Time Nick Message 14:53 ANAND MT seems to mix up normalmaps and bumpmaps in a very confusing way 14:54 ANAND enable_bumpmapping should be called enable_normalmaps, AFAIK. paramat said MT already supports per-texture normalmaps. 14:56 ANAND I haven't checked the code yet, but MT most likely generates *bumpmaps* based on per-pixel brightness, and not normalmaps 14:57 ANAND So it should be "generate_normalmaps" instead :/ 14:57 ANAND er... "generate_bumpmaps"* sorry 15:44 VanessaE ANAND: no, they are actually normalmaps. 15:45 VanessaE they just aren't very good because MT has no true directional lighting and of course flat textures don't have much in the way of 3d "cues" that the normalmap extract code could use, so it fakes it. 15:46 VanessaE RBA and I worked on that for a while before we were satisfied with what MT could do at the tiem. 15:46 VanessaE time* 15:47 VanessaE devs back then were dead set against anything that could improve MT's visuals if they thought it could give someone a gameplay advantage, however slight. 15:47 VanessaE (so advanced, moddable shaders were a "hell no") 15:47 rubenwardy heh 15:48 VanessaE am I wrong? 15:49 rubenwardy you're not wrong 15:49 rubenwardy Minetest's philosophy on shaders is annoying 15:49 rubenwardy shaders are essential in rendering pipelines, and shouldn't be considered fancy extras 15:49 rubenwardy however, fancy shaders could be considered fancy extras 15:50 rubenwardy using shaders for something like lighting is what they're intended for 15:50 VanessaE I wouldn't mind RBA's lava shader and kilbith's water shader making it in, as long as they can be turned off (I'd use them only in my HDX profile) 15:50 rubenwardy well, those are both fairly intensive so should be turn-off able 15:50 rubenwardy but I'd support their addition 15:50 rubenwardy I've posted the video of what I'd like Minetest to look like many times 15:50 VanessaE RBA's lava shader is one of these two: https://gitlab.com/VanessaE/lava_surface_shader ~~ https://gitlab.com/VanessaE/shaded_lava 15:51 VanessaE (yeah, I just happen to have a fork of his last-working code, felt like I should preserve it) 15:51 rubenwardy hmmm, how can I test lots of formspecs at once 15:51 rubenwardy I guess I want a server with UI and technic 15:52 rubenwardy :eyes: 15:52 VanessaE signs, unified dyes (airbrush) 15:53 VanessaE rubenwardy: you should take those two repos, reconcile 'em, make the result usable again. RBA would be happy :) 15:55 VanessaE https://www.youtube.com/watch?v=v52f-1YsSAM 15:55 VanessaE found it 15:55 VanessaE as I recall, it uses an animated normalmap to get that effect 15:55 VanessaE in like 2 layers 15:55 rubenwardy ah yeah 15:55 VanessaE kinda like google earth's water shader effect up close 15:58 VanessaE I have his water surface shader also, in another repo, but imho kilbith's looks better 15:59 VanessaE another red flag at the time was the fear of having to hard-code the shader to work with one specific node def. that'd be another use-case for moddable shaders. 16:00 * DS-minetest wonders how kilbith's water reflection works and whether you can see fragments in the reflection that would otherwise not be seen because of depth-testing 16:01 ANAND VanessaE: interesting 16:01 ANAND Yea our current shaders situation is messed up 16:02 VanessaE it all boils down to: MT should be able to run 100% on Intel laptop graphicds 16:02 VanessaE -d 16:02 VanessaE :P 16:02 DS-minetest from before 20 years 16:02 ANAND That matches exactly what I run MT on :D 16:02 VanessaE just about, yeah. 16:03 ANAND Running MT on such low-end specs enhances all the existing bottlenecks, making them somewhat easier to spot :P 16:04 VanessaE heh 16:04 VanessaE that's fine if you're debugging. 16:04 VanessaE definitely not fine if you're *playing* 16:05 ANAND Exactly 16:06 DS-minetest is there a minimum opengl version that minetest requires for enabling shaders? 16:06 Calinou a large part of players on indie games use Intel graphics, it would be a big mistake to bar them from playing 16:06 Calinou only AAAs can afford to require a dedicated GPU these days :) 16:06 rubenwardy I use Intel graphics 16:06 ANAND DS-minetest: We use OpenGL 1.2 16:06 ANAND uh, GLSL* 16:07 VanessaE yeah it's more a matter of OpenGL versions, not so much the underlying hardware. 16:07 VanessaE even a junky old Intel could do most stuff, just slower. 16:07 VanessaE if the GL version were up to the task 16:16 VanessaE also to be honest, I'm pretty sure the vast majority of PCs have at least some low-end dedicated GPU, and it's not at all unreasonable to expect that anyone interested in playing (ACTUALLY playing, not just coding-for, modding-for, etc) on a PC will have a reasonably good GPU, even if it's old. 16:16 VanessaE (I.e. who gives a shit if tablets/phones could do any better than they currently do, just leave them as they are) 16:18 VanessaE Hell, I'm not even close to "serious gamer" grade, and yet even I have a good GPU 16:18 ANAND I just have "Intel HD Graphics". That's it. It doesn't even have a cool number after the name :( 16:19 VanessaE what's stopping you from getting something better? 16:19 VanessaE even a cheap $50 GPU would be a big upgrade :P 16:19 rubenwardy you will have a cool number 16:19 ANAND Cost, ofc 16:19 rubenwardy VanessaE: I use laptop 16:19 rubenwardy I'm currently in a coffee shop, can't take desktops there 16:19 rubenwardy well, you could 16:19 VanessaE rubenwardy: you know full well that laptops were never truly meant for gaming. 16:20 VanessaE not high performance games that is 16:20 VanessaE and MT needs that. 16:20 rubenwardy you can still game though 16:20 rubenwardy and it's more than enough for Minetest 16:20 ANAND Over here, a GPU that'd actually be noticeably better than my iGPU would cost a LOT 16:20 rubenwardy also, graphics cards and Linux don't mix at all 16:20 VanessaE um 16:20 ANAND ^ 16:20 VanessaE no 16:20 rubenwardy I could have spend £200 more to get a graphics card 16:20 VanessaE that's untrue. 16:21 rubenwardy but it would've have made the system unstable 16:21 ANAND Drivers are hard to find 16:21 VanessaE I haven't had problems with video drivers in ages. 16:21 ANAND Although the situation is getting better these days 16:21 ANAND Esp. the red side 16:21 rubenwardy Linux doesn't support swapping between integrated and dedicated cards 16:21 rubenwardy very well 16:21 rubenwardy I forget the name for it 16:21 VanessaE well anyway 16:21 VanessaE THIS is the resistance I was talking about 16:22 rubenwardy ah 16:22 rubenwardy discrete graphics 16:22 rubenwardy what is the resistance? 16:22 rubenwardy also, my laptop can run other 3d games with reflective water shaders fine 16:22 VanessaE rubenwardy: in short, 16:22 VanessaE I only have $CRAPPY_LOW_END_MACHINE so fuck everyone else who has something better and wants better use of their hardware. 16:22 rubenwardy my laptop is high end 16:23 ANAND lol, resistance? We're just discussing about low-end graphics cards :P 16:23 VanessaE a laptop is never high end. 16:23 rubenwardy it has better specs than a flag ship macbook 16:23 rubenwardy laptops certainly can be highend 16:23 Krock a laptop can be high-end but they thermal throttle way too fast 16:23 rubenwardy yeah ^ 16:23 rubenwardy that's the problem with a thin laptop 16:23 Krock hence only real high-end if powered within a freezer 16:23 VanessaE by today's standards, my PC is low-end, but I doubt most laptops today could beat it in the long term 16:23 rubenwardy anyway, I never said that I'd oppose graphics features 16:24 VanessaE rubenwardy: no, 16:24 rubenwardy I support graphics features 16:24 VanessaE but that IS the resistance I was talking about. 16:24 rubenwardy what resistance? 16:24 VanessaE it has not changed in the entire time I've been into MT. 16:24 rubenwardy also, my CPU could probably beat yours if yours if fairly old 16:24 rubenwardy my laptop has an i7 @ 4GHz, which makes it good for compiling 16:25 rubenwardy 8th gen 16:25 VanessaE Phenom II X6 1055T (2.8 GHz x 6 cores) 16:25 rubenwardy but because of the thermal thottling, long running things like gaming will reduce performance 16:25 VanessaE with good cooling, so it never throttles 16:26 rubenwardy devs should have a mixture of hardware to be able to optimise for different devices 16:26 VanessaE indeed 16:26 Krock Q6700, 2.66 GHz. One of the good CPUs to heat your cold rooms 16:26 rubenwardy I think the problem with graphics in Minetest is more that we're clueless 16:26 VanessaE but that's a matter of money and whether they care :P 16:26 rubenwardy and performance ends up being killed for fancy features 16:26 VanessaE clueless? well, you said it, not me :) 16:27 rdococ lol 16:27 rubenwardy because RBA enables tangent space for all devices, not just ones using normal maps 16:27 VanessaE RBA used to say similar things though 16:27 ANAND Our existing shaders structure forbid us from actually making small improvements 16:27 rubenwardy RBA broke performance several times 16:27 rubenwardy but then fixed it 16:27 Krock the currently active developers don't really have a plan of shaders. what kilbith showed there would be some amazing stuff to have, but yet no PR 16:27 VanessaE yes, he did 16:27 Krock assuming it's a real shaders 16:27 VanessaE Krock: there'll never be a PR. 16:28 VanessaE all his work is in some commercial fork 16:28 rdococ commercial fork? 16:28 VanessaE of course, the source has to be available somewhere, but good luck getting a usable PR out of him 16:28 ANAND kidscode uses the same player models and textures? 16:29 VanessaE idk anything about it, really. 16:29 rubenwardy we need people to work on improving our graphics pipelines in more ways than one 16:29 rubenwardy VanessaE: he said that was in MT, he could intend to contribute it back 16:29 DS-minetest (a PR would be useless, it wouldn't be merged) 16:30 VanessaE rubenwardy: I'll reserve judgment... this IS kilbith after all :) 16:30 VanessaE DS-minetest: that too. 16:30 rubenwardy I suppose what we need is not just contributors good with graphics, but at least one developer 16:31 ANAND What's the difference? 16:31 DS-minetest one approval is not enough 16:31 Krock I assume he meant "dedicated core developer working 25/7 on Minetest's graphics stuff" 16:32 rubenwardy developers can review changes, contributors can't 16:32 rubenwardy How do I go from craft guide to the creative picker in UI? 16:32 rubenwardy not all reviews are the same 16:32 ANAND Ah, for approvals 16:32 rubenwardy at least one review needs to be informed, but the second review could be more on code structure or testing 16:32 DS-minetest click on that tool buttons 16:33 rubenwardy I think it's because I timed out 16:34 rubenwardy the only real difference between developers and contributors is that developers can actually approve PRs 16:34 * rdococ wishes he was more skilled with C++... and that C++ was less flawed 16:36 rubenwardy my laptop cost more than a gaming PC 16:36 rubenwardy as is the way 16:38 Krock rdococ: what flaws 16:38 rubenwardy templates, pointers, many ways of doing things 16:39 DS-minetest that are just things that cause the programmer to make flaws because he/she is not smart enough 16:40 rdococ unless that's sarcastic, that's a not smart thing to say 16:41 rdococ programming languages should help the programmer to avoid making flaws 16:41 rubenwardy saying the "programmer isn't smart enough" is a pretty new programmer thing to say 16:41 rubenwardy programming languages should be crafted to make it harder to create bugs, to make maintenance and testing easier 16:41 rubenwardy and code, too 16:41 Krock you can overcomplicate everything if you want to 16:42 rdococ rubenwardy: that's what I said, just stated less eloquently :P 16:42 rdococ err, by that I mean what I said was stated less eloquently 16:42 rdococ dammit english, you and your ambiguous grammar 16:42 Krock (num >= 0 && (sign = true) || (sign = false)); 16:43 Krock Lua-style conditional expression but with assignments 16:43 rubenwardy I hate assignments like that 16:43 rubenwardy also 16:44 rdococ (cond and (function () ... return true end)() or (function () ... end)()) 16:44 rubenwardy sign = num >= 0 ? true : false; 16:44 rubenwardy sign = num >= 0; 16:44 rubenwardy :D 16:44 rubenwardy I know you were purposely being obgtuse 16:44 * rdococ thinks about what his dream combination of functional and imperative programming would be like 16:45 rdococ it wouldn't be like Scala, that's for sure :P 17:12 rubenwardy argh, why is VanessaE's server model sold out 17:12 rubenwardy I blame you, VanessaE, for not mentioning the sale sooner 17:12 rubenwardy gosh 17:18 rubenwardy 2019, wifi support on Linux still sucks 17:19 ANAND ikr 17:19 Krock Works On My Machine (C)(R)(TM) 17:20 ANAND I bought a WiFi adapter last year, and it barely works without the manufacturer-provided driver 17:21 ANAND While they do have a Linux driver (to my surprise), it was made for an old version of the kernel, and doesn't compile on newer ones >.> 17:21 rdococ just about works on my shoddy laptop thing (C) 17:22 Krock Bus 001 Device 003: ID 07d1:3c16 D-Link System DWA-125 Wireless N 150 Adapter(rev.A2) <<-- can confirm working 17:22 VanessaE rubenwardy: I didn't know about the sale until the day I mentioned it, and I *did* mention my server model before :) 17:27 rubenwardy Jordach recommended digital ocean 17:27 rubenwardy then AWS and Scaleway 17:27 rubenwardy thanks Jordach 17:32 rubenwardy banned by Google play 17:32 rubenwardy if on Android, I'd report the app 17:47 Krock so I'm trying to call the notification dbus functions but this bad boy already did it https://stackoverflow.com/questions/8846671/how-to-use-a-variant-dictionary-asv-in-dbus-send/38573254#38573254 17:47 Krock nice thanks. now I can just cope&paste 18:15 rubenwardy It would be cool to have posters like these: https://supertuxkart.net/Posters 18:58 rubenwardy Would anyone be willing to help me run a stall at FOSDEM in Brussels, Belgium 1-2 Feb? 19:47 Fixer those posters are nice, but I'm not good with this