Unvanquished Development activity

Events from:

Daemon Engine
Daemon Engine
Tuesday, 21 January 2020
@illwieckz
illwieckz pushed to DaemonEngine/Daemon Jan 21, 2020
2 commits to master
  • @illwieckz ea1aeda
    cvar: make fs_legacypaks and fs_maxSymlinkDepth latched cvars
  • @illwieckz 2b97273
    cvar: make r_dpMaterial a latched cvar
@illwieckz
illwieckz merged a pull request in DaemonEngine/Daemon Jan 21, 2020
cvar: make r_dpMaterial, fs_legacypaks and fs_maxSymlinkDepth latched cvars #264

reimplement latched cvar in new cvar system make r_dpMaterial a latched cvar make fs_legacypaks and fs_maxSymlinkDepth latched cvars All I want i…

+3 -3
3 comments
@illwieckz
illwieckz pushed to DaemonEngine/Daemon Jan 21, 2020
2 commits to master
@illwieckz
illwieckz merged a pull request in DaemonEngine/Daemon Jan 21, 2020
renderer: make dpMaterials collapsed shader from the start, including non-shader images #266

now useless very ugly FindMapInStage is now dropped now useless heightmap collapsing code is now dropped extra maps loading for non-shader images …

+174 -256
1 comment
@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#266 Jan 21, 2020

tr_shader.cpp:1809:48: error: invalid operands of types ‘char [64]’ and ‘const char* const’ to binary ‘operator+’ 1809 | std::string extraMapName =…

@illwieckz
illwieckz pushed to UnvanquishedAssets/UnvanquishedTestAssets Jan 21, 2020
2 commits to master
  • @illwieckz 26de4e4
    normal: delete a now useless action leftover
  • @illwieckz d5280fd
    normal: make heightmap a loose file
@illwieckz
illwieckz pushed to UnvanquishedAssets/UnvanquishedTestAssets Jan 21, 2020
2 commits to master
@slipher
slipher commented on pull request DaemonEngine/Daemon#266 Jan 21, 2020

could use + instead of Str::Format

NetRadiant
NetRadiant
Monday, 20 January 2020

Inspired from a work I done for XQEMU, this is a simple easy build wrapper. It is meant to be kept simple without so much feature. The first purpose is to give user a simple way to build netradiant (just type ./easy-builder and voilà), the second purpose is to not add features to this script.

This script is meant to be kept the smallest and simplest possible.

It was designed to allow more advanced usage while keeping the script the most laconic possible.

So, an user can just do:

./easy-builder

If build succeeds, the user would find produced files in install/ sub-directory as usual.

If required, that user can be told to do this to produce a debug build:

./easy-builder --debug

An advanced user can do:

SUBDIR=linux-amd64 VERBOSE=1 ./easy-builder -DFHS_INSTALL=ON -j12

That would build in build/linux-amd64 and install in install/linux-amd64 (which is convenient when building for multiple os/arch), enables FHS layout, compiles with 12 parallel tasks (autodetected if nproc exists), and enables verbose compilation.

I added a commit to auto-enable Crunch compilation in CMake if Crunch submodule is checked-out. The easy-builder script automatically checkouts it if missing.

While I was at it I added a minor tweak in library-bundling to be sure it fails on error.

I edited README in order to make user know the easy-builder script exists.

Thomas Debesse (24d46c86) at 20 Jan 19:52

Merge branch 'easy-builder' into 'master'

... and 4 more commits

@illwieckz
illwieckz pushed to DaemonEngine/Daemon Jan 20, 2020
1 commit to master
  • @illwieckz 14259b3
    DarkPlaces only supports one alphaFunc operation, GE128, introduce r_…
@slipher
slipher commented on pull request DaemonEngine/Daemon#273 Jan 20, 2020

Looking at the undo history in my editor, apparently I fat-fingered it just before committing.

@slipher
slipher commented on pull request DaemonEngine/Daemon#273 Jan 20, 2020

So I expected the coordinate system for normals to be like this: But the y-axis (green) actually points in the opposite direction, giving us (agai…

Daemon Engine
Daemon Engine
Sunday, 19 January 2020
@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#16 Jan 19, 2020

Also, I don't think it's a good idea to set the version in that file. Either the engine has a version, whatever the game, and it's better to hardco…

@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#16 Jan 19, 2020

This is an idea I just got, let's imagine those two .cfg file (yes, .cfg): builtin in_name "Unvanquished" builtin in_baseName "unvanquished" builtin

@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#273 Jan 19, 2020

Note that you can reset normal scaling (inverting) by setting r_normalScale to 0, whatever the debug mode. This is currently an unsupported feature…

@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#273 Jan 19, 2020

You can find my latest build of the test-normal map there: https://dl.illwieckz.net/b/unvanquished/test-assets/ Some screenshots:

@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#273 Jan 19, 2020

Things work as expected, except that the color wheel is flipped on the y-axis. @illwieckz I recall you've posted some things about different norma…

@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#273 Jan 19, 2020
@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#273 Jan 19, 2020

I'm not sure to understand what you are trying to do, but the commit I added in my branch there: https://github.com/illwieckz/Daemon/commits/normal…

@slipher
slipher opened a pull request in DaemonEngine/Daemon Jan 19, 2020
Add tangent-space option for r_showNormalMaps #273

It's kind of hard to interpret the colors produced by r_showNormalMaps, since you have to factor out the transformation from texture-based to absol…

+50 -11
@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#258 Jan 19, 2020

What seems to be is missing is a decision of how steep of an angle a certain height delta represents. So trying to reason this out geometrically..…

Daemon Engine
Daemon Engine
Saturday, 18 January 2020
@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#258 Jan 18, 2020

Hmm, I think I wanted to do !( i == 1 && j == 1) …which must be equivalent to (i != 1 || j !=1), right.

@slipher
slipher commented on pull request DaemonEngine/Daemon#258 Jan 18, 2020

This also ignores [0][1] and [2][1] and [1][0] and [2][0]

@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#258 Jan 18, 2020

note that I'm not sure to understand everything… but other implementations I read just did similar thing (they basically unrolled loops, ignoring t…

@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#258 Jan 18, 2020

one matrix is: N N N 0 0 0 N N N while other is: N 0 N N 0 N N 0 N mat[1][1] is know to be zero in all cases, because only surrounding pixels are…

@slipher
slipher commented on pull request DaemonEngine/Daemon#258 Jan 18, 2020

This condition looks wrong

@slipher
slipher commented on pull request DaemonEngine/Daemon#258 Jan 18, 2020

What seems to be is missing is a decision of how steep of an angle a certain height delta represents. So trying to reason this out geometrically...…

@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#258 Jan 18, 2020

made the comment explicit about this.

@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#258 Jan 18, 2020

The generated normal map from heightmap is the sand on the beach (using r_showNormalMaps): So, this is with sobel filtering: With the other le…

@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#267 Jan 18, 2020

I pushed a completely new implementation, introducing r_dpBlend which can be set on and off whatever r_dpMaterial is, and can be switched at runtim…

@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#267 Jan 18, 2020

I'll make this using a dedicated cvar, not the r_dpMaterial one, maybe r_dpBlend, so in the future it would be possible to enable r_dpMaterial with…

@illwieckz
illwieckz pushed to DaemonEngine/Daemon Jan 18, 2020
1 commit to master
  • @illwieckz 5efa3f4
    builtin: import to-be-builtin assets in pkg/daemon_src.dpkdir
@illwieckz
illwieckz merged a pull request in DaemonEngine/Daemon Jan 18, 2020
builtin: import to-be-builtin assets in pkg/daemon_src.dpkdir #268

The lifegoal of those assets is to be embedded in engine, a bit like glsl shaders are. Those are the assets required by the engine to run. Before w…

+73 -0
3 comments
@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#258 Jan 18, 2020

I'll do screenshot, I also want to make a test map.

@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#258 Jan 18, 2020

For what I understand, the Z component is still reconstructed from X and Y, the later being computed from heightmap. I'm getting black artifacts wi…

@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#258 Jan 18, 2020

Yep, it's currently only done on lightMapping shader. I'll enable it in other places too.

@slipher
slipher commented on pull request DaemonEngine/Daemon#258 Jan 18, 2020

Any screenshots? What should we see if it is working?

@slipher
slipher commented on pull request DaemonEngine/Daemon#258 Jan 18, 2020

This comment doesn't make sense here, since you are getting the normals from your calculations, rather than an RGTC texture.

@slipher
slipher commented on pull request DaemonEngine/Daemon#258 Jan 18, 2020

This section is pretty long, would you mind putting it in a function?

@slipher
slipher commented on pull request DaemonEngine/Daemon#258 Jan 18, 2020

Should this be done in more places? I see this attribute being added to a lot of shaders, but it is only switched on in this one place.

Daemon Engine
Daemon Engine
Friday, 17 January 2020
@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#268 Jan 17, 2020

How is daemon.shader referenced? There is no need to reference it, the engine loads all .shader files it finds in all dpk/dpkdir it loads. The sh…

@slipher
slipher commented on pull request DaemonEngine/Daemon#268 Jan 17, 2020

How is daemon.shader referenced?

@slipher
slipher pushed to DaemonEngine/Daemon Jan 17, 2020
2 commits to master
@slipher
slipher merged a pull request in DaemonEngine/Daemon Jan 17, 2020
Implement CVAR_LATCH for new cvars #272

To do before merging: Prevent The change will take effect after restart messages from being printed when the cvars are set from command line/configs.

+60 -35
3 comments
Daemon Engine
Daemon Engine
Thursday, 16 January 2020
@illwieckz
illwieckz pushed to DaemonEngine/Daemon Jan 16, 2020
1 commit to master
  • @illwieckz 98de315
    vfs: improve some pak-loading related messages
@illwieckz
illwieckz merged a pull request in DaemonEngine/Daemon Jan 16, 2020
vfs: improve some pak-loading related messages #269

While I was toying with fs_basepath and fs_extrapaks I needed those minor improvement to really know what I was doing.

+10 -6
@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#265 Jan 16, 2020

The travis job shows the built branch was not yet rebased above this fix (not including it yet).

@illwieckz
illwieckz closed a pull request in DaemonEngine/Daemon Jan 16, 2020
quick and cheap parallax for real time lights #193

It jumps from 31fps to 95fps on hd1080p (110fps in a common scene) on an R9 390X with flame thrower on an open scene that is known to be far slower…

+29 -0
18 comments
@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#193 Jan 16, 2020

The patch saves a lot of time when using a forward renderer as, when using complex effects with many lights like a flame thrower, parallax has to b…

@briancullinan
briancullinan commented on pull request DaemonEngine/Daemon#193 Jan 16, 2020

@illwieckz saving on lighting compile time is exactly what I was hoping for! Did you already try this particular patch? I thought the parallel C++ …

Daemon Engine
Daemon Engine
Wednesday, 15 January 2020
@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#272 Jan 15, 2020

Just for my knowledge, why it has to be different? maybe a comment explaining why would be good, that's my only concern.

@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#272 Jan 15, 2020

I rebased #262 over this and it works correctly on my end. LGTM.

@illwieckz
illwieckz pushed to DaemonEngine/Daemon Jan 15, 2020
2 commits to master
@illwieckz
illwieckz merged a pull request in DaemonEngine/Daemon Jan 15, 2020
polishing: unify the way glsl shaders are written #263

unify the way glsl shaders are written: make easier to compare them while looking for bugs make factorisation or merge easier This is the way I f…

+105 -105
3 comments
@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#263 Jan 15, 2020

Thanks for the review!

@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#272 Jan 15, 2020

thanks for doing this!

Daemon Engine
Daemon Engine
Tuesday, 14 January 2020
@slipher
slipher commented on pull request DaemonEngine/Daemon#264 Jan 14, 2020

See #272 for latched cvars

@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#193 Jan 14, 2020

You mean you're trying to rebase that raytracing patch on latest ioquake3 master and in order to save render time you're interested in doing cheap …

@briancullinan
briancullinan commented on pull request DaemonEngine/Daemon#193 Jan 14, 2020

Interesting, I was planning on pulling this in https://ioquake3.org/2008/03/04/raytracing-ioquake3/ How is this related? The source patch is furthe…

Daemon Engine
Daemon Engine
Monday, 13 January 2020
@slipher
slipher opened a pull request in DaemonEngine/Daemon Jan 13, 2020
Implement CVAR_LATCH for new cvars #272

To do before merging: Prevent The change will take effect after restart messages from being printed when the cvars are set from command line/configs.

+51 -35
@illwieckz
illwieckz commented on issue DaemonEngine/Daemon#271 Jan 13, 2020

This may be somewhat safe as it's only building a string, then negative values are just printed into that string: ]/imagelist rocket -w-- -h-- -mm-…

@illwieckz
illwieckz opened an issue in DaemonEngine/Daemon Jan 13, 2020
RE_GenerateTexture: endless incrementing global signed integer is asking for overflow #271

See here: Daemon/src/engine/renderer/tr_image.cpp Lines 2964 to 2968 in 7770bfe Everytime a libRocket element is displayed for the first…

Daemon Engine
Daemon Engine
Saturday, 11 January 2020
@illwieckz
illwieckz commented on issue DaemonEngine/Daemon#270 Jan 11, 2020

An interesting read about why there such ugly artifacts on dark shadows: https://blog.demofox.org/2018/03/10/dont-convert-srgb-u8-to-linear-u8/

@illwieckz
illwieckz commented on issue DaemonEngine/Daemon#270 Jan 11, 2020

This is how looks Vega when compiled with noFast (less dark blotches because -fast has a faster faloff) and rendered with light correction. Except …

@illwieckz
illwieckz commented on issue DaemonEngine/Daemon#270 Jan 11, 2020

I attempted to do it the OpenGL way, it's far more intrusive and is more complicated to get it right as we have to make sure all images are uploade…

@illwieckz
illwieckz commented on issue DaemonEngine/Daemon#270 Jan 11, 2020

Note: I've discovered that the barricade diffuse texture we use: Is slightly darker than the one linked in its respective forum thread: I was exp…

@illwieckz
illwieckz commented on issue DaemonEngine/Daemon#270 Jan 11, 2020

The naive trick I used would be useful on pre-OpenGL 3 hardware in any way (driver without support for sRGB framebuffer and internal image formats)…

@illwieckz
illwieckz commented on issue DaemonEngine/Daemon#270 Jan 11, 2020

Note that, if I understand it well, using OpenGL features would do the opposite: making all the computation in sRGB color space and converting non-…

@illwieckz
illwieckz opened an issue in DaemonEngine/Daemon Jan 11, 2020
linear colorspace lightmap applied on non-linear sRGB colors pace diffusemap #270

This is related to #243 (sRGB lightmap) For what I know, what we call color map, diffuse map or albedo maps are in linear sRGB colors pace because …

Daemon Engine
Daemon Engine
Friday, 10 January 2020
@illwieckz
illwieckz pushed to DaemonEngine/Daemon Jan 10, 2020
1 commit to master
  • @illwieckz 7770bfe
    renderer: fix normalmap dark blotches
@illwieckz
illwieckz merged a pull request in DaemonEngine/Daemon Jan 10, 2020
renderer: fix normalmap dark blotches #260

I discovered that bug while toying with r_showNormalMaps (see #257). I was like: Wait, Wait… Wait……… As you can see the bug is also visible when…

+15 -2
12 comments
Daemon Engine
Daemon Engine
Thursday, 09 January 2020
@illwieckz
illwieckz opened a pull request in DaemonEngine/Daemon Jan 9, 2020
vfs: improve some pak-loading related messages #269

While I was toying with fs_basepath and fs_extrapaks I needed those minor improvement to really know what I was doing.

+10 -6
@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#268 Jan 9, 2020

Because we don't know how to embed those assets yet, games have to ship them themselves, but it would be nice to not require games to fix them. Eve…

@illwieckz
illwieckz opened a pull request in DaemonEngine/Daemon Jan 9, 2020
builtin: import to-be-builtin assets in pkg/daemon_src.dpkdir #268

The lifegoal of those assets is to be embedded in engine, a bit like glsl shaders are. Those are the assets required by the engine to run. Before w…

+73 -0
@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#264 Jan 9, 2020

I see, I admit all I was looking for was a message telling me to restart. It's true that LATCH does more than that.

@slipher
slipher commented on pull request DaemonEngine/Daemon#264 Jan 9, 2020

I tried the latched cvar and don't think it is working. Compare the behavior to an old-style latched cvar: ]/r_dpMaterial "r_dpMaterial" - bool - E…

@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#267 Jan 9, 2020
@illwieckz
illwieckz pushed to DaemonEngine/Daemon Jan 9, 2020
2 commits to master
@illwieckz
illwieckz merged a pull request in DaemonEngine/Daemon Jan 9, 2020
debugging: implement r_showNormalMaps and r_showMaterialMaps #257

implement r_showNormalMaps r_showEntityNormals is dropped, in any way, vertexLighting_DBS_entity and vertexLighting_DBS_world glsl are asking for m…

+51 -13
4 comments
@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#266 Jan 9, 2020

Now, the remaining bugs in solarium maps are: rendering issue of scrolling collapsed shader, which is a bug in Dæmon we have in Unvanquished if we…

@illwieckz
illwieckz merged a pull request in DaemonEngine/Daemon Jan 9, 2020
appveyor: also test Visual Studio 2017 and investigate NMake generator as we support them #262

README says the minimum supported VS is VS2017 so it's better to test it Unvanquished uses NMake Makefiles generator so it's better to test it

+25 -8
8 comments
@illwieckz
illwieckz pushed to DaemonEngine/Daemon Jan 9, 2020
2 commits to master
  • @illwieckz 9f2554e
    appveyor: drop NMake and do VS2019-64 and VS2017-32
  • @illwieckz 96d5c1f
    appveyor: also test Nmake Makefiles generator as Unvanquished builds …
  • 3 more commits »
@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#262 Jan 9, 2020

so I merged with VS2017-32 and VS2019-64 being built only note that there is there a good example of MSYS2-in-Appveyor implementation: mypaint/libm…

Daemon Engine
Daemon Engine
Wednesday, 08 January 2020
@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#263 Jan 8, 2020

In fact the best way to review this is to do things like that: meld lightMapping_fp.glsl \ vertexLighting_DBS_entity_fp.glsl \ vertexLighting_DBS_w…

@slipher
slipher commented on pull request DaemonEngine/Daemon#262 Jan 8, 2020

What about testing 64-bit on VS2019 and 32-bit on VS2017? it would reduce the build time to 10min. Yes, I think this is good enough. Rather than …

@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#262 Jan 8, 2020

Well, that would be a shame to discover the day of the release we fail to build 32-bit on top-notch setup because of a silent bug…

@illwieckz
illwieckz commented on pull request DaemonEngine/Daemon#262 Jan 8, 2020

OKOK, If you don't mind, I would add a commit to disable NMake and not squash them so we keep knowledge in history. Getting it working was not that…

@slipher
slipher commented on pull request DaemonEngine/Daemon#262 Jan 8, 2020

There's no real reason to care about the NMake Makefiles generator. I always figured that it must be needed because of the Visual Studio generator …