Page 385 of 389

Re: Dutch Trainset v2.0 being developed

Posted: 03 Aug 2014 10:28
by Densha
I noticed some oddities concerning the DDM/DDAR/DDZ coaches capacity.

The capacity for DDM/DDAR used with 'auto selection mode' with either mDDM/1600/1700/1800 locos seems to be 166 and for the driving cabs 90, but when I refit the driving cab into the special driving cab livery the capacity is suddenly 124 instead of 90.

For DDZ the standard capacity is 130 and for the driving cabs 124. But with the driving cab refit it's 90 instead of 124.

Re: Dutch Trainset v2.0 being developed

Posted: 03 Aug 2014 11:45
by FooBar
Seems I put two numbers the wrong way round. It's fixed in the nightlies.

DZZ should have lower capacity in for both regular and cab.

Re: Dutch Trainset v2.0 being developed

Posted: 04 Aug 2014 08:37
by pietjeklap
Can anyone tell me if this one is usefull?

Image

Re: Dutch Trainset v2.0 being developed

Posted: 04 Aug 2014 11:02
by Transportman
pietjeklap wrote:Don't know if it's allowed but I've made a sprite based on this one:
It is allowed under the GPL-license as long as you follow the same license and give proper credits.
Image
I think it looks nice, although the blue in the horizontal views looks a bit to gray-ish compared to the diagonal views.

Re: Dutch Trainset v2.0 being developed

Posted: 04 Aug 2014 11:15
by FooBar
pietjeklap wrote:Can anyone tell me if this one is usefull?

Image
I think it's quite bright, especially the yellow. Compare for instance the ICR coaches. Also, the side of the / views is now transparent and you haven't drawn in the colours of the palette. I appreciate the effort though! If you can make it fit in with the rest of the graphics I'm more than happy to include it.

Yoshi wrote:Something like this?
FLIRT_3_NL_preview.png
dutch_Flirt.png
Looks good. If you can put it in our standard template I can code it.

Re: Dutch Trainset v2.0 being developed

Posted: 05 Aug 2014 13:00
by FooBar
And now for something completely differentâ„¢

I think the costs in this set are too low. So I'm thinking about doubling them in the default setup. Don't worry, the parameter settings to change cost factors will stay.
In addition to that, I'm pondering variable running costs. This means that the running costs will change depending on whether a train is moving or not, and suchlike. I currently have the following cases:
  • stopped inside depot
  • stopped outside depot
  • accelerating
  • coasting at speed
  • decelerating
Stopped inside depot will be the cheapest; this is where you pay for basic maintenance but not for fuel.
Stopped outside depot and decelerating will be the same and the next cheapest; you pay a bit more for maintenance but not for fuel.
Coasting at speed (this means the train is running at constant speed) will cost you the regular maintenance as well as a bit of fuel for keeping the speed.
Acceleration is the most expensive, apart from regular maintenance a lot of fuel is consumed.

If you have an opinion on this, please do vent it. Also, maybe there are more cases (of importance) than the five listed above.

Re: Dutch Trainset v2.0 being developed

Posted: 05 Aug 2014 13:23
by PikkaBird
As I said in the other thread, I only use "waiting" and "normal" running costs now (I assume by "stopped" you mean "waiting" - literally stopped vehicles have 0 running costs always).

"Decelerating" isn't really meaningful in OpenTTD. It rarely happens (half a tile when approaching a signal or a couple of tiles when entering a station) and is indistinguishable from "accelerating" to NewGRF variables.

I'm not sure "accelerating" vs "coasting" is a good idea. Since more powerful trains reach coasting speed faster than less powerful ones, it effectively reduces the range of running costs in your set (it will make more powerful locos cheaper to run and less powerful ones more expensive). It's also not a readily obvious mechanic to players, so if it's done big-time-stylee it's confusing; if it's done subtly, it's unnoticeable, and therefore pointless.

Re: Dutch Trainset v2.0 being developed

Posted: 05 Aug 2014 14:00
by Transportman
I have to agree with PikkaBird on this, it only makes sense for waiting and moving trains, all other cases are either not possible, irrelevant or too complicated.

A possible third case would depend on the current speed, low speeds are low running costs and high/full speed means high running costs, but I'm not sure if it is worth the effort to do that.

Re: Dutch Trainset v2.0 being developed

Posted: 05 Aug 2014 14:17
by FooBar
PikkaBird wrote:As I said in the other thread...
I did see you post in the other thread, after I posted here :)

By "stopped" I did mean "stopped", but also "waiting". I didn't realise that "stopped" (by user) was already free, but then it's called "running costs" and a stopped train isn't actually running.

I have to agree with the meaningfulness of "decelerating", but I do think it's distinguishable: variable vehicle_is_stopped as I understand it is 1 when decelerating or stopped/waiting and 0 otherwise. I'll have to see if the NML documentation is correct here.

I'll have to think about the "accelerating" vs "coasting". I think it makes sense that running underpowered trains is more expensive. Less powerful trains could still be cheaper when used properly as their regurlar running costs are also lower. That doesn't hold if accelerating costs 8 times as much as coasting, but only doubling may do the trick. The running cost range may be an issue though. I only have about 25% left at the top of the range.

Thanks for shedding some different light on this!

Re: Dutch Trainset v2.0 being developed

Posted: 05 Aug 2014 15:05
by Eddi
your model seems a bit crude.

let's develop some actual theory.

"running cost" can be divided into 3 categories:
  1. personnel cost: this more or less constant and needs to be paid whether the train is moving or not, but possibly not when it's waiting in depot. older engines may have this value doubled, because they need two people to operate, but this value should probably also depend on date
  2. fuel cost: this should depend on the actual power output of the engine. an accelerating engine will always output maximum power, but a "coasting" engine only needs the power to sustain its current speed. but at max speed, this will probably be almost as high as the full power. (also, "realistic" engines have a "short-term power", a "one-hour power" and a "continuous power" value, which differ. this is not modelled by the game)
  3. wear cost: this includes all sorts of replacable parts like brakes and "tyres", which are usually calculated in distance travelled, not how fast the train is going. but this still may be higher for trains that have higher max speed, as their parts require higher quality
all this may be difficult to convey to the user, however. and as pikka mentioned, it may not make any actual impact.

Re: Dutch Trainset v2.0 being developed

Posted: 06 Aug 2014 00:41
by PikkaBird
FooBar wrote:variable vehicle_is_stopped as I understand it is 1 when decelerating or stopped/waiting and 0 otherwise. I'll have to see if the NML documentation is correct here.
I strongly suspect vehicle_is_stopped is 1 only if the train is actually stopped, manually or by sending to depot ("braking for a stop" is the feature where trains in OpenTTD don't stop instantly when you press the stop button).

You can detect other states by testing the speed; a current_speed of 0 means the train is waiting or loading, and (current_max_speed - current_speed) = 0 means the train is coasting.
Transportman wrote:A possible third case would depend on the current speed, low speeds are low running costs and high/full speed means high running costs, but I'm not sure if it is worth the effort to do that.
As I recall, in NARS2 the steam engines use this model (higher costs at higher speed) and the diesels use the "coasting" model! The fact that these two opposing mechanics produce no impact on gameplay whatsoever shows what a colossal waste of time all this "realism" is. ;)

Re: Dutch Trainset v2.0 being developed

Posted: 06 Aug 2014 01:51
by ISA
PikkaBird wrote:The fact that these two opposing mechanics produce no impact on gameplay whatsoever shows what a colossal waste of time all this "realism" is. ;)
Really? for example in my current game most of the dual headed trains running cost is above 1 mil per year and some trio headed trains even 2 mil I thought if the trains for example would be in depot with no taxes or lower taxes it would be huge impact as in my game train income is 70 mil in minus! I have tried to close the cap. So far other transportation hubs holds me on float and earn profit, even tough I use money cheat one of my games!

Re: Dutch Trainset v2.0 being developed

Posted: 06 Aug 2014 03:35
by Supercheese
(code simplified for illustration - running cost factor #defined previously)

Code: Select all

switch(FEAT_ROADVEHS, SELF, running_cost_switch, current_speed) {
    /* Running cost reduced to 1/4 if vehicle is not moving. */
	0: 				 RUNNING_COST_FACTOR/4;
	return 			RUNNING_COST_FACTOR;
}
This is the route I took for variable running costs in Fake Subways. Simple, easy, and moderately "realistic".

Re: Dutch Trainset v2.0 being developed

Posted: 06 Aug 2014 11:24
by PikkaBird
ISA wrote:in my game train income is 70 mil in minus!
I would suggest turning inflation off before starting your next game; your income problems are quite unrelated to anything we're discussing here.

Re: Dutch Trainset v2.0 being developed

Posted: 14 Aug 2014 16:53
by Voyager One
As per request, additional new livery for RET B. :wink:
Rotterdam MG2-1 Type B.PNG
Rotterdam MG2-1 Type B.PNG (10.9 KiB) Viewed 3115 times

Re: Dutch Trainset v2.0 being developed

Posted: 14 Aug 2014 17:25
by FooBar
Well, that's quick! Thanks a lot, looks good. I'll add it tonight.

Re: Dutch Trainset v2.0 being developed

Posted: 14 Aug 2014 19:58
by FooBar
It's in. Also all loading times have been revisited. For most passenger vehicles they're increased a bit and a loading speed parameter has been added. This allows you to further halve or double the now-default loading speeds. But I do think the defaults are much better now. There actually was a bug with the loading speeds of multiple units, they should've been multiplied with the amount of vehicle segments. Please do test the latest push-build: http://bundles.openttdcoop.org/dutchtrains/push/LATEST/

My last game with this set made me ponder the vehicle life. That is the time when a vehicle gets old, currently 20 years for most vehicles. I found this too short, I was replacing vehicles more than anything else.
There are options three:
1) Leave it as is.
2) Increase to some new default for all vehicles, say 25 years.
3) Set the time from real life first service to last service as vehicle life. Some vehicles may get a life of 50 years or more.
I'm mostly interested in what people say about option 3. This means that the vehicle life will be equal to the time a vehicle is available for purchase. It does not take into account that many long-lasting vehicles were once or twice refurbished (this being the reason they have such a long life).

Re: Dutch Trainset v2.0 being developed

Posted: 14 Aug 2014 20:23
by BlueEagle_nl
Hi FooBar,

With "loading speeds multiplied by amount of vehicle segments", do I get it correctly that when (for example) a VIRM-3 got a loading speed of 30 pax per tick, a VIRM-4 got a loading speed of 40 pax per tick?

For vehicle life, I'd say, pick something playable. 20 years is indeed a bit short, but for some vehicles your third option (RL In service - RL final out of service) might also be very short... I recalled somewhere the NS 5000 was in service for not even one decade...

And how would you handle trains that are currently in service, or like some DDAR, being pulled back in service? Give them a life of 25 years?

I'll be testing the set somewhere this weekend, when I come up with something during playtesting, I'll let you know.

To all of the team, keep up the good work!

Re: Dutch Trainset v2.0 being developed

Posted: 14 Aug 2014 21:30
by Densha
I don't usually use the option so I'm not exactly sure, but the current situation is that when you use automatic livery refitting trains will practically be 'refurbished' (or re-liveried) at the time when that livery is introduced right?

Re: Dutch Trainset v2.0 being developed

Posted: 14 Aug 2014 21:48
by PikkaBird
FooBar wrote:I found this too short, I was replacing vehicles more than anything else.
You don't use autorenew/replace? :P
FooBar wrote: 3) Set the time from real life first service to last service as vehicle life.
This is waaaaay too realisms. The factors which may have applied to vehicle life in reality (economics, politics) don't apply at all in TTD. For players not versed in Dutch railway history, it will make the set seem frustratingly random; why did this EMU disappear after 15 years, while this other one is still around after 50?

For a "realistic" set with a lot of vehicles, I'd say a model life of around 50 years and a vehicle life of around 25 is good - perhaps a bit longer for steam engines, and a bit shorter for MUs. Setting the retire vehicle early property to around 10 years is also a good idea; it ensures people don't buy vehicles immediately before their reliability starts to deteriorate.