Time Nick Message 06:30 kahrl lol https://github.com/freeminer/freeminer/commit/b7af2868193076985cd03a5196bbf2e1813a1692 06:31 kahrl oh god I'm tainted! I can't contribute to minetest anymore! 06:31 Kray rip 06:34 VanessaE haha 06:51 RealBadAngel should i pull that different bright for each tile thingy into mainstream? 06:52 RealBadAngel VanessaE, ? 06:53 VanessaE hi 06:54 VanessaE imho yes, but as a commit separate from your other work 06:54 VanessaE lest someone complain that you're mixing projects 06:55 RealBadAngel ofc, i wont mix them 06:55 RealBadAngel i found the idea really nice 06:56 VanessaE yes 06:57 RealBadAngel http://i.imgur.com/6GrCtos.png vs http://i.imgur.com/fAzrVVo.png 06:57 VanessaE combined with calinou's idea to put the lighting table back to linear and make moonlight just a bit brighter, I think there's something good here. 06:57 VanessaE oh G*d yes 06:58 VanessaE (I refer to https://github.com/minetest/minetest/issues/1362 ) 07:00 VanessaE er, oh G*d yes refers to the screenshots, the lighting table comment refers to 1362. 07:01 RealBadAngel different bright is made using shaders, but i should be able to easily make it without shaders too 07:01 VanessaE in addition to that, nodebox lighting - particularly nodebox-based light *sources* - needs to be fixed. 07:02 VanessaE RealBadAngel: yes, should be easily doable, so long as blocks are only differently-lighted by one manner or the other and not both at the same time. 07:03 RealBadAngel it adds more depth to the scene, i really LIKE it 07:03 VanessaE me too, that really defines the edges of the blocks nicely now 07:12 VanessaE I might go just a *tad* more subtle on that darker side. Not much, maybe 1% brighter. 07:12 VanessaE but it's really good now 07:19 RealBadAngel http://i.imgur.com/rRYwBxT.jpg vs http://i.imgur.com/pCMwQjy.jpg 07:20 VanessaE RealBadAngel: definitely better with the additional shading 07:20 VanessaE almost looks like it came from...The Game That Shall Not Be Named ;) 07:21 VanessaE it's interesting, you made it affect light sources too 07:21 VanessaE I wonder if that's a bad idea. 07:22 VanessaE light sources should imho not be shaded at all 07:22 VanessaE since they emit light 07:22 VanessaE they should have the same brightness on all faces 07:22 VanessaE (plus whatever is added to them from neighboring sources + the sun) 07:26 RealBadAngel lets see 07:29 RealBadAngel http://i.imgur.com/CUbxNqH.png vs http://i.imgur.com/gmFMjsl.png 07:31 RealBadAngel VanessaE, got the pictures at night? 07:31 VanessaE yes 07:32 VanessaE and I was right 07:32 VanessaE don't shade light sourceds 07:32 VanessaE sources* 07:32 VanessaE make all faces of a light source the same brightness 07:33 RealBadAngel shaders are not aware of that 07:33 RealBadAngel i mean if thats lightsource or not 07:33 VanessaE make them aware :) 07:34 RealBadAngel anyway i just made a simple experiment 07:34 VanessaE it looks good otherwise 07:34 RealBadAngel i asked my gf to compare two pictures, and asked which one is nicer 07:35 RealBadAngel she doesnt play mt and know anything bout coding 07:35 RealBadAngel she picked modified one 07:35 VanessaE yup, the modified one is definitely nicer 07:36 VanessaE looks like another vote for the modified one (see #mintest) 07:36 VanessaE #minetest* 10:35 proller kahrl, real problem in other place, but now server can run some hours more 11:54 celeron55 RealBadAngel: actually shaders at least were aware of light sources when i finished making them 11:54 celeron55 RealBadAngel: i don't know if you have removed the feature though 11:55 celeron55 (i'm pretty sure, altough my memory is not clear) 16:34 BlockMen what are the plans for 0.4.10? is there anything that needs to be done before next release? 16:35 BlockMen or is the plan to stop releases until the android stuff is done? 16:35 BlockMen I hope its not 16:45 sapier guys what do you think about targeting an official 0.4.10 release as of end of this month? 16:46 sapier I guess RBA's shader improvements will be finished by that time and android is ready to merge too 16:47 jin_xi your latest apk works really smoothly on huawai A6 16:54 ShadowNinja Comments? http://sprunge.us/PLSW?diff 16:56 PilzAdam ShadowNinja, why isn't there a core dev's mail address? 16:56 PilzAdam IIRC sfan5 said you could use his 16:57 sfan5 I did 16:58 sfan5 ShadowNinja: looks good otherwise 17:03 ShadowNinja PilzAdam: Because that's the file maintainers address and the file maintainer isn't a core dev. 17:03 ShadowNinja +' 17:03 PilzAdam but how does he maintain the file if he can't push to the repos? 17:03 PilzAdam aren't the core devs maintainer of all the files? 17:03 ShadowNinja He makes PRs. 17:05 sapier hmm could someone fix #1166 .... 17:05 sfan5 ShadowNinja: what if that person suddenly goes away? 17:07 ShadowNinja sfan5: Then the address can be updated. 17:08 sfan5 people usually disappear without notice 17:08 ShadowNinja Maybe something like developers@minetest.net would be good, so it wouldn't depend on any one core dev. 17:10 sfan5 how about just putting celerons email there? 17:14 sapier guess you should ask celeron about that not us ;-) 17:19 ShadowNinja celeron55: ^ 17:20 ShadowNinja Also, the website content area is too narrow for side-by-side diffs on my screen. Thoughts on widening it? 17:21 celeron55 it doesn't imply it's an email address 17:21 celeron55 just put "#minetest-dev @ Freenode.net" there 17:21 LemonLake if you do widen it, make a mobile version, otherwise mobile viewers may have a hard time 17:21 ShadowNinja LemonLake: Widening it should make it easier to view on mobile. 17:21 ShadowNinja The forum is much wider. 17:21 LemonLake fair point 17:21 LemonLake widen it then 17:22 sapier celeron55 what do you think about adding signs support by using irrlichts billboardtextnodes? 17:23 ShadowNinja celeron55: http://people.freedesktop.org/~hughsient/appdata/ indicates that it should be an email. 17:26 celeron55 sapier: i don't think that is what you think it is 17:27 sapier I read the doc as it's some mechanism to place plain text on a 3d pos within a scene ... did I missunderstand itß 17:28 celeron55 the text doesn't haw any angle, it only has a position and always faces the camera; the player names use it 17:28 celeron55 have* 17:28 celeron55 i don't understand why any of this is needed 17:28 celeron55 the image combining works fine 17:28 sapier why does it have a setRotation method? 17:29 ShadowNinja celeron55: It works, but it's ugly and uses up loads of RAM (and CPU). 17:29 sapier the image combining method causes a lot of texture memory being used ... of course that's moders problem but imho in middle to long term we need a better way to do it 17:29 ShadowNinja (/GPU) 17:30 sapier a single sign uses almost as much memory as full minetest_game texture set 17:30 celeron55 well i don't care as long as it's clean and doesn't break how people expect things to work 17:32 celeron55 ShadowNinja: well i don't want my email there, because i won't maintain it 17:32 celeron55 so put someone's who wants to 18:08 Taoki celeron55, PilzAdam, etc: I know there are a lot of bugs and changes on your list already. But please, take a look at this sometime: https://github.com/minetest/minetest/issues/1275 A Lua function one of my mods uses crashes the engine, and is making the mod pretty much unuseable due to this. I reported it some time ago but only BlockMen looked at it a little. 18:09 Taoki Short version: player:hud_set_hotbar_itemcount causes MineTest to crash. 18:09 sfan5 sounds like we already had this reported 18:10 sfan5 ah, it was that one 18:10 celeron55 i might have fixed it 18:10 celeron55 i assume you are testing it with a non-git client 18:12 celeron55 https://github.com/minetest/minetest/commit/8d315347100f1d261bb496b6c73f8cab5d2b7a41 18:12 celeron55 i fixed it because i was developing a game which caused a crash 18:12 celeron55 never seen the issue 18:13 celeron55 Taoki: you should always include a backtrace in that kind of bug reports, btw 18:13 Taoki celeron55: No, I alays run GIT. I can pull and test again in that case 18:13 Taoki celeron55: Ah, no it's not fixed. That commit says 27 april, but I tested again some two weeks ago and it was still there 18:13 sfan5 sapier: the pull you just made needs a rebase 18:14 sapier there's been a lot of changes around that line of code but I kept that fix 18:14 celeron55 Taoki: build a debug build and get a backtrace 18:15 Taoki ok. I forgot how to even get backtraces 18:15 Taoki There was a special program IIRC 18:16 sfan5 psst.. It is called gdb 18:16 Taoki right 18:17 Taoki Building a debug version now 18:17 Taoki What was the GDB command to get the backtrace once the crash takes place? I remember that you start GDB, then choose to run the file, then save the log manually or something 18:17 sapier could someone plz fix head it doesn't even build right now 18:17 Taoki *run the bin 18:18 sapier CMake Error at CMakeLists.txt:176 (install): 18:18 sapier install FILES given no DESTINATION! 18:18 sfan5 Taoki: bt full 18:18 sapier 176 is appdata 18:18 Taoki Thanks. Will test within the next minutes 18:18 sfan5 sapier: that shouldn't happen 18:19 sapier yes I agree 18:19 sapier that shouldn't happen ;-P 18:19 sfan5 APPDATADIR is set in an elseif(UNIX) and is used in an if(UNIX) 18:19 sapier but it does 18:19 sapier I don't wanna reclon 18:19 sapier e 18:21 sfan5 cannot reproduce with freshly cloned 18:21 sapier you forgot run in place case 18:21 sfan5 $ rm -Rf CMakeCache.txt CMakeFiles 18:22 sfan5 should fix it 18:22 sapier no wont 18:22 ShadowNinja Hmmm... 18:22 sfan5 sapier: no? I just tested it and it fixes it for me. 18:22 sapier you don't set this variable in RUN_IN_PLACE case but use it 18:22 sapier how is it supposed to work ? ;) 18:23 sapier add set(APPDATADIR "share/appdata") to run in place case and it's gonna work 18:24 ShadowNinja Is it APPDATADIR? 18:24 ShadowNinja Er, too late. 18:24 sapier it's done this way for XDG_APPS too 18:26 Taoki celeron55: Got it! http://pastebin.com/raw.php?i=ix2fQfV8 18:26 sapier this seems to be quite familiar 18:27 sapier wasn't there an issue like that before? 18:27 Taoki I still have the gdb session / crashed minetest process open BTW, if you need me to run more gdb commands 18:27 celeron55 good bt, it's probably fixable based on that 18:27 Taoki ok. I'll close it then and re-compile Release 18:28 Taoki I don't compile Debug because I heard it's slower, so I have to re-build from scratch each time 18:30 sfan5 Taoki: you can also compile RelWithDebInfo to get debug info while keeping the "Release" speed 18:30 Taoki Nice... didn't know about that 18:30 Taoki I should probably used that by default yes 18:31 Taoki If it's not any bit slower 18:32 ShadowNinja Comments? http://sprunge.us/XDBa?diff 18:33 sapier can you do a pull request so we have at least a chance to review? 18:33 sfan5 Taoki: it does not give as much debug info as Debug 18:34 sfan5 so... 18:35 sfan5 who is going to fix building current master now? 18:38 sapier hmm I guess the one asking ;-) 18:45 ShadowNinja sapier: You have a chance to review it now. 18:46 Taoki celeron55: Does that explain what's causing the crash? Is a fix easy to find? 18:47 sapier I won't review almost 500 lines of code change without any explanation what it's supposed to do 18:48 sapier you don't have to write a novel ;-P 18:49 celeron55 Taoki: fixing that particular crash is easy, but evidently there could be more similar ones lurking somewhere which could be easier to tackle with some kind of redesign of the interface, but dunno 18:49 Taoki I see 18:49 celeron55 i'm fine with just a simple if() there though 18:50 Taoki Maybe you can fix that particular one, and add a code comment there so others can know similar problems might be related 18:50 sapier we fixed bugs looking exactly like this one before, guess we didn't fix it completely 18:50 Taoki So at least my mod and similar ones can work for now 18:50 celeron55 the underlying problem is that the server can tell the client that some item index is being used, while the inventory isn't actually large enough for that index to actually exist 18:51 celeron55 actually 18:51 celeron55 in that case it's the client that chooses the index? 18:51 Taoki celeron55: Strange part is that even if I change the inventory size before I change the hotbar size (to be sure hotbar doesn't show more than the inventory itself), the crash still happens. 18:51 celeron55 well anyway, someone picks an index too large 18:51 Taoki My script does a set_inventory_size then a set_hotbar 18:51 celeron55 Taoki: that's about wielded item 18:52 Taoki ah 18:52 celeron55 not about hotbar drawing 18:52 Taoki I see. I kept having the idea it's hotbar drawing that caused it, since set_hotbar is the trigger 18:54 Taoki celeron55: But yeah, I'd appreciate any fix so my mod can work and others can finally test it and let me know what to change and all. If it's not the best fix, maybe add a code comment about it too, like a TODO 18:54 Taoki Not sure if my mod uses it in a very special circumstance 18:56 celeron55 not sure if i'll fix it now, i'm not interested in setting up a test case myself or asking you to test guessed fixes 18:56 celeron55 maybe someone else will 8) 18:56 Taoki ok. I pasted the BT on the bug report as well 18:58 Taoki Maybe you can post a comment there with a description as to what is happening if you conclude you shouldn't fix it now, to make it easier for other devs 19:02 celeron55 did that 19:03 Taoki Thanks 19:08 Taoki PilzAdam, sapier, BlockMen: Any chance you could look at it please? I'm here now and can test any changes to make sure the crash is gone 19:09 BlockMen ShadowNinja, your changes would break mods that use default.grow_tree() 19:10 ShadowNinja sapier: I used format-patch. The description is in the header. 19:11 ShadowNinja BlockMen: It was a broken and updocumented API. 19:11 sfan5 pushing http://ix.io/cWS in 5 mins 19:12 BlockMen ShadowNinja, so? why not checking for data and pass it to new function if present 19:13 ShadowNinja if isData(pos) then error 'default.generate_tree.' end 19:13 ShadowNinja +Use of 19:14 sapier https://forum.minetest.net/viewtopic.php?f=3&t=9389 please test the latest build. There are just a few remaining glitches. I'm not even sure if it's absolutely necessary to fix them right from beginning 19:14 ShadowNinja type(pos) == "userdata" should do. 19:14 ShadowNinja It might have a custom name though. 19:15 BlockMen in general: why is it not in builtin and called minetest.grow_tree()? 19:18 ShadowNinja BlockMen: http://sprunge.us/cXEW?diff 19:18 ShadowNinja BlockMen: Becuase treegen isn't a core part of Minetest. It isn't needed in things like space games. 19:21 BlockMen ShadowNinja, s/invalid/deprecated 19:21 BlockMen and just because some games might not use it is no valid reason. 19:22 BlockMen there might be games that dont use hotbar/statbars 19:22 BlockMen but they are aswell in builtin 19:24 ShadowNinja BlockMen: But statbars aren't specific to a specific game. Trees are specific to earth-like games. 19:26 BlockMen ShadowNinja, no. you could use trees for space like games aswell 19:26 BlockMen e.g. moon-bases 19:39 ShadowNinja BlockMen: *could*. Still earth-specific. 19:40 ShadowNinja (Since such an environment would emulate earth) 19:43 BlockMen ShadowNinja, "could" because it is an example. air is also earth specific. you want remove that now aswell? 19:56 Taoki Ugh... minetest.set_player_privs is broken 19:56 Taoki It doesn't do anything apparently 20:13 Taoki Reported here: https://github.com/minetest/minetest/issues/1366 And no, it's not due to singleplayer having base privs... it doesn't work for "fly" either 20:22 BlockMen ^ closed it 21:13 RealBadAngel celeron55, i havent removed that code, its still in 22:04 RealBadAngel ShadowNinja, fixed the style issue in #1303 , now waiting for travis 22:04 ShadowBot https://github.com/minetest/minetest/issues/1303 -- Unite nodes shaders. by RealBadAngel 22:09 VanessaE Can someone please update this to account for the mgv6_spflags setting? http://wiki.minetest.net/Map_generator/settings 22:13 RealBadAngel if there are no more objections i will merge that pull in a few hours