Unvanquished Development activity

Events from:

Unvanquished
Unvanquished
Tuesday, 02 June 2020
@illwieckz
illwieckz commented on issue Unvanquished/Unvanquished#1111 Jun 2, 2020

I don't mind if it's contemporary or not, but I mind if it's good for player or not. A “Quick play” or “I’m Feeling Lucky” mechanism seems to be go…

Unvanquished
Unvanquished
Monday, 01 June 2020
@Zilor
Zilor commented on issue Unvanquished/Unvanquished#1111 Jun 1, 2020

* [ ] open server list at startup I don't think it's contemporary pointing users to the server browser. Nowadays matchmaking is the preferred wa…

Unvanquished
Unvanquished
Sunday, 31 May 2020
@DolceTriade
DolceTriade pushed to Unvanquished/Unvanquished May 31, 2020
1 commit to rmlui/sync
  • @DolceTriade 3ffe28c
    Fix circle menu selection
Unvanquished
Unvanquished
Saturday, 30 May 2020
@DolceTriade
DolceTriade pushed to Unvanquished/Unvanquished May 30, 2020
1 commit to rmlui/sync
Unvanquished
Unvanquished
Friday, 29 May 2020
@slipher
slipher commented on issue Unvanquished/Unvanquished#1185 May 29, 2020

OK, the yaw angle, i.e. the same one which you could control by moving around the target in a circle, is what I originally thought. But then my ori…

Unvanquished
Unvanquished
Thursday, 28 May 2020
@illwieckz
illwieckz commented on issue Unvanquished/Unvanquished#1185 May 28, 2020

When you're building on an horizontal floor, I'm speaking about using the vertical axis that go through the center of the buildable to rotate the b…

Unvanquished
Unvanquished
Wednesday, 27 May 2020
@slipher
slipher commented on pull request Unvanquished/Unvanquished#1201 May 27, 2020

The target is selected when you press +deconstruct and can't be changed without starting a new charge.

@slipher
slipher commented on issue Unvanquished/Unvanquished#1185 May 27, 2020

So the pitch angle is the one you want to change?

@illwieckz
illwieckz commented on pull request Unvanquished/Unvanquished#1201 May 27, 2020

The target lock requires you to stay near the targeted buildable when the charge completes. It intentionally does not require you to keep aiming a…

@illwieckz
illwieckz commented on issue Unvanquished/Unvanquished#1185 May 27, 2020

Hmm, I did not care about hitbox when I opened that issue. My concern is the rotation of the model itself, for example the direction you face when …

@slipher
slipher commented on issue Unvanquished/Unvanquished#1185 May 27, 2020

Now that I am remembering the fact that buildable bounding boxes are always axis-aligned, maybe it is not so interesting to be able to rotate them.…

@slipher
slipher commented on pull request Unvanquished/Unvanquished#1201 May 27, 2020

I implemented a form of target lock-on (supporting the early error message when the target is not valid for deconstruction) and allowed deconstruct…

NetRadiant
NetRadiant
Monday, 25 May 2020

It's a refactoring that came along some VFS work done for game features. It would have been possible to add more convolution to not factorize that code but that did not seem a good idea. Having a dedicated VFS code for notex only is as absurd as having some texwindow code only for WAD, and it looks like you're not happy yourself with that dedicated WAD texwindow code. So, you may be happy, VFS factorization is already done, no dedicated VFS for notex anymore!

Well, at least introduced notex mess has nothing to do with game features.

Thomas Debesse (acd4026b) at 25 May 10:06

q3map2 * fix: do not affect styled lightmaps by floodlight

Thomas Debesse (3d71aa7b) at 25 May 08:55

gamepacks: add Alien Arena (GPL)

  • load .pak and .wad files, fix #139
  • add .pakdir support for free (but that's not yet useful before map compilers handle it)
  • fix loading PNG fallbacks (notex, etc.) when the game does not support PNG, fix #141
  • do not load engine path twice when home path is engine path (happens when the game does not have home path)
  • properly list wads as texture directories, with working tree, per wad texture listing and wad name in texture window title

Reported by jar-el and PeterInRealLife in xonotic's mapping channel.

Initial report was: if the mapper does not select a texture before creating a brush, netradiant would crash:

This comes from the fact the game does not provide textures/radiant/notex and for some reason the one shipped by netradianbt is not loaded, despite the vfs is loaded.

vfs directory: /home/illwieckz/dev/netradiant/install/linux-amd64/base/
vfs directory: /home/illwieckz/Jeux/PC/alienarena/data1/
vfs directory: /home/illwieckz/Jeux/PC/alienarena/data1/
filesystem initialised
Texture load failed: "textures/radiant/notex"

It looks a bit like to #139.

Reported by Lava Croft on Quetoo channel.

It still doesn't read the wads from my quake/id1/ folder like before
Normally you set your engine path and radiant will read the wads you have stored in the id1 folder
speaking for q1 specifically, of course
yeah, it dont output anything useful in the console either
ah, it also dont seem to load the standard blue blocky notex placeholder
Texture load failed: "textures/radiant/notex"

vfs directory: E:/dev/netradiant_1.5.0-20200419-illwieckz-windows-amd64/base/
vfs directory: E:/Quake - FTEQW/id1/
vfs directory: E:/Quake - FTEQW/id1/
filesystem initialised

That may be an issue with the VFS.

Reported by Lava Croft on Quetoo channel.

It still doesn't read the wads from my quake/id1/ folder like before
Normally you set your engine path and radiant will read the wads you have stored in the id1 folder
speaking for q1 specifically, of course
yeah, it dont output anything useful in the console either
ah, it also dont seem to load the standard blue blocky notex placeholder
Texture load failed: "textures/radiant/notex"

vfs directory: E:/dev/netradiant_1.5.0-20200419-illwieckz-windows-amd64/base/
vfs directory: E:/Quake - FTEQW/id1/
vfs directory: E:/Quake - FTEQW/id1/
filesystem initialised

That may be an issue with the VFS.

Thomas Debesse (9c6a79a9) at 25 May 08:54

Merge branch 'vfs' into 'master'

... and 7 more commits

Better implementations of notex fix and wads loading are undoing commits, which introduced these problems.

Those commits were made to support more game features. The only way to go is to fixes potential regressions or rewrite it a better way. Remember what I said when I said I don't want people having to chose between features and support for their game.

Case sensitivity fix is ok, i'v been noticing this sort of problem in numerous spots besides this.

OK.

As for only show per-wad textures, while this behavior is wanted, implementation is maximally rusty. I prefer having none, than one, done this way, although wads handling in radiant is way different to q1/hl tools anyway.

To be honest I would prefer to not have those ugly “if wads” tests everywhere. But as you said, the behavior is wanted, so I'm not looking for contributions to delete it but I'm looking for contributions to make it better.

So there is no suggestion for a better implementation. Since the features are wanted and those implementations are good enough, they will make their way in to master.

Don't hesitate to come back if you have a better implementation for those features, or at least if you find an issue or a regression related to those implementations. Users will appreciate it.

@slipher
slipher pushed to Unvanquished/Unvanquished May 25, 2020
2 commits to 0.52.0/sync
  • @slipher 4969522
    Control all building-related cheat behavior by g_instantBuilding
  • @illwieckz 2707a57
    commands: make deconstruction code reusable
@slipher
slipher merged a pull request in Unvanquished/Unvanquished May 25, 2020
Use g_instantBuilding for decon checks #1204

... instead of g_cheats (i.e. whether devmap was used) as before. Pulled in one of illwieckz's refactoring commits since it moves the location of t…

+100 -68
@DolceTriade
DolceTriade commented on issue Unvanquished/Unvanquished#1202 May 25, 2020

Yeah, it seems probably that a lot of the engine logic that requires these buttons could be moved to the gamelogic as well. The main thing that see…

@DolceTriade
DolceTriade commented on issue Unvanquished/Unvanquished#1203 May 25, 2020

fwiw, I've been using GCC10 to do testing of the stuff for the RmlUi branch.

@slipher
slipher opened a pull request in Unvanquished/Unvanquished May 25, 2020
Use g_instantBuilding for decon checks #1204

... instead of g_cheats (i.e. whether devmap was used) as before. Pulled in one of illwieckz's refactoring commits since it moves the location of t…

+100 -68
Unvanquished
Unvanquished
Sunday, 24 May 2020
@DolceTriade
DolceTriade pushed to Unvanquished/Unvanquished May 24, 2020
1 commit to rmlui/sync
  • @DolceTriade 8a34f27
    Fix memory leak when registering custom RmlUi elements.
@slipher
slipher commented on pull request Unvanquished/Unvanquished#1182 May 24, 2020

Actually I was thinking the opposite regarding translation. The more complete of a sentence you have, the more likely the translation is to be corr…

@illwieckz
illwieckz commented on pull request Unvanquished/Unvanquished#1182 May 24, 2020

The name lookup seems to be more future-proof regarding translation.

@illwieckz
illwieckz commented on pull request Unvanquished/Unvanquished#1182 May 24, 2020

I used dash word because that's the word used in War§ow for a similar movement, and I lack imagination.

Better implementations of notex fix and wads loading are undoing commits, which introduced these problems.
Case sensitivity fix is ok, i'v been noticing this sort of problem in numerous spots besides this.
As for only show per-wad textures, while this behavior is wanted, implementation is maximally rusty. I prefer having none, than one, done this way, although wads handling in radiant is way different to q1/hl tools anyway.

@illwieckz
illwieckz opened an issue in Unvanquished/Unvanquished May 24, 2020
Testing game, engine and tools compilation with GCC10 #1203

GCC10 is known to break compilation of many idtech-related stuff: SpiKe reported me compilation issues on q3map2 (and dæmonmap) I fixed; ElBarto r…

@slipher
slipher commented on pull request Unvanquished/Unvanquished#1182 May 24, 2020

I like attack2 for start following. Then we have attack2 for cycling following vs. free-flying, and attack3 for cycling view mode. To me this is a …

@slipher
slipher commented on pull request Unvanquished/Unvanquished#1182 May 24, 2020

I think this should just be else { ? Since you can switch camera type regardless of whether you are on the spectator team.

@slipher
slipher commented on pull request Unvanquished/Unvanquished#1182 May 24, 2020

Nit: just Hold %s to use the jetpack seems better, no need for a name lookup.

@slipher
slipher commented on pull request Unvanquished/Unvanquished#1182 May 24, 2020

Maybe "lunge" instead of "dash" would be a better description, although it's not perfect.

@slipher
slipher commented on pull request Unvanquished/Unvanquished#1182 May 24, 2020

itemact not +itemact

@slipher
slipher opened an issue in Unvanquished/Unvanquished May 24, 2020
Move `BUTTON_` definitions into gamelogic #1202

The constants like BUTTON_ATTACK, BUTTON_ACTIVATE etc. should be moved from the engine into the gamelogic. A few of them are referenced in the engi…

@slipher
slipher commented on pull request Unvanquished/Unvanquished#1182 May 24, 2020

Instead of this can we: Leave all existing buttons in the same order Define the new button in the gamelogic (I did this in my BUTTON_DECONSTRUCT PR)

@slipher
slipher commented on pull request Unvanquished/Unvanquished#1182 May 24, 2020

This function should not be needed, just do BG_Weapon(n)->name

@slipher
slipher commented on pull request Unvanquished/Unvanquished#1182 May 24, 2020

I think it would be simpler to iterate the range [cg.snap->ps.weapon + 1, WP_NUM_WEAPONS) and then [WP_NONE + 1, cg.snap->ps.weapon). Then no extra…

Unvanquished
Unvanquished
Saturday, 23 May 2020
@illwieckz
illwieckz commented on pull request Unvanquished/Unvanquished#1201 May 23, 2020

The decon protection continues to work for me. Probably you are running into the unintended side effects of devmap again weary. I'll try to create…

@slipher
slipher commented on pull request Unvanquished/Unvanquished#1201 May 23, 2020

Anyway, your implementation does not prevent to decon the last spawn or the rc/overmind. The decon protection continues to work for me. Probably …

@illwieckz
illwieckz commented on pull request Unvanquished/Unvanquished#1201 May 23, 2020

Hmm, sorry, it looks like I checked out the branch in the wrong repository. The charge bar works. Anyway, your implementation does not prevent to d…

@slipher
slipher commented on pull request Unvanquished/Unvanquished#1201 May 23, 2020

The logic works on my end but the charge bar is not displayed. Even using the UnvanquishedAssets/unvanquished_src.dpkdir#17 branch?

@illwieckz
illwieckz commented on pull request Unvanquished/Unvanquished#1201 May 23, 2020

The logic works on my end but the charge bar is not displayed.

Well, seems I not tested these changes well. Closing for now, it's not ready for merge.

C++11 provides default copy assignment operators, so manual implementations of operator= now is deprecated and unnecessary (en.cppreference.com/w/cpp/language/copy_assignment).

@illwieckz
illwieckz pushed to UnvanquishedAssets/map-vega_src.dpkdir May 23, 2020
1 commit to master
  • @illwieckz af40508
    q3map2: daemon is not sRGB ready yet (default compilation profile is …
@illwieckz
illwieckz pushed to UnvanquishedAssets/UnvanquishedAssets May 23, 2020
2 commits to master
  • @illwieckz 3456b52
    sync submodules
  • @illwieckz ea7b067
    makefile: allow to checkout arbitrary git ref
@illwieckz
illwieckz pushed to DaemonEngine/Daemon May 23, 2020
2 commits to master
  • @dimhotepus 6b94e2e
    [tinygettext] Precache string length in loop (DEBUG builds optimization)
  • @dimhotepus 2960251
    [pdcurses] Copy argv + allocate space for '\0'
  • 3 more commits »
@illwieckz
illwieckz merged a pull request in DaemonEngine/Daemon May 23, 2020
PVS Studio fixes for used libs #340

Some fixes from PVS Studio static analyzer. Product page: https://www.viva64.com/en/pvs-studio/ How to use for free: https://www.viva64.com/en/b/0600/

+6 -10
3 comments
@illwieckz
illwieckz merged a pull request in DaemonEngine/Daemon May 23, 2020
Show detailed CRN image load warnings #342

Show detailed explanation of CRN image load errors for client. Related to #220

+39 -4
1 comment
@illwieckz
illwieckz pushed to DaemonEngine/Daemon May 23, 2020
1 commit to master
  • @dimhotepus 05140e8
    Show detailed CRN image load warnings

@winterheart do you have any idea why those little changes completely break NetRadiant?

NetRadiant
NetRadiant
Friday, 22 May 2020

There was some commented-out broken code to only display the filename of the wad like usual texture directories. I deleted it and entirely re-implemented it properly. Also, it's better since the tree was still done on _ detection.

I also made sure clicking a wad only show textures for the given wad (like usual texture directories), and that wad name is displayed in texture window title:

@Garux, this fixes notex loading when it's missing (like Quake1), and it fixes .WAD listing in texture browser on Linux (case sensitivity issue). See #139 (closed) and #141 (closed). If you have ideas for better implementations please share your thoughts, let's make NetRadiant great again. 🙂

Note that in the end I would like to see all editorImages to be loaded with all supported image plugins whatever the game supports them or not (they are meant to not be displayed by the game after all, and are usually alternatives for weird game-specific formats), same for the XYWindow background (see #142), but I haven't yet found an easy way to keep that information all the way down to QERApp_LoadImage. At least the current implementation prevents some crashes, which is good enough.

The texture browser now tests .wad file extension case-insensitively on all systems:

Lava Croft linked me to QUAKE101.WAD. I'm currently facing case-sensitivity issue: the console saids the wad is loaded but in fact the texture browser displays nothing. But when I rename the wad as quake101.wad, the magic happens:

NetRadiant now seems to load the .pak and .wad files:

vfs directory: /home/illwieckz/Jeux/PC/quake/id1/
pak file: /home/illwieckz/Jeux/PC/quake/id1/PAK1.PAK
pak file: /home/illwieckz/Jeux/PC/quake/id1/PAK0.PAK
pak file: /home/illwieckz/Jeux/PC/quake/id1/gfx.wad
filesystem initialised

I also fixed the bug that made the engine path being loaded twice, and the bug that made the fallback textures (notex, etc.) not being loaded. See !166 (merged).

By the way, I'm yet to see any Quake textures in texture browser. I know nothing about .pak and .wad files, so I don't know how to debug more.

At least, when show_wads is enabled in q1.game file (default behaviour), the texture browser list the wad. But I don't get more.

Note: the gfx.wad I used to test things is one I extracted from pak0.pak for quick testing purpose only, I'm not meaning it's the way to do it. I just needed a .wad file.

Note: NetRadiant can also load .pakdir (.pak can be extracted using tools like unpak from quake-cli-tools, or UnPAK), or pakextract from Yamagi Quake II (I only tested the first one). By the way, this feature is useless until map compilers support the feature. This is still useful to debug stuff.

  • load .pak and .wad files, fix #139
  • add .pakdir support for free (but that's not yet useful before map compilers handle it)
  • fix loading PNG fallbacks (notex, etc.) when the game does not support PNG, fix #141
  • do not load engine path twice when home path is engine path (happens when the game does not have home path)
  • properly list wads as texture directories, with working tree, per wad texture listing and wad name in texture window title

The XYWindow background image is loaded using QERApp_LoadImage which loads file from the VFS, ignoring all the formats that are not explicitly listed in .game file.

Because of this:

  1. a PNG XYWindow image can't be loaded if the game only supports TGA (similar to #141).
  2. a PNG XYWindow image can't be loaded from arbitrary directories like the user home directory.

In both case, I don't think so much restrictions makes sense. There can be legit uses case where the XYWindow background image is part of the mapper's workflow but does not have to be included in to-be-released game assets.

@illwieckz
illwieckz pushed to Unvanquished/pluto-devfeeds May 22, 2020
1 commit to master
  • @illwieckz eda45e3
    fetch feeds with a custom User Agent, GitLab seems to have blackliste…
NetRadiant
NetRadiant
Thursday, 21 May 2020

When I merge, compile and run this I get weird bugs in absolutely unrelated place. For example at startup NetRadiant loses the engine path for the current selected game, then on next startup it starts with another game… etc.

Or worst, starts a game but prefill the engine path from another game. For example start Heretic2 with Smokin' Guns engine path.

And segfaults when quitting.

Default textures/radiant/notex is not loaded because the provided fallback is in PNG format while AlienArena game pack only enables TGA and JPG image plugins.

C++11 provides default copy assignment operators, so manual implementations of operator= now is deprecated and unnecessary (en.cppreference.com/w/cpp/language/copy_assignment).

Because textures/radiant/notex is not loaded, creating a brush before selecting a texture may crash netradiant, see similar issue with PAK archives: #141.

While GtkRadiant provides vfspak, vfswad and vfspk3 plugins, NetRadiant only provides vfspk3 but also provides archivezip, archivepak, and archivewad. Maybe the vfs code was modified to be unique whatever the package format, but there is some bugs when the VFS is not pk3-based, because it's less tested, and maybe wads are not loaded at all.

While GtkRadiant provides vfspak, vfswad and vfspk3 plugins, NetRadiant only provides vfspk3 but also provides archivezip, archivepak, and archivewad. Maybe the vfs code was modified to be unique whatever the package format, but there is some bugs when the VFS is not pk3-based, because it's less tested.

Unlike Quake1 (see #139) there is no issue to load Alien Arena textures because they are distributed unpacked.

Reported by jar-el and PeterInRealLife in xonotic's mapping channel.

Initial report was: if the mapper does not select a texture before creating a brush, netradiant would crash:

This comes from the fact the game does not provide textures/radiant/notex and for some reason the one shipped by netradianbt is not loaded, despite the vfs is loaded.

vfs directory: /home/illwieckz/dev/netradiant/install/linux-amd64/base/
vfs directory: /home/illwieckz/Jeux/PC/alienarena/data1/
vfs directory: /home/illwieckz/Jeux/PC/alienarena/data1/
filesystem initialised
Texture load failed: "textures/radiant/notex"

It looks a bit like to #139.

Thank you for the patches. I applied the same fixes on daemonmap submodule.

For some reason GitLab did not detect the merge. So I'm closing it (it is merged).

Libraries under libs should be forced STATIC since there circular dependencies and implicit linking with external libraries. Otherwise compilation will fails with underlinking errors.

Thomas Debesse (ec255b56) at 21 May 09:26

Fixing compilation with -DBUILD_SHARED_LIBS=ON

... and 1 more commit

@illwieckz
illwieckz pushed to DaemonEngine/daemonmap May 21, 2020
2 commits to master
  • @illwieckz 1ce452e
    more fixes for compilation with '-DBUILD_SHARED_LIBS=ON'
  • @winterheart 292c76b
    Fixing compilation with -DBUILD_SHARED_LIBS=ON

Libraries under libs should be forced STATIC since there circular dependencies and implicit linking with external libraries. Otherwise compilation will fails with underlinking errors.

Unvanquished Assets
Unvanquished Assets
Wednesday, 20 May 2020
Unvanquished
Unvanquished
Tuesday, 19 May 2020
@slipher
slipher pushed to Unvanquished/Unvanquished May 19, 2020
2 commits to playerstate/sync
@slipher
slipher opened a pull request in Unvanquished/Unvanquished May 19, 2020
+deconstruct: hold to force-decon #1201

My attempt at a press-and-hold-to-deconstruct implementation, based on the work by @illwieckz. I'm not comfortable overloading +attack2 with decons…

+569 -153
@DolceTriade
DolceTriade pushed to Unvanquished/Unvanquished May 19, 2020
1 commit to rmlui/sync
  • @DolceTriade 04c2dc3
    Use the builtin RmlUi progressbar instead our custom one for our prog…
@DolceTriade
DolceTriade pushed to UnvanquishedAssets/unvanquished_src.dpkdir May 19, 2020
1 commit to rmlui/sync
  • @DolceTriade 58d0669
    Fix progressbars to use the new RmlUi progressbar
@slipher
slipher pushed to DaemonEngine/Daemon May 19, 2020
1 commit to playerstate/sync
  • @slipher cf6b891
    Netcode: fix non-byte-multiple signed integers
Unvanquished Assets
Unvanquished Assets
Monday, 18 May 2020
@illwieckz
illwieckz pushed to UnvanquishedAssets/unvanquished_src.dpkdir May 18, 2020
2 commits to rmlui/sync
  • @illwieckz ce1604b
    Fix image color syntax in hud_alien.rml and hud_human.rml
  • @illwieckz 5c10bdd
    Fix decorator syntax in hud_human.rml
@illwieckz
illwieckz commented on issue Unvanquished/Unvanquished#1200 May 18, 2020

The game suddenly quit the game and returned to the main menu. There was that in log: Warn: RE_AddRefEntityToScene: bad reType 1048033888

@DolceTriade
DolceTriade pushed to UnvanquishedAssets/unvanquished_src.dpkdir May 18, 2020
2 commits to rmlui/sync
  • @DolceTriade 8aa2e60
    Fix decoratorsyntax in window.rml
  • @DolceTriade 637d681
    Replace deprecated s/t decorator syntax with a spritesheet for checkb…
Unvanquished Assets
Unvanquished Assets
Sunday, 17 May 2020
@DolceTriade
DolceTriade pushed to UnvanquishedAssets/unvanquished_src.dpkdir May 17, 2020
2 commits to rmlui/sync
  • @DolceTriade 40adb7d
    Set default cursor in the global body.
  • @DolceTriade 119081a
    Fix missed merge conflict.
@DolceTriade
DolceTriade pushed to Unvanquished/Unvanquished May 17, 2020
1 commit to rmlui/sync
  • @DolceTriade 5ad64d6
    Implement mouse cursors utilizing rcss properly.
@DolceTriade
DolceTriade pushed to Unvanquished/Unvanquished May 17, 2020
1 commit to rmlui/sync
  • @DolceTriade e47c389
    Add rocketFocusManager's events in the capture phase.

Thomas Debesse (5980dc6b) at 17 May 15:53

Thomas Debesse (458a91b4) at 17 May 15:53

Merge branch 'illwieckz/gcc10' into 'master'

... and 9 more commits

I compiled NetRadiant as explained in the README:

cmake -G "Unix Makefiles" -S. -Bbuild -DCMAKE_BUILD_TYPE=Debug

cmake --build build -- -j$(nproc)

Here is the output (things start to get weird after 52% progress): paste.debian.net/hidden/77e43a05/

It is known to build on:

  • Linux Ubuntu 20.04 with GCC 10.0.1
  • Linux Ubuntu 20.04 with GCC 9.3.0
  • Linux Ubuntu 18.04.1 with GCC 7.5.0
  • FreeBSD 12 with Clang 6.0.1
  • Windows 10.0.18362 with GCC 9.2.0 (MSYS2/Mingw64)

I haven't tested the build themselves.

@SpiKe can you build this and compile a map?

Don't forget to update submodules after having checked out the branch:

git fetch
git checkout 'illwieckz/gcc10'
git submodule update --init --recursive

I also built some Unvanquished maps without seeing any issue.

Thomas Debesse (5980dc6b) at 17 May 15:50

daemonmap: update submodule

@illwieckz
illwieckz pushed to DaemonEngine/daemonmap May 17, 2020
1 commit to master
  • @illwieckz 2534f9d
    q3map2: declare or define with Q_EXTERN
@DolceTriade
DolceTriade pushed to Unvanquished/Unvanquished May 17, 2020
1 commit to rmlui/sync
  • @DolceTriade c486fa2
    Hardcode custom cursors in the rocket.txt for now.