Topic: [Bug] Bus wouldn't leave the busstop. Accelerating (smoke) but speed stays 0.

[Erik]

Each time I'm running the game somewhere a bus will get stuck. The bus accelerates (smoke) but goes nowhere.
It is each time on one busstop on top of a slope. If I change the road around it. (Get the slope further of the busstop.) It will happen again by a other busstop with stands also on top of a slope.

After saving the game and reloading it, the bus will leave, but after a while a other bus on the same busstop has the same.

Here a save game.
Bug staying bus.sve
The current busstop with the problems is "Vlissingen Kolonie Zuid bushalte". (on the middle of the screen.)

I have used pak128  2.0.0 and two add-ons.
(The add-ons are on the attachment.)

I run it with 10.10 Linux 64bit.

The settings inside the game are 100 m/tile and stations has 4 blocks radius.
(With the 100m/tile the breaking of the trains looks quite nice.)

[jamespetts]

I've just noticed that I haven't replied to this topic - apologies. I have asked Bernd Gabriel, who developed the physics code to look into it. He has not had time to look into this fully yet, but he thinks that it might be related to a known issue whereby a vehicle that does not accelerate to at least 1m/s within the first calculation cycle can become stuck. The 'bus might be underpowered - you could increase the global power factor percent in simuconf.tab as a quick fix for this for the time being - even a small increase might well fix this.

Sorry that I haven't responded sooner, and sorry that you've had trouble.

[Erik]

I'm happy that the issue is known. I think you are on the right way of reason of the issue.
A bus with a stonger engine seems not to have the problem. (By that the problem solves itself bij playing along.  )

[greenling]

Erik
Can You Please Replly The Scoure your Pakfiles how you devlopemd?

[Erik]

I'm started to try to solve this.

I try it by running and analyse the game with gdb. But I have to relearn how to use that.
So it will take a while.

@Greenling: Sorry but I totally don't understand what you try to ask.

[jamespetts]

Erik - thank you for looking into this. Do let me know how you get on!

[greenling]

Erik i Have be ask after the scoure for this Pakfiles:

ns-sein.pak (13.29 KB - runtergeladen 9 Mal.)
ns-einde-keuze-sein.pak (4.68 KB - runtergeladen 21 Mal.)

[Erik]

Yes, now I understand the question.
Here the topic about this signal project: http://forum.simutrans.com/index.php?topic=6239.0 (Is in Dutch)

[greenling]

Erik
Thanks you!

[Bernd Gabriel]

Erik, I tried to reproduce the the issue at last.  In my devel branch the busses didn't stuck but were teleported.

I assumed that both stucking and teleporting had the same reason namely the still persistent drag of the passed slope.
I changed the behaviour to creeping with at least 4 km/h until the drag effect decreases.

[Milko]

Hello Bernd

From github: "Fixed: This is the end of the teleporting era. " 

Thank's

Giuseppe

[Erik]

Thanks guy's.

With gdb I didn't has much success.
My programming skills are still not too well, but I'm learning.

I haven never seen a teleporting bus but because it happens on the same area it could be close to this bug.
From with version has the Teleport fix been implemented? Perhaps it has also effect this bug.

[HeinBloed]

From github: "Fixed: This is the end of the teleporting era. " 

This got me quite excited, but apparently it's referring to a different problem than the old XP/release library bug. That one isn't fixed unfortunately, just tested it.