Ceres

Hello! When 2 PCs run the latest Windows 10 release of 0 A.D., should I expect problems when hosting or joining from a device that runs one of the recent (or future) developer versions? If this has not yet been tested, I can just try it and report back.

Everybody can win from multiple cultures - also in gaming. The art works you guys do, are really impressive!

Now I would like to try feeding 'matchsettings.mp.json' to the mumble client and to have the players join the respective channels... "PlayerData": [{ "Name": "Player 1", "Civ": "spart", "Color": { "r": 21, "g": 55, "b": 149 }, "Team": 0, "AI": false }, { "Name": "Player 2", "Civ": "sele", "Color": { "r": 150, "g": 20, "b": 20 }, "Team": 0, "AI": false }, { "Name": "Player 3", "Civ": "ptol", "Color": { "r": 86, "g": 180, "b": 31 }, "Team": 1, "AI": false Player 1 + Player 2 (both in 'Team 0') will join channel 'All' and channel 'Team 0'. Player 3 (in 'Team 1') will join channel 'All'. Thus, all 3 players can talk via channel 'All', but the 2 allies in 'Team 0' can also talk among themselves, without Player 3 hearing them.

Well, it still works with GLSL enabled and even with all the other gfx options enabled and set to the maximum, except for the fps limit, which I set to slightly above 60, because my monitor runs at 60 Hz (1920x1080 pixels); vertical sync is also enabled. The fps is then miserable (avg. 15, max. 20), but the game does not crash anymore on that PC. It seems to be likely that the Intel iGPU driver update (version 27.20.100.9466, dated 16 Apr 2021) plays a role, but I am not sure, nor would I know what exactly could have contributed. Do you have some further hints or thoughts what I could do retrospectively?

Maybe putting the PCs to sleep would be an option. However, when we did that and later woke them up again, it was not possible to continue the multiplayer game. The message was about a player behind or something like this. I can check again.

Thank you! I found the file there (also as a note for myself): https://github.com/0ad/0ad/blob/f314b5b9d2be852a4d25f4b9ef40ab4acf4f2002/binaries/data/mods/public/gui/gamesetup/Persistence/PersistentMatchSettings.js ... PersistentMatchSettings.prototype.ConfigName = "persistmatchsettings"; PersistentMatchSettings.prototype.PersistedSettingsFileSingleplayer = "config/matchsettings.json"; PersistentMatchSettings.prototype.PersistedSettingsFileMultiplayer = "config/matchsettings.mp.json"; PS also for others interested - in Windows (10), the files are there: C:\Users\<username>\AppData\Roaming\0ad\config\ matchsettings.json matchsettings.mp.json user.cfg There is a nice (IMO) online JSON validator: https://jsonlint.com

Would it be feasible to add an option that pauses any drawing and looking-around-possibilities when the game is paused? Pause vs. "Sleep"?

@m7600 Cool! Which tool do you use for these great works?

Hm, yes, there was a driver update from Intel. I will look into this and also about the other options. Don't get me wrong - I will still post here in this thread and let you know what I found. Just for today, I would like to call it a day. Have a nice evening!

Sorry, this was not correct, I guess. I meant the latest available exe file for download on the play0ad website.

What shall I say? Believe it or not, but now the last auto-built version works also on that PC with GLSL enabled. This drives me crazy. I have not changed anything about the game. The PC windows resolution is also the same etc. I feel really embarassed that I have stolen your precious time for replying so kindly and patiently to my many questions. Even more I want to express my sincerest thanks for your help. And I have not yet given up, because I have already learnt a lot, want to give something back to the community, and also because I am really curious to find out what in fact was the problem. And as it happens sometimes, maybe the problem returns, and then I shall have made some progress at least in better understanding git, trac, etc.

Is it possible for testing purposes to have public.zip extracted when running pyrogenesis.exe? I ask because public.zip is always repacked when I replace files by earlier changesets. Or shall I use "pack-only" without compression? Would that work?

Yes, I know, but I am interested to understand whether these settings are then stored somewhere and if yes, where?

Thank you! This is what I have on my mind but was unable to describe properly. Would that work for multiplayer games as well?

If I see it right, I have to extract these and test the game each time: changeset_ps_trunk_binaries_data_mods_public_shaders_24777.zip changeset_ps_trunk_binaries_data_mods_public_shaders_25018.zip changeset_ps_trunk_binaries_data_mods_public_shaders_25114.zip changeset_ps_trunk_binaries_data_mods_public_shaders_25118.zip changeset_ps_trunk_binaries_data_mods_public_shaders_25271.zip changeset_ps_trunk_binaries_data_mods_public_shaders_25283.zip changeset_ps_trunk_binaries_data_mods_public_shaders_25285.zip Will have lunch now, though.

Somewhere between https://trac.wildfiregames.com/changeset/24938/ps/trunk/binaries/data/mods/public/shaders and https://trac.wildfiregames.com/changeset/25356/ps/trunk/binaries/data/mods/public/shaders (I will first check 'shaders')

I have not red everything under that link, but if I understand it correctly, it's the same as already mentioned there by @Yekaterina (see also Wiki). I was not sure yet, if that helps also during a paused game. I will give it a try.

Yep, so I will have a look there... https://trac.wildfiregames.com/changeset?old=24938&old_path=%2Fps%2Ftrunk%2Fbinaries%2Fdata%2Fmods%2Fpublic%2Fshaders&new=25356&new_path=%2Fps%2Ftrunk%2Fbinaries%2Fdata%2Fmods%2Fpublic%2Fshaders

Is there a way to DL from Trac auto-built archives that include 'system' and 'data'? This would very much help me finding out what the problem was and where it got fixed etc.

Unfortunately, even the latest autobuilt game crashes on the PC when GLSL is enabled, i.e. simply replacing 'pyrogenesis.exe' does not seem to be a solution to find out which version is the last one that crashes and which is the first one that does not crash anymore. Could it be that some other files are the "culprit", so I have to test them all (i.e. the full set, respectively)?

There are 411 steps between 'pyrogenesis_25349.exe' and 'pyrogenesis_24938.exe' (I renamed them like this after DL). Is it appropriate to halve this and go down from 25349 by 205 to 25144 and then (if it still works) again halve this and go down by 205 to 24939, etc. (roughly)? PS: How I can see only auto-built pyrogenesis.exe files in Trac between 24938 and 25356?

Oh, great hint! I was not aware that the binaries are all there. Will try this! Maybe wraitii could nevertheless give me another hint about bisecting. (yeah, I know that I am slow )

Where in the 0 A.D. code can I look up the team/player selection? (to understand in which variables etc. this info is stored) And is it also stored in a file during gameplay? If yes, where? Thanks again for your help! PS: Sorry, if my questions maybe do not make sense (I am still trying to learn).

Building does not work either: 1>------ Build started: Project: Collada, Configuration: Release Win32 ------ 2>------ Build started: Project: atlas, Configuration: Release Win32 ------ 3>------ Build started: Project: engine, Configuration: Release Win32 ------ 4>------ Build started: Project: graphics, Configuration: Release Win32 ------ 3>precompiled.cpp 3>Unknown compiler version - please run the configure tests and report the results 1>precompiled.cpp 1>c:\users\master\documents\0ad\0ad\libraries\source\fcollada\include\fmath\fmvector3.h(107): warning C4458: declaration of 'x' hides class member I do not understand why VS now tells me that the compiler version is unknown, when 10 minutes ago (starting from scratch and cloning the latest 0 A.D. version from GitHub), building was fine (and the game also worked). What does bisect do with regard to the compiler version?

I cannot execute 'update-workspaces.sh' when bisecting: master@HP-PC MINGW64 ~/Documents/0AD/0ad (master) $ git bisect start --term-old broken --term-new fixed master@HP-PC MINGW64 ~/Documents/0AD/0ad (master|BISECTING) $ git bisect fixed master master@HP-PC MINGW64 ~/Documents/0AD/0ad (master|BISECTING) $ git bisect broken A24b Bisecting: 1308 revisions left to test after this (roughly 10 steps) [ab3f99641ef65f471886987a102adfa8793032f6] Merge 'remotes/trunk' master@HP-PC MINGW64 ~/Documents/0AD/0ad ((ab3f99641e...)|BISECTING) $ cd build/workspaces/ master@HP-PC MINGW64 ~/Documents/0AD/0ad/build/workspaces ((ab3f99641e...)|BISECTING) $ ./update-workspaces.sh Updating bundled third-party dependencies... ./update-workspaces.sh: line 85: make: command not found ERROR: FCollada build failed master@HP-PC MINGW64 ~/Documents/0AD/0ad/build/workspaces ((ab3f99641e...)|BISECTING) What am I doing wrong?