Rigs-of-rods: Broken/malfunctioning vehicles (aka 'parser issues')

[tritonas00]

[EDIT by _only_a_ptr_] See up-to date TODO list

[ulteq]

Re-tested with the new truck parser: https://github.com/RigsOfRods/rigs-of-rods/pull/1137

Edit: All of the remaining issues seem to be caused by invalid truck file entries.

20:10:05: Cache: Parsing vehicle 'Audi80Quattro.truck' yielded following messages:

#FFFF00 WARNING #FFFFFF (Section cinecam)
    Line (4253): position_nodes
    Message: Not enough arguments, 11 required, got 1. Skipping line.
#FFFF00 WARNING #FFFFFF (Section cinecam)
    Line (4254): include 8-21
    Message: Not enough arguments, 11 required, got 2. Skipping line.
#FFFF00 WARNING #FFFFFF (Section cinecam)
    Line (4255): exclude 0-7, 22-290
    Message: Not enough arguments, 11 required, got 3. Skipping line.

20:10:07: Cache: Parsing vehicle 'BoxSamurai.truck' yielded following messages:

WARNING (Section meshwheels)
    Line (# 883): -6
    Message: Encountered node with illegal negative number: '-6', parsing as positive '6' for backwards compatibility. Please fix as soon as possible.
ERROR (Section axles)
    Line (# 888): 3 6 7 1
    Message: Invalid property, ignoring whole line...
ERROR (Section axles)
    Line (# 889): 2 4 5 0
    Message: Invalid property, ignoring whole line...

20:10:20: Cache: Parsing vehicle 'accord.truck' yielded following messages:

#FFFF00 WARNING #FFFFFF (Section flares)
    Line (3628): position_nodes
    Message: Not enough arguments, 5 required, got 1. Skipping line.
#FFFF00 WARNING #FFFFFF (Section flares)
    Line (3629): include 0-123
    Message: Not enough arguments, 5 required, got 2. Skipping line.
#FFFF00 WARNING #FFFFFF (Section flares)
    Line (3630): exclude 90-113
    Message: Not enough arguments, 5 required, got 2. Skipping line.

[DirtGamer301]

If I'm not mistaken the Mercedes W460 had some problems somewhere after RoR 0.4.0.7: http://www.rigsofrods.org/old-forum/forums/117-Light-Vehicles/71945/page-0008.html
All the "openable parts" (beams which are supposed to extent/break for example in a crash) open automatically when spawning or driving (so the beams either extent even if there's not enough force applied to them or break to easily, but I don't know much about that so it might be something else).
Also, the wheels are nearly white, but that's a know bug with the parser making materials nearly entirely white when the material is in the managedmaterials section but doesn't have any specular file.

The Lotus Esprit may also have some bugs, but it, by what I read, is made a bit differently, so that's most likely bugs you won't encounter with any other vehicle (I can't exactly remember what happaned here, and sadly can't test it right now, I just wanted to post those two vehicles now before I forget about it again): http://www.rigsofrods.org/old-forum/forums/117-Light-Vehicles/98617/page-0001.html

[ulteq]

From here: https://github.com/RigsOfRods/rigs-of-rods/pull/1123#issuecomment-251425969

2012 Salf-T-Liner HDX:
https://cloud.githubusercontent.com/assets/2660424/19080727/9e923b66-8a60-11e6-8b27-4f9ba943451a.png

[ulteq]

Yes, but BoxSamurai.truck and CH-54-Skycrane.airplane still produce alarming warning messages in the log file.

[tritonas00]

P-38 is ok now, but as soon you accelerate, it breaks: video

Other than that and 2012 Salf-T-Liner HDX, the new parser rocks for me :+1:

[ulteq]

Parser::GetArgPropulsion is different than before:

New: https://github.com/RigsOfRods/rigs-of-rods/blob/master/source/main/resources/rig_def_fileformat/RigDef_Parser.cpp#L4000-L4011
Old: https://github.com/only-a-ptr/ror-legacy-svn-trunk/blob/master/source/main/physics/input_output/SerializedRig.cpp#L2478

Example truck: https://forum.rigsofrods.org/thread-627-post-6026.html#pid6026

[only-a-ptr]

@ulteq You're right about Parser::GetArgPropulsion being different

The original code propagated the raw truckfile value into simulation and did 3 types of check: equality to 1, equality to 2 and > 0 (many different places.) This means even an invalid value yielded a propulsed wheel.

My logic resets invalid truckfile value to 0, which means all those >0 yield a non-propulsed wheel.

[only-a-ptr]

TODO list, april 2017

Crane collapses on spawn: https://archives.rigsofrods.org/repo/repofiles-4th-batch/LiebherrV54.zip

Audi80Quattro.truck

Pete379.truck

BoxSamurai.truck

2012 Salf-T-Liner HDX

(Ulteq) Rear axle issue when you accelerate:
1747b3b8deb2fed6ee7d50f397e322fa78ee9d66_dakardaf.zip
-> 6f52UID-daf.truck

[only-a-ptr]

Problem introduced in https://github.com/RigsOfRods/rigs-of-rods/pull/1271#issuecomment-306873472

Skyvan - spawns with incomplete cab mesh


Skyvan download: https://www.dropbox.com/s/8d8o3isw3rlsxys/Skyvan.zip?dl=1

[tritonas00]

Loading the first vehicle (2012 Saf-T-Liner HDX) from the link below crashes the current master: https://pastebin.com/4BaZKCzt

The fps version 2012 Saf-T-Liner HDX (FPS) works fine

https://www.dropbox.com/s/3clcpvol3jxm80b/a4f210f9f88e78ccaf4cac3a51d7c114b1d78ff1_HDX5.19.12b.zip?dl=1

It worked fine before #1271

[only-a-ptr]

@tritonas00 Looks like the tracking videocamera is causing issues. Testing...

[ghost]

They work if you press CTRL+V to re-enable videocameras.

[tritonas00]

Didn't know that, thanks!

[only-a-ptr]

658072f doesn't have the above issue, so #1362 broke it

@tritonas00 Thanks for checking, but it's not true - I've experienced it earlier. The trick is, the commit history isn't linear:

UPDATE2: I broke it! 😛 And this is the crime scene: https://github.com/RigsOfRods/rigs-of-rods/commit/3cb4da833c56f19e269a89b4a98770a763e3d6cc

UPDATE3: So dumb! I increase the wheel counter and THEN read the current wheel index from it :blush: https://github.com/RigsOfRods/rigs-of-rods/blob/3cb4da833c56f19e269a89b4a98770a763e3d6cc/source/main/physics/RigSpawner.cpp#L4237-L4242 - no wonder it always breaks the first wheel defined.

UPDATE4: Skidmarks are also created with wrong wheel ID. So I can kill 2 flies with one strike.

[tritonas00]

https://www.dropbox.com/s/0kazidh61bzyd5u/TJ_s_WTJ__V5.zip?dl=1

spawns like:

same also in latest master

[ghost]

Renaming ddescription to description fixes it. Surprised the mod author never noticed this error.

[tritonas00]

try spawn this: https://www.dropbox.com/s/wdx2mm6y2mi8v3a/b6688609f0581c05114049e8bb578a11344ed752_tatangjosesLilgtiToyotaFJ40strechedtruggy.zip?dl=1

[tritonas00]

something is wrong with this https://www.dropbox.com/s/5isu0wka5ehn18l/SDSArgo_1.zip?dl=1

[only-a-ptr]

@tritonas00 Strange - OGRE can't find material tracks/SDSargo although it parses the script file 'SDSargo.material' first without reporting problems. Anyway, this got broken in the "materialflare/skinzip/videocamera/whatever" patch, I need to resolve it.

[only-a-ptr]

@tritonas00 I found the problem with SDSArgo - there's an error in it's .material file - a missin } in the first material definition. Apparently OGRE tolerated this before but doesn't anymore.

Closing this because all other things are old - Skyvan is totally fine now.

[tritonas00]

please check this https://github.com/RigsOfRods/rigs-of-rods/issues/1008#issuecomment-307080609 also

[only-a-ptr]

@tritonas00 Allright

[only-a-ptr]

Allright, the mirrors are working again, but not the way I wanted :laughing:

[tritonas00]

Is this a parser or vehicle issue?

[tritonas00]

Load simple test terrain with water and the c1boat. Press UP and let it drive. After a while the game will crash. Backtrace leads to https://github.com/RigsOfRods/rigs-of-rods/blob/master/source/main/physics/collision/PointColDetector.cpp#L180

Full bt: https://pastebin.com/fkWx72Qb
Boat: C1boat.zip

[tritonas00]

https://www.dropbox.com/s/irv2zlinsadbuct/UnimogU5000Dakar.zip?dl=1 crashes the game:

error loading mesh: dashboard1.mesh: OGRE EXCEPTION(2:InvalidParametersException): Couldn't read 16 bit header value from input stream. in Serializer::determineEndianness at /home/babis/Downloads/rigsofrods-next-build/ogre/src/ogre_src_v1-8-1/OgreMain/src/OgreSerializer.cpp (line 75)

bt: https://pastebin.com/WZBv8VYW

[only-a-ptr]

@tritonas00 UnimogU5000Dakar.zip - Either the ZIP is corrupted (most likely) or the dashboard1.mesh format is so outdated that OGRE no longer recognizes it.

[ghost]

dashboard1.mesh is supposed to be a blank dashboard mesh used for the steering wheel, OGRE apparently now has issues loading it.
Here's a fixed mesh (in zip format since GitHub does not allow .mesh filetype):
dashboard1.zip

[tritonas00]

Closing this