News:

Simutrans.com Portal
Our Simutrans site. You can find everything about Simutrans from here.

Tram considerations influenced by Pak128.Britain-ex and pak128 graphics

Started by Mariculous, January 21, 2020, 10:57:41 PM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

Mariculous

I guess tram improvementshave been discussed quite a lot, most of them being "please allow bi-directional service". I want to discuss a different approach here.
Note that I will simply call any rail vehicle that can use road-bound rails a tram.

So What is the situation?
Trams have for ages been implemented in Simutrans with all of their current strengths and weaknesses.

Their pros are generally quite large capacities, the trams itself are relatively inexpensive to operate, they are not affected by congestion and they can make use of normal rails.

Cons are generally quite expensive ways and the inability of bi directional operation (at the same time) at a single tile.

Additionally, depending on the pakset and the point in time, trams can be a rather quick mode of transport (generally in pak128 at any time) or a rather slow on roads (Pak128.Britain-Ex any time after ~1930)

The issue is that the largest disadvantage of not being able to operate bidirectionally on the same road is quite unrealistic, whilst their biggest advantage of not being affected, nor themselves affecting road feels is even more unrealistic.
Additionally their strength of using normal rails is most often practically impossible for lack of space.

Thus, some call them overpowered (compared to the real world) due to not being affected by traffic, whilst others call them too weak (due to not being able to travel on roads bidirectionally and consuming even more space on their own tracks)

Sadly, we can not solve the bidirectional issue without very large changes to simutrans implementation of trams.
Affecting trams by private cars and vice versa, thus removing their biggest advantage, should not be done without adding another advantage to trams.

In the real world it is quite usual to build tram tracks in between the two directions of large roads.
In this case, trams will be unaffected by private cars on the road but will for sure affect private cars attempting to cross the tracks at a crossing.
e.g. like here: https://qimby.net/cache/ccfb8bb9/av6076b5281dae8092d08.jpeg

Given a graphics scale of 30m/tile in length, divided by the width and height factor of 1,25, We get a street width of 24m. Given roughly 4m sidewalks on either side there are 16m left. In reality this is slightly too few for one road lane per direction (~3m each) and bidirectional tram tracks in between (~11m), however it is absolutely enough for two road lanes and one direction of tram tracks.
I am aware that we can't move the road itself to one side "outer" side of the road and the tram track to the inner, at least not without a huge code project, but simmulating this kind of track layout by a single track in between the lanes would be a good compromise imho.

So what would this mean technically?
We need two different types of tram tracks that can be placed "on roads"
- A road bound type, which is visually the same as current tram tracky in Pak128.Britain-Ex, where trams technically get occupied by road traffic in driving direction or at least at crossings.
- A dedicated tram track, which visually differs from the road. Technically this type would ensure, that trams don't get occupied by road traffic e.g. by slightly adjusted level crossing code.

See the attachment for an example taken from pak128. From right to left there are road bound tracks, road bound with dedicated lane (thus unoccupied but still slow), grass tracks (unoccupied and slightly faster) and another fast type.

This would allow for fast 2nd gen tram service within cities without the requirement for dedicated tracks within cities, keeping in mind the space requirements for such.
It is much more practical to get two adjacent or nearby parallel roads equipped with such a type of fast track than building an entire track somewhere within city borders as this will most often require scaping whole districts.

jamespetts

This is an interesting thought, but I am not convinced by the utility of it. What appears to be shown is tram tracks with a single track down the middle and a single carriageway of road on either side, which is quite different to what is shown in the picture to which you link. I am not aware of this arrangement existing anywhere in the world, whereas, although double track tram lines on roads are more common, a single track in the centre of a bidirectional road is within the bounds of possibility.

From a Pak128.Britain-Ex point of view, the current speed restrictions on trams are entirely realistic: trams that were in roads were in fact limited in speed until later times (i.e. the 1990s, I believe, when second generation trams were introduced).

It is hard to see why, if the speeds were faster for the type of integral segregated roads that you propose would be faster, any player would ever have an incentive to build any other type of tram track. In reality, segregated tram tracks in the UK were usually totally separate from the road, rather than in the middle of it, which is what is more closely simulated by the current arrangement.

Also, you refer to tram traffic being affected by road traffic: it is likely that implementing this would be a huge project in itself, as currently the two types of vehicles are on completely separate waytypes with their own rules and data structures that are totally independent from and incompatible with each other.

Additionally, I am not sure that I understand how the centre lane system would work for crossing private cars: if this were real, private cars would not be able to cross at this point, and thus intersections on this type of tram track would have to be prohibited or direction restricted to only left turns; but this would make the coding necessary even more enormous.

I understand the issues with trams, but I am afraid that I do not think that this is a solution to them. Thank you, however, for your careful thoughts on the matter.
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.

Rollmaterial

I think one thing to bear in mind would be that the reason cities nowadays have wide avenues is precisely that districts have been scaped. Paris is a prime example of this: the city's large avenues were built under Napoleon III to make them more difficult for rioters to barricade.

Mariculous

Quote from: jamespetts on January 22, 2020, 12:09:31 AMIt is hard to see why, if the speeds were faster for the type of integral segregated roads that you propose would be faster, any player would ever have an incentive to build any other type of tram track. In reality, segregated tram tracks in the UK were usually totally separate from the road, rather than in the middle of it, which is what is more closely simulated by the current arrangement.
Because they were cheaper in construction and maintainance and the others won't be available early anyway.

Quote from: jamespetts on January 22, 2020, 12:09:31 AMAlso, you refer to tram traffic being affected by road traffic: it is likely that implementing this would be a huge project in itself, as currently the two types of vehicles are on completely separate waytypes with their own rules and data structures that are totally independent from and incompatible with each other.
Trains reservations can already occupy roads using level crossings. With some adjustments and automatically placing these on any tram track with road crossing tile, this should be sufficient for the unoccupied method.

The only "real" solution to trams were bi-directional tracks, either on roads or on their own but to maintain their rail nature this should also be compatible to normal rails, so this would be just like re-implementing trams from scratch.
Implementing trams as trolleybussed may also work well but won't allow for later light rail services.

This one whereas would reduce the space issues within cities without reimplementing the whole thing.

Quote from: jamespetts on January 22, 2020, 12:09:31 AMAdditionally, I am not sure that I understand how the centre lane system would work for crossing private cars: if this were real, private cars would not be able to cross at this point, and thus intersections on this type of tram track would have to be prohibited or direction restricted to only left turns; but this would make the coding necessary even more enormous.
To be honest I had a level crossing like reservation in mind, so nobody can enter the crossing at that time. However, allowing left turns and straight cars parallel to the track would indeed be more realistic.

jamespetts

I do not think that one could practically use level crossing logic - it would have bizarre consequences (e.g. as to left turns, etc.) and would require the data structures currently existing only where there are level crossings to exist on all tiles of road with tram tracks, which would be very cumbersome and complex to implement.

Also, are there any data about these types of tram tracks, their costs and introduction dates? The proposed economics of this do not seem to be very reality based given my knowledge of UK tramways at least.
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.

Matthew

I had a little thought that may or may not be helpful, as I don't fully understand Freahk's proposal yet. It is not the perfect solution for trams, but it gives a small improvement for a medium-sized (as opposed to huge) coding effort. A step in the right direction, but not a revolution.

I believe THLeaderH's OTRP patch is intended to make it possible to have real dual-carriageway roads (that is, traffic on one line of tiles travels in one direction and traffic on a parallel line of tiles travels in the opposite direction). If the overall aim is to have the functionality of bidirectional trams, the OTRP would make a limited implementation possible. You would build a dual-carriage road (Rollmaterial's wide avenue) and then have one tramline on each side.

I know we can build parallel roads for such parallel trams at the moment, but the player spends money and badly damages the city's economy for a relatively minor benefit. Building an OTRP dual carriageway also creates faster traffic flows for private carriages, omnibuses, etc., so the cost/benefit ratio is better for the player (and the poor Simutransians whose homes are torn down!).

I know that reincorporating the OTRP patch would itself be a major piece of work and was rejected from both Standard and Extended for code quality reasons. But IMHO it's a more achievable aim than re-writing all the tram code, since some of the code has already been written. And it might mean that the Japanese playerbase for OTRP could be reunited with Extended, which would increase the number of players, artists, and coders in the long-term. But of course, this 'bright'(??) idea comes with the caveats that I am unable to carry out this work myself, and I don't even know enough to reliably estimate the scale of the work required.
(Signature being tested) If you enjoy playing Simutrans, then you might also enjoy watching Japan Railway Journal
Available in English and simplified Chinese
如果您喜欢玩Simutrans的话,那么说不定就想看《日本铁路之旅》(英语也有简体中文字幕)。

Ranran(retired)

At present the road is behaving like a four lane with overtaking mode.
So I hope that something like a four-lane road could be implemented.
I prefer to link the appearance and functionality rather. In other words, the road determines whether or not it is possible to pass from the beginning instead of changing the attribute of the road later. Generally, the road that cannot be overtaken is narrow, and the road that can be overtaken is wide or has many lanes.

Road vehicles can only enter one vehicle in one direction, so citycars are more likely to cause traffic congestion than in reality.
Roads are chaotic because no traffic lights are automatically installed at intersections.
An increase in lanes leads to an increase in capacity. I think it is difficult to implement. Yes, this is a selfish delusion. I just think it would be nice if that could be achieved.  :-*


Quote from: Freahk on January 21, 2020, 10:57:41 PMAdditionally their strength of using normal rails is most often practically impossible for lack of space.
I strongly agree.
In simutrans, the roads are very large relative to the building, so the two roads side by side will have a very high occupancy of the city.
Double track trams can be seen in many places in the real world (unfortunately Japan has fewer trams than Europe), but to operate it with simutrans would have to severely destroy the city's buildings, which would be costly in the center of the city, and the removal of buildings would make loss passengers demand there.
There are many passengers, so I want to make a double track line, but I have to lose much demand. (´・ω・`)
Road expansion will require unrealistic costs if buildings are priced properly.
ひめしという日本人が開発者達の助言を無視して自分好みの機能をextendedに"強引に"実装し、
コードをぐちゃぐちゃにしてメンテナンスを困難にし(とりわけ道路と建物関連)、
挙句にバグを大量に埋め込み、それを知らんぷりして放置し(隠居するなどと言って)別のところに逃げ隠れて自分のフォーク(OTRP)は開発を続けている
その事実と彼の無責任さに日本人プレイヤーは目を向けるべき。らんらんはそれでやる気をなくした(´・ω・`)
他人の振り見て我が振り直せ。ひめしのようにならないために、らんらんが生み出したバグや問題は自分で修正しなくちゃね(´・ω・`)

jamespetts

An improvement of the road code to allow multiple lanes of traffic in one direction consistently would be good indeed; but I anticipate this being a lot of work to do properly. Likewise, it would be good for road vehicles not to be confined to one per tile in each direction, but rather be able to occupy fragments of a tile based on their length (as in the .dat files), but, again, I suspect that this would be a lot of work.
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.

Octavius

Quote from: Freahk on January 21, 2020, 10:57:41 PMGiven a graphics scale of 30m/tile in length, divided by the width and height factor of 1,25, We get a street width of 24m. Given roughly 4m sidewalks on either side there are 16m left. In reality this is slightly too few for one road lane per direction (~3m each) and bidirectional tram tracks in between (~11m), however it is absolutely enough for two road lanes and one direction of tram tracks.
It would be more than enough for bidirectional tram tracks. I just spotted an example in Rotterdam with, in 22 metres, 2 footpaths (2m each), 2 bikepaths (2m each), 2 car lanes (4m each) and 2 tram tracks (3m each). Trams are narrower than railway trains. Most roads with bidirectional tram tracks are wider (about 35m), but have additional car lanes, parking lanes, bidirectional bikepaths on both sides or room for platforms. Another common trick is to combine tram tracks with bus lanes.

BTW, trams have been in continuous use there since 1879 and have been continuously upgraded, so generations don't make sense. But I understand that for British trams it makes sense to talk about two generations.

Mariculous

I had taken these dimensions from from Google maps in Essen and Kassel in multiple places but all of these had ~10-12m in between the road directions. However, it seems these had a quite huge spacing in between the road and the rail and also in between the rails themselves.
I had yet located some places where they indeed got two tracks in between road directions at 6-8 m and the whole road including sidewalks, one road lane per direction and two tracks was indeed below 24m.
At a 1m gauge tram track in Oberhausen i could even measure less than 24m for a sidewalk on either side, two lanes per direction and two tram tracks.
However, it feels like two trams, and 4 busses would not fit onto a sigle tile would they? Thus I guess their scale is larger. than 24m per tile in width.