View Issue Details

IDProjectCategoryView StatusLast Update
0004854The Dark ModPhysicspublic16.12.2022 06:59
Reporterstgatilov Assigned To 
PrioritylowSeverityminorReproducibilitysometimes
Status confirmedResolutionopen 
Product VersionTDM 2.06 
Summary0004854: dmap: player can go through brush sides (large coordinates)
DescriptionIf the map is located at large coordinates, all sort of precision issues can happen.

This issue in particular is about a bug happening on a testmap test/4853_stuckTest2.map (in assets SVN). It was creates by Jesps:
  http://forums.thedarkmod.com/topic/19510-cauldron-clipping-issues/page-2#entry424145

There are some solid columns created with brushes, and you can go through some of their polygonal sides.
Steps To Reproduce1. Compile: dmap test/4853_stuckTest2.map
2. Start the map: map 4853_stuckTest2.map
3. Hit the leftmost switch on the wall --- you'll get teleported to another simiar room.
4. Go to the other half the room where columns are located. Find the column where arrows point (with text label "Walk through here").
5. Go through the labeled face.
Tagsprecision

Activities

stgatilov

stgatilov

30.06.2018 09:14

administrator   ~0010620

From what I see, the problematic polygon consists of vertices:
  /* 652 */ ( 41994.21484375 50187.23046875 0 )
  /* 651 */ ( 41993.421875 50232.4765625 0 )
  /* 646 */ ( 41994.1875 50187.2734375 128 )
  /* 645 */ ( 41993.41015625 50232.4765625 128 )
And edges:
  /* 969 */ ( 645 646 ) 0 2
  /* 977 */ ( 645 651 ) 0 2
  /* 978 */ ( 651 652 ) 1 1
  /* 979 */ ( 652 646 ) 0 1
And the polygon is:
  4 ( -969 977 978 979 ) ( -0.9998477101 -0.0174524076 0 ) -42863.703125 ( 41993.421875 50187.23046875 0 ) ( 41994.21484375 50232.4765625 128 ) "textures/darkmod/stone/natural/dark_dirty"

The only weird thing here is that numUsers of 978 and 979 edges is wrong: it must be 2 instead of 1. Something goes wrong here.

One the other hand, if I try to remove everything except this room and this column from the map, then the issue goes away.
nbohr1more

nbohr1more

16.12.2022 05:44

developer   ~0015577

I am unable to reproduce the problem with either test map in SVN and with the map from the thread.

I presume that collision fixes between 2.06 and 2.11 have somehow addressed this.
stgatilov

stgatilov

16.12.2022 06:19

administrator   ~0015578

I can reproduce it exactly as it was.
nbohr1more

nbohr1more

16.12.2022 06:33

developer   ~0015579

Last edited: 16.12.2022 06:59

Sorry
I was trying to reproduce getting stuck on the rock edge per the map.
The tracker replication steps work, but getting stuck on the brush in the original map doesn't...

Issue History

Date Modified Username Field Change
30.06.2018 09:10 stgatilov New Issue
30.06.2018 09:10 stgatilov Tag Attached: precision
30.06.2018 09:14 stgatilov Note Added: 0010620
30.06.2018 09:16 stgatilov Steps to Reproduce Updated
16.12.2022 05:44 nbohr1more Note Added: 0015577
16.12.2022 05:45 nbohr1more Status new => closed
16.12.2022 05:45 nbohr1more Resolution open => unable to reproduce
16.12.2022 05:45 nbohr1more Fixed in Version => TDM 2.11
16.12.2022 05:45 nbohr1more Target Version => TDM 2.11
16.12.2022 05:46 nbohr1more Assigned To => Springheel
16.12.2022 06:19 stgatilov Assigned To Springheel =>
16.12.2022 06:19 stgatilov Status closed => confirmed
16.12.2022 06:19 stgatilov Note Added: 0015578
16.12.2022 06:27 nbohr1more Fixed in Version TDM 2.11 =>
16.12.2022 06:27 nbohr1more Target Version TDM 2.11 =>
16.12.2022 06:33 nbohr1more Note Added: 0015579
16.12.2022 06:33 nbohr1more Resolution unable to reproduce => open
16.12.2022 06:59 nbohr1more Note Edited: 0015579