illwieckz commented Sep 23, 2021

This is bikesheding, but the native CMake boolean values are ON and OFF, CMake automatically converts 0, 1, true and false to either ON or OFF.

illwieckz commented Sep 22, 2021

I remind we still don't test Linux build since Travis was dropped, so I don't know if there can be a collateral regression using the build checks. …

illwieckz commented Sep 22, 2021

Ok that's fine then. 👍 I'm OK with dropping VS2017, I just wanted to be sure we don't miss anything.

slipher commented Sep 22, 2021

We went testing VS2017 because our README says VS2017 is supported My commit updates the README to stay in sync. The latest Visual Studio works j…

illwieckz commented Sep 22, 2021

So we drop Visual Studio 2017? We went testing VS2017 because our README says VS2017 is supported, see #262 and Daemon/README.md Line 23 in 83…

slipher commented Sep 20, 2021

there is a regression in your version though: While g_disabledEquipment "lcannon" works and g_disabledEquipment "gren" works, g_disabledEquipment …

necessarily-equal commented Sep 20, 2021

there is a regression in your version though: While g_disabledEquipment "lcannon" works and g_disabledEquipment "gren" works, g_disabledEquipment "…

slipher commented Sep 20, 2021

I tried it out and you can indeed inspect non-alive entities. Hence the !CG_IsAlive check disrupts the functionality. .

necessarily-equal commented Sep 20, 2021

indentation went wrong here

necessarily-equal commented Sep 20, 2021

neat

necessarily-equal commented Sep 20, 2021

I mean, if it's a mover then I'd be surprised if it's in a team, but yeah, that changed. The whole point of this code is to "tag" the crosshair dep…

slipher commented Sep 19, 2021

I checked and everything is already done in the desired order: 1. set cvar according to "spawned string" or default 2. load default map config 3.…

illwieckz commented Sep 19, 2021

I checked and everything is already done in the desired order: 1. set cvar according to "spawned string" or default 2. load default map config 3. …

slipher commented Sep 19, 2021

I made a version fixing my comments (and other stuff) here: #1503

slipher commented Sep 19, 2021

One other thing that needs to be checked is whether this change compatible with server owners setting variables in the configs are executed on map…

Gireen commented Sep 17, 2021

never mind bt syntax doesn't like me 😅

illwieckz commented Sep 17, 2021

Right, I modified accordingly. I also added some comments to tell future ourselves to not forget to add early return when audio is not initialized …

slipher commented Sep 17, 2021

I prefer to not add anything in UpdateReverbSlot, especially since that allows the warning to be seen. I would rather only put the checks on the en…

illwieckz commented Sep 17, 2021

Ah yes, with not initialized early return in SetReverb it looks like I don't need such test in UpdateReverbSlot to silent a missing preset log. @sl…

slipher commented Sep 17, 2021

You missed SetReverb. LGTM otherwise

Gireen commented Sep 16, 2021

Maybe limit the number of maps to skip to 10 or 20 and if that happens fall back to the last played map.

Gireen commented Sep 16, 2021

Oh that's perfect 😃

Gireen commented Sep 16, 2021

Why? I see no benefit benefit in it, only reduced flexibility. The way it is now we could for example add rectangular bounding boxes.

illwieckz commented Sep 16, 2021

Here is the error window: https://github.com/UnvanquishedAssets/unvanquished_src.dpkdir/blob/master/ui/error.rml If I'm right the feature disappear…

illwieckz commented Sep 16, 2021

In the past we had an error window like that (screenshots from Unvanquished 0.35 and Unvanquished 0.17): The error depicted in those screenshot w…

illwieckz commented Sep 16, 2021

Worked around in #312.

illwieckz commented Sep 16, 2021

As a side topic, we need a way to print significant error messages on main menu. We have some other errors that can happen, for example by failing …

illwieckz commented Sep 16, 2021

The one reporting the issue had a broken sound device and wanted to play the game anyway.

slipher commented Sep 16, 2021

Do you think that playing with sound is something non-dev users ever want to do? It's really difficult to play without sound, for me anyway. If thi…

illwieckz commented Sep 16, 2021

I discovered that a possible cross-platform workaround is to set ALSOFT_DRIVERS='null' environment variable to tell OpenAL to use a dummy sound dev…

illwieckz commented Sep 16, 2021

I noticed that with released game, one can use ALSOFT_DRIVERS='null' to use a dummy audio OpenAL backend and workaround the Sys::Error. We may even…

illwieckz commented Sep 16, 2021

On principle, sure. I don't really know how to test this though. You can use ALSOFT_DRIVERS='none' environment variable to set a non-existing Ope…

slipher commented Sep 16, 2021

I'm mainly concerned with the development burden dead code imposes just by existing. Let's say you want to figure out how fog is implemented an sea…

necessarily-equal commented Sep 16, 2021

On principle, sure. I don't really know how to test this though.

illwieckz commented Sep 16, 2021

Before: ALSA lib pcm_dmix.c:1089:(snd_pcm_dmix_open) unable to open slave AL lib: (EE) ALCplaybackAlsa_open: Could not open playback device 'defaul…

necessarily-equal commented Sep 16, 2021

so, to give some hint to someone who would want to implement that. To be able to trace a bounding box currently, the game will on each frame call G…

necessarily-equal commented Sep 16, 2021

static inline, if you want the inline

necessarily-equal commented Sep 16, 2021

I'm not sure dropping the shaders are an improvement. I have the impression the only benefit here is to allow for a slower compilation time, but th…

illwieckz commented Sep 16, 2021

I configured a special schroot environment to build Unvanquished. I usually build it on Ubuntu 20.04 LTS as my testing environment is 20.04 based (…

illwieckz commented Sep 15, 2021

Great! 👍

slipher commented Sep 15, 2021

I replaced it with a command to delete all empty subdirectories, which seems like a good idea and is also shorter.

slipher commented Sep 15, 2021

Fixes #480 !

bmorel commented Sep 15, 2021

That would be nice... inside a BBox, for perf reasons (reduce the problem, and then, do the subtle things)

illwieckz commented Sep 15, 2021

Now that we have bones-based model, we may even have bone-based composite BBox.

Viech commented Sep 15, 2021

This is nothing that can be fixed on map level and I don't think there is a reasonable workaround. Closing as also that particular elevator was rem…

illwieckz commented Sep 15, 2021

Ah ok, then you can do that: find "${PKG_PREFIX}/bin" "${PKG_PREFIX}/include" "${PKG_PREFIX}/lib" -maxdepth 0 -empty -exec rmdir {} \; I verified i…

illwieckz commented Sep 15, 2021

There is already another "volumetric fog" shader by the way. If I'm right gl_fogQuake3Shader is about fog that is not map-wise but brush-wise. Th…

slipher commented Sep 15, 2021

Yes but no one will know we had some volumetricCloud code if we don't list it somewhere. Not true, the command I gave lists all deleted GLSL shad…

illwieckz commented Sep 15, 2021

For those GLSL shaders being deleted because of not being used, we may want to do a wiki page to reference them, with link to the deletion commit…

slipher commented Sep 15, 2021

For those GLSL shaders being deleted because of not being used, we may want to do a wiki page to reference them, with link to the deletion commit,…

slipher commented Sep 15, 2021

The directory should always exist (thanks to the other lines added in this commit); the idea is to remove it if it's empty. I see there's an rmdir …

illwieckz commented Sep 15, 2021

For those GLSL shaders being deleted because of not being used, we may want to do a wiki page to reference them, with link to the deletion commit, …

illwieckz commented Sep 15, 2021

Hmmm, wait, it would still return false if it does not exist and stop the execution… Better do the opposite: ! [ -d "${PKG_PREFIX}/bin" ] || rmdir …

illwieckz commented Sep 15, 2021

Maybe doing it that way would be better? [ -d "${PKG_PREFIX}/bin" ] && rmdir "${PKG_PREFIX}/bin"

Gireen commented Sep 14, 2021

or maybe better move this directly into the health component HealthComponent::HandleDamage there happens godmode checks

Gireen commented Sep 14, 2021

Then how about g_bot_dretchDamageStruct ?

bmorel commented Sep 14, 2021

why not using std::string::empty() instead of this comparison?

bmorel commented Sep 14, 2021

This only allows to use the symbol. Technically, they still can't buy it nor throw it.

illwieckz commented Sep 14, 2021

When carrying the chaingun I get: Debug: R_AddMDVSurfaces: bad frame -2147483648 to -2147483648 for 'models/weapons/chaingun/chaingun.md3' (1) Whe…

illwieckz commented Sep 14, 2021

I also notice that logs.suppression.enabled 1 does not suppress those.