• We have updated our Community Code of Conduct. Please read through the new rules for the forum that are an integral part of Paradox Interactive’s User Agreement.
someone made in paradox a shitupdate this tool no more working a 3 days ago was.. DO SMTH!
Thank you for detailed instructions of how to reproduce the problem.

i'm sorry but the tooltips dont show up for me. how do i fix this?
Unsubscribe and subscribe again, but make sure you are using the latest version. It is possible to select specific (previous) version using PDX Mods website, so use in-game mod UI for managing subscriptions (it always selects the latest version).

I would like to see the ability to show only one direction at a time while setting up an intersection. That would remove the confusing amount of lines on large intersections with multiple types. Being able to toggle all, or a single direction's connections visible would be a big improvement.
I plan to introduce fully configurable color sets(including presets for colorblind), so you will be able to select colors of overlays and their opacity, but first I wanted to release road/lane priority tool (~2 weeks if everything goes according the plan).
 
  • 3Like
Reactions:
I would like to see the ability to show only one direction at a time while setting up an intersection. That would remove the confusing amount of lines on large intersections with multiple types. Being able to toggle all, or a single direction's connections visible would be a big improvement.
I agree, it would be great if the player could toggle an approach direction and the tool would highlight all movements originating from that approach and gray out or fade the rest of the intersection movements. That would make visibility a lot easier when working with complicated intersections.

Edit - just saw the response from krzychu124 in the post above. Thanks!
 
For future ideas, specific despawn disabling options for Transit vehicles would be awesome. Having full buses and trams disappear and drop hundreds of people at busy intersections is very annoying.
 
  • 2Like
Reactions:
For future ideas, specific despawn disabling options for Transit vehicles would be awesome. Having full buses and trams disappear and drop hundreds of people at busy intersections is very annoying.
Thanks, noted. I'll check what can be done, full "no-despawn" like we had in TM:PE in CS1 will be hard to achieve, but limited one might be a lot easier. The reasons are quite obvious: introducing such feature will require full understanding and future maintenance on my side (for you it's one click). I need to figure out what are exact reasons vehicles despawn and see if they could be tweaked in any significant way.
 
  • 2Like
Reactions:
Hey @krzychu124, really appreciate your work. From the repo I can tell it was an absolute bear to get working - really good job!

Since Roundabouts aren't supported by the mod yet, here's my attempt at a Turbo Roundabout made out of 4 road segments - sadly foiled by the car AI's ability to switch lanes at any time..
1716403176083.png


Given that supporting roundabouts directly is quite tricky, being able to turn off lane-switching for a road segment would be fantastic as it would enable similar functionality with these 'pseudo-roundabouts'.

All that aside, if you need any help with QA I'd be happy to do so :)
 
Last edited:
Hey @krzychu124, really appreciate your work. From the repo I can tell it was an absolute bear to get working - really good job!

Since Roundabouts aren't supported by the mod yet, here's my attempt at a Turbo Roundabout made out of 4 road segments - sadly foiled by the car AI's ability to switch lanes at any time..
View attachment 1137308

Given that supporting roundabouts directly is quite tricky, being able to turn off lane-switching for a road segment would be fantastic as it would enable similar functionality with these 'pseudo-roundabouts'.

All that aside, if you need any help with QA I'd be happy to do so :)
I don't think that I can easily turn off lane switching. I've checked potential options, since the feature was already requested on Discord (TMPE/Traffic). The easiest way is to remove(or skip adding) "AllowChange" lane flag which increases lane change penalty 5x but doesn't block it completely, vehicles will simply choose other place to change lane, but it can be very problematic when overused (applied to multiple segments(edges) in a row), since the path (unlike in CS1) doesn't tell the vehicle which lane it should use, it only tells which segment (edge) allows for following the path. The lane is selected dynamically (usually few segments(edges) ahead), which means blocking lane changes at certain lanes may lead to vehicle despawn - dynamic lane selection may select sequence of perfectly valid lanes, but because of high lane change penalty, vehicle may easily stuck at incorrect lane (when next set of lanes has to be calculated), and at worst case simply despawning vehicle because of "penalty overflow".

The green path is actually correct knowing the set of restrictions the vanilla auto lane connector has - at least one lane has to be connected with all other segments (which prevents from drawing an intersection with unreachable segments). So one going left+u-turn, one straight, one right, since it's technically 3lanes => 3x2-lane pairs (though the result will be slightly different with highway roads)
 
  • 1Like
Reactions:
Hey @krzychu124, really appreciate your work. From the repo I can tell it was an absolute bear to get working - really good job!

Since Roundabouts aren't supported by the mod yet, here's my attempt at a Turbo Roundabout made out of 4 road segments - sadly foiled by the car AI's ability to switch lanes at any time..
View attachment 1137308

Given that supporting roundabouts directly is quite tricky, being able to turn off lane-switching for a road segment would be fantastic as it would enable similar functionality with these 'pseudo-roundabouts'.

All that aside, if you need any help with QA I'd be happy to do so :)
The IA is still not optimal (to be respectful), the cims behavior is too aggressive in the usage off all lanes, and creates all type of problems. Also thanks to the path mod, you can see that cims do crazy things, choose estrange routes and make last second merging.
Probably locking one segment, will move one segment before the problem, and still not solving anything, because isn't a lane configuration problem, is how aggressive the cims changes lanes for no reason. Example: cars instead of go through, change to the deceleration lane, and at the last second merge again. This crates congestion in a perfect free straight road. And a lot of examples could be seen if you paid attention to traffic.

cities2_2024_05_21_19_13_07_574.png


cities2_2024_05_21_19_13_35_602.png
 
Example: cars instead of go through, change to the deceleration lane, and at the last second merge again. This crates congestion in a perfect free straight road. And a lot of examples could be seen if you paid attention to traffic.

View attachment 1137407

View attachment 1137408
Yes, this is a major problem in the traffic model that will cause unnecessary congestion. I work with traffic simulation professionally, and in our models we have two parameters that can be set at each connector to govern lane selection. One tells the vehicle the distance prior to their turn that it should start to change lanes/merge over to position itself in the correct lane to access the turn, the other tells vehicle the farthest point it can travel without being in the correct lane. If those two parameters are not set correctly in the professional grade simulators, then even the professional grade traffic simulators will look like the chaos we see sometimes in CS2.

@krzychu124 It would be a huge improvement if you are able give the player an option to set a distance upstream of each connector that allows vehicles to “see” the connector. Then make any vehicle within the player determined set distance to “see” the detector that is not in the connector’s lane have a desire to make lane changes to position themselves in the correct lane to access that connector. Once a vehicle is within the player determined distance and in the correct lane, it should be given a large lane change penalty so it is incentivized to stay in that lane.

You could call this setting “initiate pre-position distance” or something like that.

With a system like this, @elGendo87 could solve the unnecessary lane change problem by setting an “initiate pre-position distance” for the straight through connector to something a little longer than the off-ramp deceleration lane. If it works correctly it should keep the through traffic in the straight through lane, and allow the traffic destined for off-ramp to change lanes and exit without a problem.
 
  • 1Like
Reactions:
The IA is still not optimal (to be respectful), the cims behavior is too aggressive in the usage off all lanes, and creates all type of problems. Also thanks to the path mod, you can see that cims do crazy things, choose estrange routes and make last second merging.
Probably locking one segment, will move one segment before the problem, and still not solving anything, because isn't a lane configuration problem, is how aggressive the cims changes lanes for no reason. Example: cars instead of go through, change to the deceleration lane, and at the last second merge again. This crates congestion in a perfect free straight road. And a lot of examples could be seen if you paid attention to traffic.

View attachment 1137407

View attachment 1137408

Did you touch the selected node? I found a bug in lane connector which may generate lane connections in wrong order so (DLS) Dynamic Lane Selection (counting lane changes) may think that certain connections are "cheaper" than others. I'll try to fix it till the next update (with priority signs).
Here is how it works: network system is created from edges(segments) and lanes, if road has more than one lane going the same direction, additional lane is created called "master lane", it stores the information about how many lanes should be considered as group of lanes going the same direction. On your screenshot you can see 2 nodes and 5 edges in total. Counting from left:
1) one two way highway edge (no master lanes)
2) one two way, two lane highway (2 additional master lanes grouping 2 lanes going the same direction)
3) 3 edges (2 one-way highway for merging and diverging) + two way highway
The problem is (as mentioned in my previous post) vehicles need to check at least second node to "see" the road is diverging because their path says: follow edge x (1), switch to edge y (2 - select optimal lane from master lane "a") then switch to edge z (3 - optimal master lane "b" at node - there will be two), so for DLS middle edge by itself doesn't tell vehicle which lanes are better or worse, there are technically 2 lanes, bottom one might look very tempting if only handful of vehicles go off highway (will have significantly better traffic flow/congestion stats than the one it should actually use going straight)
 
Did you touch the selected node? I found a bug in lane connector which may generate lane connections in wrong order so (DLS) Dynamic Lane Selection (counting lane changes) may think that certain connections are "cheaper" than others. I'll try to fix it till the next update (with priority signs).
Here is how it works: network system is created from edges(segments) and lanes, if road has more than one lane going the same direction, additional lane is created called "master lane", it stores the information about how many lanes should be considered as group of lanes going the same direction. On your screenshot you can see 2 nodes and 5 edges in total. Counting from left:
1) one two way highway edge (no master lanes)
2) one two way, two lane highway (2 additional master lanes grouping 2 lanes going the same direction)
3) 3 edges (2 one-way highway for merging and diverging) + two way highway
The problem is (as mentioned in my previous post) vehicles need to check at least second node to "see" the road is diverging because their path says: follow edge x (1), switch to edge y (2 - select optimal lane from master lane "a") then switch to edge z (3 - optimal master lane "b" at node - there will be two), so for DLS middle edge by itself doesn't tell vehicle which lanes are better or worse, there are technically 2 lanes, bottom one might look very tempting if only handful of vehicles go off highway (will have significantly better traffic flow/congestion stats than the one it should actually use going straight)
I don't remember, but as the u turns and the unsafe turns are still on, i would say no, is vanilla.
I have 5 of this "T" interchanges and not always happen this. But I saw this behavior in a lot of intersections. There lane system gizmo in Dev mode shows the priority of the lanes? I Can help you with a screenshot of that if so...

Cheers
 
I don't remember, but as the u turns and the unsafe turns are still on, i would say no, is vanilla.
I have 5 of this "T" interchanges and not always happen this. But I saw this behavior in a lot of intersections. There lane system gizmo in Dev mode shows the priority of the lanes? I Can help you with a screenshot of that if so...

Cheers

In dev mode lane gizmo you can see lane priority in sense of which lane yields at next intersection. I don't think there is a view/gizmo showing lane traffic flow, though that data is collected (I'll check if I could add it maybe as Traffic debug gizmo). You can enable navigation gizmo and follow selected vehicle to see what it's doing (pink line will appear to show willingness of lane change and place where it aims to change it).
I haven't delved too deep what colors mean (yet). I'll focus on such tweaks and improvements once I release priority tool (unless devs do that earlier)
 
  • 1Like
Reactions:
In dev mode lane gizmo you can see lane priority in sense of which lane yields at next intersection. I don't think there is a view/gizmo showing lane traffic flow, though that data is collected (I'll check if I could add it maybe as Traffic debug gizmo). You can enable navigation gizmo and follow selected vehicle to see what it's doing (pink line will appear to show willingness of lane change and place where it aims to change it).
I haven't delved too deep what colors mean (yet). I'll focus on such tweaks and improvements once I release priority tool (unless devs do that earlier)
cities2_2024_05_23_15_02_16_404.png
cities2_2024_05_23_15_01_05_316.png
cities2_2024_05_23_15_00_58_656.png


Look at this guy... Plenty of time and space to change lane, and did it in the last second. They are bad drivers... LOL (with path gizmo activated)
 
The IA is still not optimal (to be respectful), the cims behavior is too aggressive in the usage off all lanes, and creates all type of problems. Also thanks to the path mod, you can see that cims do crazy things, choose estrange routes and make last second merging.
Probably locking one segment, will move one segment before the problem, and still not solving anything, because isn't a lane configuration problem, is how aggressive the cims changes lanes for no reason. Example: cars instead of go through, change to the deceleration lane, and at the last second merge again. This crates congestion in a perfect free straight road. And a lot of examples could be seen if you paid attention to traffic.

View attachment 1137407

View attachment 1137408
I did some experimenting with highway segments. It looks like the best solution is to keep the deceleration lane as short as possible and use the dashed connector where the highway tapers out from one lane to two lanes. With longer deceleration lanes it seems like some of the vehicles can’t “see” far enough to realize there is an off-ramp at the end.
 
  • 1
Reactions:
I did some experimenting with highway segments. It looks like the best solution is to keep the deceleration lane as short as possible and use the dashed connector where the highway tapers out from one lane to two lanes. With longer deceleration lanes it seems like some of the vehicles can’t “see” far enough to realize there is an off-ramp at the end.
Hi, i was testing your idea and taking measures. At 80 meters the ramp conserve a shorter gap between dash marking, and cims apparently see as a ramp. At 100 meters, the markings changes and a longer stripe appear, but almost every cim still don't use it as a lane. At 120 meters, more cims start to switch between lanes.
I need to do more testing with a denser traffic to find the sweet spot of ramp length :D
Using the "direct" lane connection is worst, so it should remind as unsafe lane change, with the dashed connector as you said.

EDIT:
@PTOE Plays i answered in a new thread.
Sorry @krzychu124 from deviating this thread.
 
Last edited:
  • 1Like
Reactions: