stgatilov

I'm afraid I have no idea what could have caused it.

I think this is the bug: the number should increase to 1. The problem happens in idAI::PerformVisualScan: if ( !canWalkToPlayer || ((m_LastSight - physicsObj.GetOrigin()).LengthFast()*s_DOOM_TO_METERS ) <= cv_ai_sight_combat_cutoff.GetFloat() ) { SetEnemy(player); // set flag that tells UpDateEnemyPosition() to NOT count this instance of player // visibility in the mission data m_ignorePlayer = true; // grayman #3063 - don't count this instance for mission statistics (defer until Combat state begins) } else // player is too far away, but AI will continue to move because he can walk to the player { newAlertLevel = thresh_5 - 0.1; alertInc = newAlertLevel - AI_AlertLevel; } The flag m_ignorePlayer says to not increase the number right now. It is intended that it would be increased later, but it seems that for non-combat AI this "later" never happens. I have created an issue 5286 for this.

Oh crap, I forgot about it! Committed in svn rev 8807. Thanks!

I have prepared some sort of alternative. But it makes sense only if you want to install TDM from scratch, i.e. you have no TDM installation on your machine yet. Here are instructions to install TDM 2.08 from another server: Make directory somewhere, where you can easily write files, e.g. C:\Games\TheDarkMod. Download archive by this link and unpack files from it to the directory. Now run the curl_install_tdm208.cmd batch file. Wait for it to download all the stuff. Unpack all files with .zip archives: tdm_shared_stuff.zip, tdm_update_win.zip, and tdm_update_linux.zip Delete the file curl_install_tdm208.cmd to ensure that you won't run it again accidentally. You are ready to play now! Basically, it uses the famous curl tool (built for Windows) to download all the files one by one. The files are downloaded from CDN endpoint, so it should be relatively fast. It took me 15 minutes to download the whole TDM 2.08. Also, with some technical skills, you can extract direct links to all the files from script and put them to some other software. Note that it is still HTTP, not a torrent

I'd like to ask people who took part in 2.08 beta testing: was downloading beta versions significantly faster?

Why not dive in and find the reason yourself?

This is really interesting. Did you look when exactly these errors happen? Right when TDM starts? Do these trailing errors "Value out of range" happen in constant number, or they continue spamming as long as your main menu stays idle? Of course, it makes sense only if these logfile is flushed after every message, which is how I would do it myself... UPDATE: Try to disable HRTF in main menu and restart TDM. See if 64-bit version starts working this way.

I think when debug cvars are enabled, then TDM actually crashes here. Or stops working in some other way. I might be that debug functions are sometimes supported (maybe drivers update since then). Maybe it is recording issue. Anyway, this debugging puzzle is not the main problem. @Dracula: When you record a better trace, please make a condump too and post console log here again.

Here is the riddle for fellow coders ( @duzenko @cabalistic). According to the log posted here, the driver does not support GL_KHR_debug. I have attached the end of the trace. According to it, execution has got to the end of OpenGL initialization, queried all features, then started executing debug commands. I am absolutely sure that the call to GLimp_CheckRequiredFeatures has just ended, and the following code has executed: if( glConfig.debugGroupsAvailable ) { qglDebugMessageCallback( R_OpenGLDebugMessageCallback, nullptr ); if( r_glDebugOutput.GetBool() ) { qglEnable( GL_DEBUG_OUTPUT ); } else { qglDisable( GL_DEBUG_OUTPUT ); } if( r_glDebugOutput.GetInteger() == 2) { qglEnable( GL_DEBUG_OUTPUT_SYNCHRONOUS ); } else { qglDisable( GL_DEBUG_OUTPUT_SYNCHRONOUS ); } } How the hell is it possible? Extension is not available, but the if is executed?!

This one is almost empty, it did not even get to the main menu. Was it the run which quickly closed by itself? Could you please try the way when TDM did not close automatically. The file in question should be about 20-40 MB in size. Everything under 1 MB is not going to work.

Yes, zip it, put it somewhere, and post a link here.

I think not. It stays forever without script, and I think it should do the same with it. Try this one, maybe it would be better. apitrace_record.cmd

Please download this archive. It contains single directory with name starting with "apitrace". Unpack this directory and rename it to C:\apitrace (path C:\apitrace\bin\apitrace.exe should point to executable). Then download the attached script apitrace_record.cmd into the directory where TDM is installed. Run the script. It should start TDM. When TDM window appears, wait 5 seconds, then close it (e.g. with Alt+F4). Now find file name myw64.trace in the same directory, archive it and share it somewhere (google drive or something similar). apitrace_record.cmd

I think this problem will return to you, to all of you If you are willing to look at the code, it is most likely in BinaryFrobMover.cpp.

What about these settings, did they help?

I think the main problem could be: ALCplaybackAlsa_mixerNoMMapProc: available update failed: Broken pipe It occurs randomly in both logs. Does sound work well in 32-bit? Did you try playing something for 5 minutes to check if it is buggy? You can enable OpenAL logs by settings environment variables: export ALSOFT_LOGLEVEL=3 export ALSOFT_LOGFILE=openal_log.txt Make sure the TDM executable is run with them enabled. After run you can inspect the logs in openal_log.txt. You might want to generate the log both for 32-bit and 64-bit, and check the diff between them. And just look for anything suspicious there. Also, you can change OpenAL backend in alsoft.ini. Uncomment line which assign drivers. You can try setting oss, port, jack, alsa (I guess alsa is default). The last thing is to build TDM including all dependencies from sources. If you want to try it, ask me: I'll explain how to build third-party libs.

Regarding highlight of lockpicking: something nondeterministic happens to it. It looks different to every player: some people have it non-transparent, some have it huge, now you see it yellow.

Does it work? I was under impression that GUI editor is long dead in TDM due to path changes.

It does not matter whether you have hotfix for 2.07 or not (and most likely you have it). Because savegames compatibility between 2.07 and 2.07 hotfix is guaranteed. However, in general when you change version of TDM, savegames stop working. Thus people usually finish all missions in progress before updating.

Could you also post console log from 32-bit build (for comparison) ?

Another try is: r_glCoreProfile 0 r_uniformTransforms 0 r_useFenceSync 0 But please make sure to revert these settings after you try (deleting the whole darkmod.cfg is the most reliable way to reset it to defaults).

We can try some more complicated ways to debug the problem, but they would need more patience from @Dracula. For simple checks, could you please set in your darkmod.cfg: r_glDebugContext 1 r_glDebugOutput 2 logFile 1 (in case you have not already) Then post contents of qconsole.log here. Please put large text files into spoiler environment (including your earlier post). That's the icon looking like eye in post editor.

I'm not sure what you are talking about. Rendering in browser lags behind desktop rendering in terms of features by several years. Until WebGL 2.0 came out, I estimated the difference to be about 10 years. Moreover, even if someone managed to build working TDM for browser, working with it would be a major pain for programmers (less/worse tools, more problems, everything is slow, etc). Just forget about browser, it is completely pointless.

Definitely not. It would be a massive time waste for a big step into the past.

I like the idea, but I'm afraid it is not possible to get the key from settings into the map right now. There is method idCommon::KeysFromBinding right now. I think it is possible to add a bit of code which would allow referencing the in a map. And I think there is already some way for a mapper to show text on screen.