tuxayo

Thanks for that huge migration work. So it's a git+svn workflow? Does that mean the game files are now spread across two repositories? On part on the git repo and the rest on an svn repo?

@Vantha Can you check the logs of the game after it crashes? On linux the logs are in /home/YOUR_USER/.config/0ad/logs/

Hi thanks for sharing this! Elytra is a nice step from Petra. Is it possible to have a version for a27? Now that the RC1 has been released. https://wildfiregames.com/forum/topic/107313-alpha-27-pre-releaserelease-candidate-build-testing/ https://trac.wildfiregames.com/wiki/PortA26ToA27 It output some errors so there are at least a few things to check. I would like to playtest the RC and doing that against a better AI would be more fun

Makes sense, the RC actually had some prepackaging so going blindly with source built dev version isn't equivalent (beside the few more recent patches since the RC). Thanks for clearing that up ^^ Anyway, it also works on my AMD HD 7950 (GCN 1.0) So hopefully there shouldn't be a bad surprise coming from the firsts Intel and AMD GPUs that had (experimental/incomplete) Vulkan support. I was worried about that.

Sorry, since we reached RC, I didn't think to check for manual stuff to add for Vulkan support. Ok so for now everyone should just use use the spir-v mod or the zip manually I guess. The thing in the main menu right? Not the validation you mention in the 1st post.

Oh crap, that might then be a packaging issue then. But the build instructions don't mention vulkan/spir/shader https://trac.wildfiregames.com/wiki/BuildInstructions Here is the packaging script: https://aur.archlinux.org/cgit/aur.git/tree/PKGBUILD?h=0ad-git Ho, the 0ad-spirv mod made it work! So false alert about the HD 4000 experimental support crashing the game. I will try to test with an AMD 7950 (GCN 1.0) so the 1st AMD GPUs with Vulkan support (but also experimental, I'm lucky with my GPUs XD). So what would be the right way to include the spir-v shaders when building packages? (distros might have bad surprises if they forget this and users are locked out of the game until package fix or config file edit) And also for the build instruction documentation? In case there is something different.

> Did your game not do that, or did you not give it the chance to do that? Is just crashes. With a clearly incompatible driver it triggers stuff like this: WARNING: Failed to create a Vulkan window: Installed Vulkan doesn't implement the VK_KHR_surface extension ERROR: Unable to create device for Vulkan backend, switching to GL. Like what is happening with the AppImage (it mistakenly doesn't find support with fully compliant GPU): https://github.com/0ad-matters/0ad-appimage/issues/19 In this case, the HD 4000 has experimental support enable by default in the official Linux Intel driver (Mesa). Which output this every time in the terminal: MESA-INTEL: warning: Ivy Bridge Vulkan support is incomplete But it mostly works. I can run the VKCube demo. Run Veloren (with some glitches but it's too slow anyway) and I can use Proton to run Windows games which uses DXVK to translate Direct3D calls to Vulkan. (I might not be able with the latest Proton&DXVK version which requires Vulkan 1.3 support)

Hi I built the WIP dev version (0ad-git AUR package on Arch Linux) and with rendererbackend = "vulkan" in user.cfg it crashes on start before the menu. So I can't switch back to GL via the GUI, so this looks bad if the same would happen to anyone with the release and certain GPUs About that: I have an Intel HD 4000 (gen7 ivy bridge) so basically the first Intel with Vulkan support (but incomplete/unofficial!). Right at the frontier in the middle of the minefield. Sorry to not have tried sooner since it's and interesting case: not uncommon GPU and partial vulkan support enabled by default. There doesn't seem to be an existing ticket. Is that correct? So should I create one and which additional info should I provide? In case there isn't already everything there. My build has the debug symbols. I can also make a debug build if needed. And run it with GDB. (I forgot how to make GDB spit the relevant info after a crash, or maybe it already does) ~ ❯❯❯ 0ad TIMER| InitVfs: 333.458 us FILES| Main log written to '/home/MIA_UZANTO/.config/0ad/logs/mainlog.html' FILES| Interesting log written to '/home/MIA_UZANTO/.config/0ad/logs/interestinglog.html' TIMER| CONFIG_Init: 5.08008 ms Sound: AlcInit success, using OpenAL Soft MESA-INTEL: warning: Ivy Bridge Vulkan support is incomplete FILES| UserReport written to '/home/MIA_UZANTO/.config/0ad/logs/userreport_hwdetect.txt' TIMER| RunHardwareDetection: 4.37839 ms FILES| Hardware details written to '/home/MIA_UZANTO/.config/0ad/logs/system_info.txt' TIMER| write_sys_info: 16.534 ms ERROR: CCacheLoader failed to find archived or source file for: "shaders/spirv/overlayline.xml" ERROR: Program 'spirv/overlayline' with required defines not found. ERROR: "USE_OBJECTCOLOR": "1" ERROR: Failed to load shader 'spirv/overlayline' ERROR: CCacheLoader failed to find archived or source file for: "shaders/spirv/overlayline.xml" ERROR: Program 'spirv/overlayline' with required defines not found. ERROR: "IGNORE_LOS": "1" ERROR: "USE_OBJECTCOLOR": "1" ERROR: Failed to load shader 'spirv/overlayline' ERROR: CCacheLoader failed to find archived or source file for: "shaders/spirv/overlayline.xml" ERROR: Program 'spirv/overlayline' with required defines not found. ERROR: Failed to load shader 'spirv/overlayline' ERROR: CCacheLoader failed to find archived or source file for: "shaders/spirv/foreground_overlay.xml" ERROR: Program 'spirv/foreground_overlay' with required defines not found. ERROR: Failed to load shader 'spirv/foreground_overlay' ERROR: CCacheLoader failed to find archived or source file for: "shaders/spirv/overlay_solid.xml" ERROR: Program 'spirv/overlay_solid' with required defines not found. ERROR: Failed to load shader 'spirv/overlay_solid' TIMER| InitRenderer: 13.8888 ms TIMER| ps_console: 3.17901 ms TIMER| ps_lang_hotkeys: 2.51741 ms TIMER| common/modern/setup.xml: 413.352 us TIMER| common/modern/styles.xml: 139.937 us TIMER| common/modern/sprites.xml: 889.833 us TIMER| common/global.xml: 384.85 us TIMER| common/setup.xml: 214.359 us TIMER| common/sprites.xml: 332.596 us TIMER| common/styles.xml: 101.024 us TIMER| pregame/backgrounds/: 503.768 us TIMER| pregame/sprites.xml: 118.376 us TIMER| pregame/styles.xml: 34.883 us TIMER| pregame/mainmenu.xml: 30.4513 ms ERROR: CCacheLoader failed to find archived or source file for: "shaders/spirv/canvas2d.xml" ERROR: Program 'spirv/canvas2d' with required defines not found. ERROR: Failed to load shader 'spirv/canvas2d' Assertion failed: "0 <= pass && pass < static_cast<int>(m_Passes.size())" Location: ShaderTechnique.cpp:59 (GetGraphicsPipelineState) Call stack: (0x5628a69d3d6f) /usr/bin/pyrogenesis(+0x8a7d6f) [0x5628a69d3d6f] (0x5628a69832e8) /usr/bin/pyrogenesis(+0x8572e8) [0x5628a69832e8] (0x5628a6984cd9) /usr/bin/pyrogenesis(+0x858cd9) [0x5628a6984cd9] (0x5628a6985607) /usr/bin/pyrogenesis(+0x859607) [0x5628a6985607] (0x5628a66656af) /usr/bin/pyrogenesis(+0x5396af) [0x5628a66656af] (0x5628a65fc053) /usr/bin/pyrogenesis(+0x4d0053) [0x5628a65fc053] (0x5628a65fa06a) /usr/bin/pyrogenesis(+0x4ce06a) [0x5628a65fa06a] (0x5628a696d2b2) /usr/bin/pyrogenesis(+0x8412b2) [0x5628a696d2b2] (0x5628a68e31d6) /usr/bin/pyrogenesis(+0x7b71d6) [0x5628a68e31d6] (0x5628a694839c) /usr/bin/pyrogenesis(+0x81c39c) [0x5628a694839c] (0x5628a68b2282) /usr/bin/pyrogenesis(+0x786282) [0x5628a68b2282] (0x5628a66aa055) /usr/bin/pyrogenesis(+0x57e055) [0x5628a66aa055] (0x5628a66aa736) /usr/bin/pyrogenesis(+0x57e736) [0x5628a66aa736] (0x5628a66ad5ea) /usr/bin/pyrogenesis(+0x5815ea) [0x5628a66ad5ea] (0x5628a61eab8b) /usr/bin/pyrogenesis(+0xbeb8b) [0x5628a61eab8b] (0x5628a61ecf55) /usr/bin/pyrogenesis(+0xc0f55) [0x5628a61ecf55] errno = 2 (Error during IO) OS error = ? ShaderTechnique.cpp(59): Assertion failed: "0 <= pass && pass < static_cast<int>(m_Passes.size())" /usr/include/c++/13.1.1/bits/stl_vector.h:1144: std::vector<_Tp, _Alloc>::const_reference std::vector<_Tp, _Alloc>::operator[](size_type) const [with _Tp = CShaderPass; _Alloc = std::allocator<CShaderPass>; const_reference = const CShaderPass&; size_type = long unsigned int]: Assertion '__n < this->size()' failed. /usr/bin/0ad: linio 9ª: 88824 Abortita (nekropsio elŝutita)"$pyrogenesis" "$@"

@Yekaterina Thanks for all the positive stuff over the years, the advice when I was learning to play while testing a25, the learning resources on the forum or videos you made. (I'm still finding them now that I'm rusted and am searching to revise usual build orders)

Just coming around to says that it's a very pleasant surprise to see restorative justice advocated. Thanks @Norse_Harold and @leopard for pushing for it. That's a thing pretty much all online (and IRL) communities would benefit for handling conflicts no matter who says or leaves in the end. edit: no matter who *stays*

What is your Linux distribution (name and version) and your level of familiarity with command line tools? The Snap part of the install instructions is too oriented for users familiar with command line tools. It should mention that «See https://snapcraft.io/0ad for more information. » is enough. And more experienced users can continue reading for more possibilities. One should be able to review a bit the PKGBUILD and their diffs on updates. And not just hit enter blindly. Otherwise it's too much blind trust from random people. And as for 0ad-git package, one must also know where make cleanup. I have more than 30 GiB of files sitting in my AUR helper directory after the build. And only less than 10 (the git clone) are actually reused when updating. > (I am using one called "yay", there is a new one called paru) There is also pikaur, it's so nice for general updates because it orders the packages by importance of updates. For example with most versions being X.Y.Z-A it will list first the packages where the version change is on X and then Y and then Z and then A. So it filters a lot of noise from rebuilds and allow to see immediately the major version changes. So this example. There are a lot of unofficial repos that are easy to install for the more niche software. (and GeckoLinux has them setup) No distro is at the level of the AUR but in the many more of less obscure packages I use, I don't think I'll miss a lot. Check the following with you relevant packages: https://software.opensuse.org/ https://pmbs.links2linux.de/search Yes, IIUC their focus on IoT usecase (99% of IoT should die, it's a waste of resources and pollution) has caused some technical choices that are terrible for launch time. Despite massive efforts to fix that, Chromium on Ubuntu (which is now forced as a snap and not a .deb) is horribly slow to start even on a recent PC. Errata from a previous message from me here: the Snap store website is libre but it's backend is not. So it's not possible to make an alternative Snap repo. (unless recoding) Like if Debian had chosen to keep the server part of APT closed (it's just an HTTP server serving files but imagine it's more complex) so only the Debian repos would exist. So, Ubuntu is still making a horrible choice, even worse than if it was for the website front end.

Regardless of the discussion channel, IRC, XMPP, Matrix, Revolt, etc, when starting a match, there should be no problem in using an external tool for voice provided it's very easy to use and requires nothing to install. And example is https://entraide.chatons.org/en/ => Video-conference => create. (use the circular arrows to switch to another server if it doesn't work) It will pop you in a Jitsi Meet room and the link can be shared on the discussion channels for the played to join. Leaving only the microphone enabled and here you go, voice chat channel joinable and leavable. It's libre, and on a network of ethical hosters (CHATONS) so no centralization and no hassle to pick up a server. (If using regularly, go to the website of the hoster you are using and give a few buck from time to time. That's how that can scale.)

Does anyone know where to find the Konoha Extreme AI? cc @Yekaterina As for Catilina it's at https://github.com/Nagasushi2021/Res_Gestae There might a confusion, Catilina is in nagasushi's repo. So the above quote might have the AI last authors mixed.

Part of the issue makes sense. If there are close units that physically block the view it's not possible to shot over. But depending of the terrain (ranged units on a steep hill), unit numbers and placement (the close unit front must be as wide as the units behind to act as a shield) and size (cav or elephants behind on-foot units), then it's should be possible to shoot units that are further. But being able to have tanks/meatshields units is important. To not have healers, heroes, catapults killed too easily.

Hi are you still maintaining your AI?

Amazing that it's possible to get a new AI via a mod. Just just tried 2 quick matches vs Petra, Kiara won Thanks for sharing!

They baited me! I didn't expected a troll strategy. Now I want to try spamming female citizen from houses. Good thing forge upgrades don't apply to them. I think it would make the strat viable. And break the meta. I chocked at them replacing the victory music with the Esperanto "anthem". I should have seen them coming given their nickname. Anyway, I should reach them to help me continue the translation ^^

Hi all Also, having an upper limit for frames per seconds should help about power consumption and heat. options => graphic general => FPS limit in the game. That's a bit extreme but I double checked in a replay and due to not having a good enough CPU, when the game does many things, I can get down to 4 FPS for extended periods of time. And I'm not feeling it! I can focus on managing an attack and economy. I suppose that depends on the person. So maybe you can try and see if a lower than 30 FPS limit will keep the game enjoyable. Anyway, great job for the developer that made the controls & stuff not being impacted by the number of FPS of the game. Does the relief of the map impacts performance? That could seriously be nice. It could still be visually enjoyable. Like in Minecraft/Minetest, it's the art style that matters the most. And even without yet a well refined art style, the option would be worth it. For big team games and low specs PCs. It's like if 0ad would more look like a tabletop game which is fine. It depends on the big picture of the house consumption. Check the bill would allow to check if there is something to actually gain by paying attention to the PC consumption. A low gaming time VS the other uses of the PC would make the power usage when gaming irrelevant. As well as having hot water heated with electricity, and electric heaters or air conditioning.

Yup, snap has non-libre components. At least the web app at https://snapcraft.io/ IIUC. About Arch, the user must know what they are getting into. Among the popular distros, it's the one that requires the most knowledge. What the is use of apt for building the dev version of projects from source? Yes, such a breeze I'm trying Tumbleweed (GeckoLinux actually) on another PC to see if it would fit people with minimal knowledge. So they could have up to date packages. https://repology.org/project/0ad/versions https://pkgs.org/search/?q=0ad&on=name The situation is very sad many months after the release. It was worse one month after the release. Since then IIUC only Solus, OpenSuse Leap have caught up but late. So the thirds party repos like the mentioned PPA. Distros should treat multiplayer games like the browsers. Meanwhile in a group of friends playing, those on Windows had the new version on day one. That's an additional step of knowledge requiered. @gui456wSERTDYF What's your experience with Linux distributions?

IIUC, using the dev version one should have a multithreaded pathfinder. I tried an experiment about that. The idea is that, if there is are intensive pathfinding calculations. Then the 0ad process should use more than on thread of CPU usage. That is to say, for a 4 thread CPU we should see a CPU usage > 25% (for the 0ad process) I pulled atlas, made two groups of 200 cavs earch. Made a lot of house. I made the two groups run past each other, cross the houses. And most of the time the 0ad process CPU usage is a 23/24% with a peak at 28% Is this test valid? If so then it means that all and the path finding is in one thread. Even further that would mean that the rendering is in the same thread.

Is there a place where they can be seen? To see more details about OpenGL, follow the changes in resolutions in the future and see what else is available there.