Announcement

Collapse
No announcement yet.

Darkplaces Error Reporting Thread

Collapse
This is a sticky topic.
X
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • adib
    replied
    Brush entities "flicker" at the player's sight

    -Happens to all of them all the time. They become visible when viewed at some angles or some distances. I've seen this in Quake and WinQuake while my map wasn't VISed yet. But using Darkplaces, it happens even when there's absolutely no leaks.

    - Just tie a brush to some func_wall entity and take a walk around it to replicate the bug.

    -The bug only happens in fullscreen. When running a windowed Darkplaces, all entities are rendered fine. Again, it doesn't happen in WinQuake, JoeQuake-GL or ezQuake-GL, only in Darkplaces.
    Last edited by adib; 03-28-2013, 11:14 PM.

    Leave a comment:


  • adib
    replied
    decode error on copying the texture data out of the bsp, the wrong surface size is being calculated, this can be due to floating point math errors
    Interesting that you mention. Here is one of the brushes' data:

    Code:
    {
    ( 304 -368 160 ) ( -208 -368 160 ) ( -192 -352 160 ) BCPLAST02 [ 0 1 0 32 ] [ 1 0 0 64 ] 0 1 1 
    ( 288 -352 120 ) ( -192 -352 120 ) ( -208 -368 120 ) MBTRIM04 [ 0 -1 0 -110.345 ] [ 1 0 0 88.9656 ] 0 0.29 0.29 
    ( -192 -352 120 ) ( 288 -352 120 ) ( 288 -352 160 ) MBTRIM04 [ -1 0 0 -64 ] [ 0 0 -1 -124 ] 0 0.3 0.31 
    ( 304 -368 120 ) ( -208 -368 120 ) ( -208 -368 160 ) MBTRIM04 [ 1 0 0 21.3334 ] [ 0 0 -1 -124.032 ] 0 0.3 0.31 
    ( 288 -352 120 ) ( 304 -368 120 ) ( 304 -368 160 ) BCPLAST02 [ 0 1 0 0 ] [ 0 0 -1 -96 ] 0 1 1 
    ( -208 -368 120 ) ( -192 -352 120 ) ( -192 -352 160 ) BCPLAST02 [ 0 1 0 0 ] [ 0 0 -1 -96 ] 0 1 1 
    }
    ... and they can get worse:

    Code:
    {
    ( 160 -392 520 ) ( 248 -392 608 ) ( 248 -400 608 ) BCTRIM03 [ 0.707107 -6.12303e-017 0.707107 14.6274 ] [ -8.65927e-017 -1 -8.65927e-017 -120 ] -90 1 1 
    ( 256 -392 600 ) ( 256 -400 600 ) ( 248 -400 608 ) PLAST02 [ -1 0 0 104 ] [ 0 1 0 56 ] 0 1 1 
    ( 256 -400 600 ) ( 168 -400 512 ) ( 160 -400 520 ) BCTRIM02 [ -0.707107 0 -0.707107 -64.3604 ] [ 0.707107 0 -0.707107 -31.2081 ] 45 0.8 0.8 
    ( 160 -392 520 ) ( 168 -392 512 ) ( 256 -392 600 ) BCTRIM02 [ -0.707107 0 -0.707107 -64.4315 ] [ 0.707107 0 -0.707107 -31.1371 ] 45 0.8 0.8 
    ( 168 -400 512 ) ( 168 -392 512 ) ( 160 -392 520 ) BCTRIM03 [ -0.707104 9.57786e-017 0.707104 22.6273 ] [ 0 -0.999996 0 -120 ] -90 1 1 
    ( 168 -392 512 ) ( 168 -400 512 ) ( 256 -400 600 ) BCTRIM03 [ -0.707107 6.12303e-017 -0.707107 -22.6274 ] [ 0 -1 0 -120 ] 90 1 1 
    }
    Floats appear on texture definitions too often, mostly in the offset fields.

    I'm using Worldcraft 3.3 and Bengt Jardrup's tools bundled with it:

    TxQBSP 1.12
    Vis 2.31
    Light 1.43

    I never defined texture offsets like -124.032 in the editor, they are always integers. Seems that Worldcraft is generating these floating point numbers while translating my model to MAP.

    LordHavoc, I can send you the map file, textures, whatever you need, if it helps.


    P.S.: all the brushes are snapped to an 8 point grid, there are no vertices off the grid.

    Leave a comment:


  • LordHavoc
    replied
    Never seen this behavior before, so everything below is purely guessing...

    Perhaps there is a .lit file included with the .bsp and that .lit file is not valid? Renaming it so it loads only the lighting within the .bsp might help.

    It's also possible there's a different rounding behavior on the floating point math in darkplaces which just happens to affect this bsp due to some extremely unlikely numbers in it for that surface.

    I will note that the artifact being shown is a decode error on copying the texture data out of the bsp, the wrong surface size is being calculated, this can be due to floating point math errors (the q1bsp format uses floating point numbers, and expects the compiler and engine to round those numbers to integer identically).

    Leave a comment:


  • adib
    replied
    Light artifacts that doesn't happen in other engines:

    http://i3.photobucket.com/albums/y80...ps9a316090.jpg

    There's nothing specially complex about this surface or neighbor brushes. Lighting is just ambient (80) and sunlight (255) so far.

    There are few more random dark artifacts in this map.

    I'm new to DarkPlaces. Is anyone else aware of this?

    Leave a comment:


  • Koakuma
    replied
    *"Controls wont save"*
    -Closing out DarkPlaces saves my controls, but remaps the default controls too. (Simplest way I can describe it, sorry :u.)
    -Close and reopen DarkPlaces
    -This happens on Quake 3: Team Arena also, but on there it happens when I close the menu itself.

    Leave a comment:


  • Supa
    replied
    Originally posted by Patrol1985 View Post
    Hip2m3 (The Catacombs) - Vores don't spawn and prevent further progress

    - Bug description: When I entered the room with a horn of conjuring (the one which summons a Death Knight) I killed a shambler... and didn't know what else to do. I checked every corner of the level and finally decided to watch a walkthrough. In the video, the player enters the room where a shambler appears along with two vores. After killing everyone a message "sequence complete" appears and some doors open leading to stairs and further part of the level. When I played, the vores never spawned thus preventing my progress.
    This has been fixed in the latest stable.

    Originally posted by _Smith_ View Post
    Fiends, after a jump, can get stuck if they land on the edge of level geometry. It looks like DP does not detect a touch in such case.
    It's a very frequent event with darkplaces and it's hard to not encounter it on many levels.
    Thanks, we've been able to reproduce this one but haven't had the time to diagnose the root cause yet.

    Leave a comment:


  • _Smith_
    replied
    Fiends, after a jump, can get stuck if they land on the edge of level geometry. It looks like DP does not detect a touch in such case.
    It's a very frequent event with darkplaces and it's hard to not encounter it on many levels.

    An easy place to replicate is e3m5 – go through the first windtunnel and trick a fiend to jump out of the platform.
    It'll keep getting stuck on the edge of the platform for 3s after each jump.

    There is a 3s timer in original qc to unstuck it in such case, but I don't recall it actually happening with other engines.
    It's not related to new builds, all build I've used had this problem.

    ***
    Small issue with new builds: mouselook is enabled on intermission screen when using our csqc hud mod.

    Leave a comment:


  • _Smith_
    replied
    Please, refrain from offtopic discussion in this thread. Google will tell you what normal texture map is.

    Leave a comment:


  • Kracov
    replied
    thanks. what is the normal texture addon?

    Leave a comment:


  • _Smith_
    replied
    You have probably too little video memory. Rygel's ultra pack textures are huge.
    Enable textures compression in Options -> Change Video Mode. It will make loadings extremely long though.

    Also, consider following a link from my sig.

    Leave a comment:


  • Kracov
    replied
    Lag/freeze bug
    When loaded with rygel's hi-res texture pack there are a lot of lags/freezes every 3-10 seconds, and last about 1-4 seconds each time. Any way to correct this? I've turned threading optimization off in my graphics driver options. I tried the "gl_finish 1" even though I may be using the regular version of Quake and I don't have opengl.

    I'm using Windows 7, and other 3D games run smoothly like Torchlight with plenty of mods installed.

    Leave a comment:


  • _Smith_
    replied
    The only thing I can add is that it is not present in this build 2011_07_02 and was introduced somewhere around September 2011. Thank you for all your work.

    Leave a comment:


  • LordHavoc
    replied
    Originally posted by _Smith_ View Post
    It looks like a part of the player model, that is on the other side of the water surface, is rendered when "r_water" "1"
    I am aware of this issue but don't have a solution - hiding the player model in water refraction causes other issues...

    Leave a comment:


  • _Smith_
    replied
    Issue with r_water 1

    It looks like a part of the player model, that is on the other side of the water surface, is rendered when "r_water" "1"

    Leave a comment:


  • talisa
    replied
    yep i can confirm this. just installed new 310.33 beta too myself, and tried running DP and no errors anymore

    so anyone else who downgraded to 301.42 before cuz newer drivers gave errors and had dp half-broken, get new 310.33, they work fine

    Leave a comment:

Working...
X