stgatilov

Speaking of licenses. I think there is no reason to copy/paste TDM license, because it is huge and involves some third-party licences which your repos don't store. It also has tendency to change, so your version would quickly get out of date. If you want to ensure that materials in the repos could be included into the core mode, then you should have: GNU GENERAL PUBLIC LICENSE Version 3 for scripts. Creative Commons Attribution-Noncommercial-Share Alike 3.0 for models and prefabs (textures, etc).

There is betamapper SVN repo, which can serve as a place to share assets. Although I'm not sure what is its access policy today. I don't understand why you suggest using GitHub as file hosting site. How is it better than google drive and similar cloud storage? What will you do if repos grow large? With textures, you can reach 1 GB after some time. By the way, there is long-standing plan to create official GitHub mirror for engine source code. I'll have to create GitHub organization for TDM team, because the mirror repo should have several owners. But it's not a problem to have two TDM organizations, of course.

The next dev build will include more dmap changes, but this time about improving precision. Corresponding issues are: 5486 and 5580 Unfortunately, I did not put a cvar for disabling changes of 5580 (it would need duplicating too much code). So dmap in the next dev build and in 2.10 will differ a lot in its numeric behavior, and there would be no way back. Please keep this in mind when using new builds.

Be careful! Having target version 2.10 does not mean it will be in 2.10 On rare occasions even seeing issue as "resolved" does not mean it will be in the next release February 2022.

@teddybear, did you also update video drivers? Or install new GPU? Or plugged a different GPU to the new monitor? (in case you have several, like integrated Intel HD + discrete AMD Radeon)

You can delete it. It is important that this file is present on TDM server, because it is used for self-update. But after you download zip file, it has no use.

Yeah, I hope to fix it this year --- tracked as 5341.

I think power imbalance is an entirely different element, not necessarily tied to age. Abusing power imbalance is a problem regardless of age. The main reason why sex with underage person is forbidden is that he/she is deemed not mature enough (mentally and psychologically) to decide whether he/she wants it or not, and understand the consequences. For that reason there is no such concept as "voluntary" from the point of law.

Well, I guess I have found why he left FSF in 2019: https://selamjie.medium.com/remove-richard-stallman-fec6ec210794 Basically, he wrote that it is wrong to call something "sexual assault" if there was no physical force or violence applied to the victim.

These torches need some love! The real light source is very large, and the occluder is very close to it: the shadows should be blurry as hell. Up to the point that individual shadow edges would be hard to see. I guess it would look just like mild ambient darkening. With just a light source moving around and ordinary shadows, it looks completely wrong. It looks like that in TDM, but maybe it can be improved in UE4?

@AluminumHaste, I guess I should repeat it: Basically, find the internalVar->floatValue in debugger, and set "Break When Value Changes".

Not sure that Eigen is needed. If you wanted to compute SVD of NxM matrix, then indeed writing it yourself would be a problem. But what problems you can't solve easily with a 4x4 matrix? Do you perform something like Principal Components Analysis? Does 4x4 matrix multiplication pose a problem to anyone? I'm very skeptical about any performance improvements due to this change. Yes, you have 4xfloat and 2xdouble with SSE2, but it seriously matters only when you have arrays of data and do purely numeric operations with them. The submodule is of course a nuisance. For instance, having a submodule breaks GitHub's integration with Hg and Svn: only Git client can fetch them. Luckily, nobody will need to actually work with the submodule, so it won't add much confusion. To me it looks like adding dependency without reason.

Yet another case of mass madness: Remove Stallman: https://rms-open-letter.github.io/ Don't mock Stallman: https://rms-support-letter.github.io/ Basically, people want Richard Stallman to be removed from FSF because of ... some buzz? I tried to find why exactly people went against him, but I could only find a bit of careless words about sexual abuse. In fact, the letter which suggests removing Stallman does not even describe why. It just says "He has shown himself to be misogynist, ableist, and transphobic, among other serious accusations of impropriety. These sorts of beliefs have no place in the free software, digital rights, and tech communities." without any proof.

Yes, most likely. If someone wants to have a look, I suggest setting memory breakpoint on the cvar value. To be honest, I don't see why players should be able to modify these cvars at all, and I guess rare player thinks otherwise

I think @Gadavre uses Linux, since Alt+Tabbing works fine on Windows in any mode.

Ok, you are probably right here. Although I downloaded the same Q2 version and it works fine with OpenGL renderer without any additional actions, it might be some tweak on driver side. On the other hand, I think your extension string is reduced too: do you see ARB_bindless_texture in the list? I don't see it in my list on NVIDIA either. One possible explanation about it is that old games were written in C, and they often copied OpenGL extension string into a fixed-size C string. At that time number of extensions was very low, so the programmers didn't think about how long it will grow in 20+ years. I guess NVIDIA supports more extensions in raw number, meaning that they faced this problem earlier than AMD. Curiously, GL3 introduced a different mechanism for querying extensions in order to force programmers not doing this mistake. The new mechanism is somewhat less comfortable...

Don't see anything there. ref_gl.dll is Quake renderer, it prints version 0.01, which has no relation to OpenGL. OpenGL has version 4.6 Compatibility profile --- that's what any OpenGL application will get unless it specifically asks for Core profile. Version 3.20 is one of the latest versions of quake, and I find it plausible that it was released in 1998. Where is "special compatibility mode", where is "extensions override" ?

It is pretty hard to believe it.

Bindless textures are used as extension, which does not require GL version to go up --- it can stay at 3.3 (even at 1.1 probably) In fact, bindless textures are not in core OpenGL yet: there is no GL version that provides them, extension is the only way. By the way, you can try to enable compatibility profile: cvar "r_glCoreProfile 0" + fresh restart. I'm pretty sure it will print "GL 4.6 compatibility profile" then. But better return cvar back after you try.

Well, in TDM you will see 3.3 Core Profile, like in ioQuake3's "New Renderer". Regardless of which vendor, GPU, and driver you have. UPDATE: And the profile reported by ioQuake in these screenshots is OpenGL profile, I'm pretty sure.

This thread is about different LCC compiler, which handles C++ too. And yes, calling the new compiler "LCC" was a very bad decision in the long run.

As far as I remember, @cabalistic wanted to restore back the resolution-switching code to not betray users habits.

Could you be more precise with this "compatibility profile" term? From OpenGL standpoint, TDM runs under "core profile" now, unless you changed it manually via cvar.

No, you cannot.

How many physical cores do you have? TDM fully uses two threads, plus some minor time is taken by async thread and OpenAL mixing. If you have two cores, then it is not surprising that OBS cannot find processing time. Also there is "Frontend Acceleration", which assigns some job for more threads, probably you should disable it. Did you try setting high priority to OBS in Process Explorer ?