen that some buildings (stations) have light-pixels.
"English is a easy language. But not for me." ;) sojo

follow simutrans_de on Twitter
- A home for Simutrans (in german)

The Hood

@sojo

When you say light-pixels do you mean lights that come on in night view? I know some of these need fixing, but thanks for the reminder...

sojo

Yes, I mean walls with lightning pixels.
"English is a easy language. But not for me." ;) sojo

follow simutrans_de on Twitter
- A home for Simutrans (in german)

wlindley

The #1 issue making it difficult to lay roads and track is that the "grid" color hardly shows up against the transparent trees, and Underground mode is seemingly black-grid-lines-on-black-background... perhaps red and yellow grids more like standard pak64?

The Hood

Sounds like a good suggestion to me.  Does anyone know how to change it?  I will investigate and perhaps do a poll with some options once I've figured it out.

jamespetts

Hmm, I quite like the subdued colour...
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

wlindley

Quote from: wlindley on August 13, 2009, 11:47:46 PM
In the 1920s there is a very limited number of houses and buildings available, so the cities look remarkably monotonic.

Started a game in 1830 and the cities look a lot nicer.  Maybe some of the "old" buildings could have their "built until" dates extended into the early 1900s? 

Similarly, Textile Mills are only built until 1840 -- so unless you start a game *very* early, you will never have one... even though surely there were already-extant mills operating much later.  Lowell, Massachusetts was founded in the mid-1800s and had textile mills until the 1980s.

Maybe there needs to be some way of having "old" buildings and industries built as part of a new game even if they will not appear as new during gameplay?

jamespetts

Wlindley,

there was an extension request at one stage for older buildings to be built on map generation, to simulate the fact that the cities would have been around for a long time and would not just have recent buildings. I think that Prissi said that was on his to-do list somewhere.

As to industries, note that there are lots of different versions of the same industry, with different introduction and retirement dates, and different capacities. It is only the early type textile mill that is built until 1840 - after that, a new type comes along.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

The Hood

There are several different textile mills, with increasing production over time.  A textile mill can be created right through the timeline...

wlindley

The Pharmaceutical Factory occurs starting in 1860 but requires Chemicals which are not available until the Oil Refinery appears in 1909...?

I see that only the latest (1972) Paper Mill requires Chemicals, with the older ones only requiring Woodchips... perhaps the Pharmaceutical Factory can do likewise?

jamespetts

Ahh - the reason for that was that it was (and, I think, still is eventually) intended to have gasworks, which take an input of coal and produce an output of chemicals (albeit in a far lower proportion than coal, since most of it is turned to gas, which does not need transporting in the same way). I think that I had originally set the earlier pharmaceutical factories to accept coal, but this seemed a bit odd. You're not any good at using Blender by any chance, are you...?
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

The Hood

Thanks for spotting this.  I will include a fix in the next release that works around this problem until some more industries are drawn (contributions welcome, as always...)