Minetest logo

IRC log for #minetest, 2019-02-17

| Channels | #minetest index | Today | | Google Search | Plaintext

All times shown according to UTC.

Time Nick Message
00:00 swift110 joined #minetest
00:04 SHINIGAMI joined #minetest
00:06 Tux[Qyou] joined #minetest
00:12 cheapie Ow, my eyes... remind me what exactly the "tone mapping" setting is supposed to be for, again?
00:22 paramat websearch 'filmic tonemapping'
00:24 cheapie paramat: So it's supposed to make it look more like a movie or something? These results aren't real helpful.
00:25 cheapie I'll admit I haven't watched a movie in ages, but I don't remember them being insanely bright with everything washed out.
00:26 paramat specifically its 'Hable's UC2 filmic tonemapping' one tone mapping standard
00:27 cheapie I keep seeing references to HDR and stuff... does this require hardware I don't have and that's why it looks bad?
00:27 JDCodeIt Anyone know why these functions don't seem to be available in lua when writing mods? They are listed in lua_api.txt - minetest.get_humidity(pos) - Seems one needs to get the perlin params then call perlin noise directly
00:28 paramat http://filmicworlds.com/blog/filmic-tonemapping-operators/
00:28 paramat no it doesn't require anything
00:28 paramat just colour post-process
00:28 cheapie paramat: Yes, I found that, and it seems to show different versions of it, but it doesn't say what tone mapping is in the first place, what it's supposed to do, or why you would want it.
00:30 paramat HDR = high dynamic range
00:30 paramat see 'tone mapping' on wiki
00:30 paramat -pedia
00:32 cheapie So it's supposed to make things more realistic-looking?
00:33 cheapie If so, it must be broken, because it completely ruins colors when I turn it on.
00:33 paramat i don't use it but many like it. MTG textures may be partly why it looks bad to you
00:33 paramat if broken post an issue
00:34 paramat should look like this https://github.com/minetest/minetest/pull/3612
00:34 cheapie OK, so it's supposed to look that bad :P
00:39 paramat heh
00:40 paramat at least it's not a bug
00:59 paramat JDCodeIt maybe test with the code i tested the PR with? https://github.com/minetest/minetest/pull/6455
00:59 paramat it worked when i tested it
01:05 paramat they might not work in mgv6 though
01:07 paramat as that doesn't use the biome API
01:13 paramat left #minetest
01:25 paramat joined #minetest
01:29 sveta thanks for adding the cern paragraph to the minetest draft article at wikipedia, and for mentioning it paramat jluc, I did some formatting fixes for it. paramat the opensource.com link is already there
01:30 sveta there is several research publications which involve using minetest in a class once where they demonstrate some success which is good
01:31 sveta but finding some places where it is offered continuously every year would be even better
01:31 sveta (link: https://en.wikipedia.org/wiki/Draft:Minetest for those who weren't here a few hours ago)
01:49 Lone-Star tone mapping does nothing for default MTG textures. if using a good texture pack, it makes it really stand out
01:50 Dr-Frankenstone joined #minetest
01:53 Dr-Frankenstone Is there a way to get "ignore" nodes into a .mts? I want the .mts to overwrite nodes in the world, but because the empty nodes in the mts are set to air it removes everything in the world inside its bounds.
01:53 Dr-Frankenstone The schematic loading function converts ignore into air - https://github.com/minetest/minetest/blob/master/src/mapgen/mg_schematic.cpp#L292
01:54 PND joined #minetest
01:54 Dr-Frankenstone but the schematic placing function knows not to overwrite nodes with ignore - https://github.com/minetest/minetest/blob/master/src/mapgen/mg_schematic.cpp#L144
01:55 Dr-Frankenstone so if there's a way to have 'ignore' nodes in the .mts, then it should work
01:55 paramat yes .mts can contain ignore nodes, they're just never placed
01:56 Dr-Frankenstone how are ignore nodes specified in the .mts so that Minetest doesn't convert them to air?
01:56 Lone-Star think there is a flag you can set to replace all nodes on place.
01:56 paramat actually, until now i've recommended using 'air, prob = 0' for 'void' in schematics, but using 'ignore' for void is better, and is slightly faster during schematic placement
01:57 Lone-Star this may be a worldedit command though
01:58 paramat to avoid a node in a schematic overwriting a node in the world, set it's 'prob' to 0. you may have to create the schematic using a lua table to do this, but some schematic saving mods can set per-node prob too
01:58 Dr-Frankenstone ahh, that might be the clue I need. I'm trying to fix up the old script from https://forum.minetest.net/viewtopic.php?id=6007 , I didn't realise I could specify a probability for each material
02:01 paramat that deserialize function may not be ideal, i need to study that
02:03 JDCodeIt Using 0.4.17 stable branch I did not see the function - i did print(dump(minetest)) to a file, but the functions are not listed - did this PR #6455 not make 0.4.17 stable branch?
02:03 ShadowBot https://github.com/minetest/minetest/issues/6455 -- Add 'get heat', 'get humidity', 'get biome data' APIs by paramat
02:06 paramat i misunderstood. using 'ignore' in a schematic may not be a good idea, or, perhaps it should be 'ignore prob = 0'
02:09 paramat probabiliites are per-individual node, not per-node type, see MTG lua tables https://github.com/minetest/minetest_game/blob/master/schematic_tables.txt
02:10 paramat get biome data/heat/humidity is a feature so wouldn't have been added to 0.4.17
02:10 paramat 0.4.17 was bugfixes only
02:10 JDCodeIt Ok, got it. Looking forward to a stable 5.0 :-)
02:13 paramat ok, looks like it may be best to use 'air, prob = 0' for non-placed void in schematics
02:20 swift110 joined #minetest
02:21 paramat the mod 'schemtools' can save a schematic from the world, then convert it to a lua table for further editing
02:23 paramat then you can alter air to air prob 0 to not be placed
02:32 Dr-Frankenstone Thanks paramat. I think I'll do that, this old script is a bit primitive. Fixing it up in a lua table is probably the way to go.
02:35 Linda_Wolfy joined #minetest
02:35 Linda_Wolfy J
02:35 Linda_Wolfy W
02:35 Linda_Wolfy S
02:35 Linda_Wolfy S
02:35 Linda_Wolfy D
02:35 Linda_Wolfy E
02:35 Linda_Wolfy A
02:35 Linda_Wolfy S
02:35 Linda_Wolfy D
02:35 Linda_Wolfy E
02:35 Linda_Wolfy A
02:36 paramat ^ VanessaE
02:39 Emerald2 joined #minetest
02:44 sergiuslane joined #minetest
02:45 sergiuslane thanks for this forum..https://downloadlagu-mp3.club/
02:45 Ronika joined #minetest
02:45 Ronika Shush
02:48 Miner_48er joined #minetest
02:49 Lone-Star the nuts are out tonight
02:51 puzzlecube joined #minetest
02:53 swift110 joined #minetest
02:55 MinetestBot [git] paramat -> minetest/minetest: lua_api.txt: Various additions and improvements (#8245) 00a22ec https://git.io/fh5gb (2019-02-17T02:52:53Z)
03:00 ichoquo0Aigh9ie joined #minetest
03:06 PNd joined #minetest
03:07 ichoquo0Aigh9ie joined #minetest
03:12 JDCodeIt left #minetest
03:36 ichoquo0Aigh9ie joined #minetest
04:07 Cornelia joined #minetest
05:05 Dr_Frankenstone joined #minetest
05:09 Wikiwide joined #minetest
05:10 Wikiwide Minetest crashes with error about write_F1000 function : assertion i is between F1000_Min and F1000_max failed. What could it be?
05:11 sveta unload all modules, check again
05:11 sveta is it easy to reproduce
05:20 swift110 joined #minetest
05:46 Wikiwide Sveta: easy to reproduce, yes. Especially near a particular location.
06:00 kaeza joined #minetest
06:00 SanskritFritz joined #minetest
06:03 Ruslan1 joined #minetest
06:07 Ruslan1 Hello
06:46 Emerald2 How about finishing/polishing up some of the games and shipping them with Minetest? MTG is really basic after all.
06:50 Tux[Qyou] joined #minetest
07:24 SanskritFritz joined #minetest
07:25 Weedy joined #minetest
07:28 SanskritFritz joined #minetest
07:28 CWz joined #minetest
07:31 SanskritFritz joined #minetest
07:35 Dr-Frankenstone joined #minetest
07:37 DrFrankenstone joined #minetest
07:39 SanskritFritz joined #minetest
08:15 Gael-de-Sailly joined #minetest
08:48 Krock joined #minetest
09:09 Beton joined #minetest
09:21 YuGiOhJCJ joined #minetest
09:38 Dr_Frankenstone joined #minetest
09:40 Dr-Frankenstone joined #minetest
09:56 Hijiri I think the general attitude is that it's a not a specific priority of MTG or engine team, and that the game developer would need to make it ready for inclusion and propose it
10:08 kaeptmblaubaer joined #minetest
10:19 aheinecke joined #minetest
11:00 Fixer joined #minetest
11:26 Taoki joined #minetest
11:39 svetlana_ joined #minetest
11:48 kaeza joined #minetest
11:57 Krock joined #minetest
12:13 sveta joined #minetest
12:26 calcul0n joined #minetest
12:33 jluc joined #minetest
12:47 kaeza joined #minetest
12:51 svetlana_ joined #minetest
12:59 fwhcat joined #minetest
13:23 MindUser0_0 joined #minetest
13:23 MindUser0_0 joined #minetest
13:25 entuland joined #minetest
13:27 BillyS joined #minetest
13:31 BillyS joined #minetest
14:03 calcul0n joined #minetest
14:27 Nametest joined #minetest
14:34 Nametest Hello. Anyone knows what are the corner node positions of a world mapblock? For example, 0,0,0 and 16,16,16?
14:35 Krock well, depends on where the mapblock is
14:35 Krock node_pos = mapblock_pos * 16
14:35 galaxie left #minetest
14:36 Krock whereas node_pos here is the minimal position inside the mapblock (i.e. 0,0,0)
14:36 Nametest Hmm... so that means mapblock_pos = node_pos % 16  ?
14:37 Nametest Sorrry, I meant integer division, not %
14:37 Krock yes
14:37 Nametest Okay. Thanks!
14:37 Krock so in Lua you'll have to use math.floor
14:38 Krock !next
14:38 MinetestBot Another satisfied customer. Next!
14:56 p_gimeno what's the reported Y position in the F5 screen? is it the position of the camera?
14:58 Nametest It's the position of the player's feet, I think.
15:00 Taoki joined #minetest
15:00 Nametest The camera position should be player:get_pos() + player:get_properties().eye_height + player:get_eye_offset()
15:05 Krock p_gimeno: feet + 0.5m
15:05 Krock don't ask me why
15:05 Krock err lol
15:05 Krock it's exactly the feet position
15:06 Krock Nametest: not quite. the eye offset is rotated with the player
15:06 Sokomine any ideas how to turn a nodebox (looks pretty ok as inventory image) into a texture i could use for the corresponding wield item?
15:07 Krock uhm.. why not just use the wielditem entity type?
15:07 Krock Minetest does calculate that automatically
15:07 Nametest I hate rotations. They make the math so much worse.
15:08 Sokomine does that work for entities as well?
15:08 Krock Sokomine: well, it's what you get with dropped items
15:09 Nametest ... do you want the wield image to be the same as the inventory image, or?...
15:09 Krock Nametest: until there's a magic function called rotate_vector(v, pitch, yaw, roll)
15:11 Nametest Krock: rotating entities by roll axis also rotates their yaw and pitch axis so if you want to set yaw, pitch and roll separately... you can't
15:11 Nametest Fortunately everything is a cube
15:11 Nametest ... or cuboid shape
15:11 p_gimeno Nametest: RC1?
15:12 Nametest p_gimeno What is RC1?
15:12 p_gimeno Minetest 5.0.0 Release Candidate 1
15:13 shangul joined #minetest
15:13 Nametest Ah, well, I'm using the latest 5.0.0, so... probably yes?
15:13 p_gimeno latest released? or latest from github self-compiled?
15:14 shangul in other words the latest stable or just the latest?
15:14 Krock latest 5.0.0 means latest in-development build, most likely current HEAD
15:14 Krock since there was no stable 5.0.0 yet
15:14 p_gimeno roll is broken in set_rotation in the released RC1, hopefully that will be fixed in today's release (if there's one)
15:15 p_gimeno which was where I was getting at :)
15:16 Nametest Latest stable that passed. I'm probably behind as of now, but it should be at most a week old. I don't compile from source, not enough cost-benefit
15:16 Nametest However, I spent yesterday "fixing" my rotation code so that set_rotation did what I wanted it to do... and now you tell me it's a bug...
15:17 p_gimeno sounds like you're affected by #7927
15:17 ShadowBot https://github.com/minetest/minetest/issues/7927 -- get/set_rotation and get/set_yaw problems
15:19 Nametest Amazing that set_rotation is such a recent addition. I would think it would be part of the fundamental entity toolset
15:19 p_gimeno for 5.1 there will (hopefully) be a relative rotate()
15:21 Nametest From what I have gathered, the vector is actually X=pitch, Y=yaw, Z=roll, but roll also rotates the pitch and yaw axis such that with Z=90, X=-yaw and Y=pitch
15:21 Krock pitch & yaw alone would probably be easier to deal with
15:21 Krock but then you need some more code to avoid locks
15:22 shangul Which key is used to change the appearance of nodes?
15:22 Nametest So, my own solution was to set pitch, yaw and roll independently, and then use the following code:
15:22 Nametest v.x = v.x * math.cos(v.z) - v.y * math.sin(v.z)
15:22 shangul I used to use a key and then appearance of nodes was changed, crosses or something instead of textures
15:22 Nametest v.y = v.x * math.sin(v.z) + v.y * math.cos(v.z)
15:25 p_gimeno Nametest: the way it's designed in the underlying engine, X is roll and Y is pitch, but they behave as pitch and yaw respectively due to Minetest's choice of world axes, as long as Z is kept at zero. In the underlying engine Z is yaw, but with MT's axis conventions it causes a mess.
15:26 Krock shangul: F5 with the debug privilege makes them appear as a grid
15:26 Krock press it like 4 times
15:26 Sokomine or rather: can someone do me a favor and turn the image as created by mt into a texture with the necessary transparency and not too high a resolution?
15:26 Nametest lol. That's amazing p_gimeno
15:27 Nametest Sokomine: have you heard of [inventorycube{<top>{<left>{<right> ?
15:27 shangul Krock, thanks
15:27 Krock !next
15:27 MinetestBot Another satisfied customer. Next!
15:27 Krock Nametest: cube. does not work for any other shape
15:28 shangul :D
15:28 Nametest Right... I forgot.
15:29 p_gimeno Krock: back to the vertical position: if it's feet position then I don't get why the feet position is 0.5 when in contact with the node. Horizontally, 0.5 is the node centre, which makes sense. It would make more sense if it was feet + 0.5m as you first said.
15:29 Sokomine Nametest: isn't that for cubes only? can it handle nodeboxes?
15:29 Krock p_gimeno: the node centre is 0,0,0
15:29 Krock extent is then 0.5m in each direction
15:29 Nametest Sokomine: You can take a screenshot of the nodebox in question with high-saturation, uncommon-hue nodes behind (as in bright-magenta nodes). Then it is a very simple job of removing said color
15:29 p_gimeno Krock: oops, that's right (and WTFy)
15:30 Krock and you're standing on the top side of the node, which is 0.5m above the node centre, usually.
15:30 Krock it's indeed confusing at the first glance
15:30 Nametest Nodeboxes are center-aligned to the grid. Entities usually have their feet as the origin
15:31 Nametest I meant nodes*.
15:31 Sokomine nametest: that involves texture editing. i'm very bad at that :-( image manipulation tools seem to conspire against me
15:31 Nametest Sokomine: If it's just one image, I can try doing it for you.
15:32 Krock enable the inventory animations and screenshot when it's in the right angle
15:32 Sokomine that would be great :-)
15:32 Sokomine rotating a texture for the wieldimage would then no longer be a problem either
15:32 shangul left #minetest
15:37 ckeltz joined #minetest
15:37 ckeltz joined #minetest
15:52 FreeFull joined #minetest
15:57 cddepppp256_ joined #minetest
16:06 riff-IRC joined #minetest
16:09 Sokomine now nametest has created a hopefully nice texture for me :-) if only i could get it now :/
16:10 jluc joined #minetest
16:43 JDCodeIt joined #minetest
16:47 JDCodeIt Hi All: any perlin noise experts here today? I'm trying to figure out where "seeddiff" comes from that one should pass to get_perlin...
16:49 Sokomine JDCodeIt: paramat or nore definitely know a lot about perlin noise. other mapgen creators perhaps as well
16:49 Krock that's just the map seed and it adds some constant value to it
16:49 Krock so you don't have to generate a random value there
16:50 Ruslan1 joined #minetest
16:50 JDCodeIt If I wanted the humidity noise - do I pass the "seed" found in map_meta.txt in the humidity section for the seeddiff?
16:52 Krock https://github.com/minetest/minetest/blob/master/doc/lua_api.txt#L3925
16:54 Krock for manual re-implementation, you take mg_biome_np_humidity and mg_biome_np_humidity_blend directly with the seed values from the map_meta file
16:54 Krock then add the noise result of each to have accurate results
16:54 Krock oh, and it's a 2d noise
16:59 JDCodeIt ...and if the world is v6 this is still valid or there we use the v6 specific parameters?
17:01 JDCodeIt mgv6_np_humidity?
17:02 Krock https://github.com/minetest/minetest/blob/master/src/mapgen/mapgen_v6.cpp#L365
17:02 Krock yes, the noise name is different
17:04 JDCodeIt And v6 doesn't have one for the blend component - is that assumed zero?
17:05 ckeltz joined #minetest
17:05 ckeltz joined #minetest
17:09 JDCodeIt cpp code there is computing a 1-D index, guess for speed - pre-calculated noise.
17:46 proller joined #minetest
17:49 Krock you're answering your own questions. task completed :D
17:49 Krock !next
17:49 MinetestBot Another satisfied customer. Next!
18:15 JDCodeIt Any advice on interpreting the temperature at a pos in v6 maps? There is a mgv6_np_biome perlin noise parameters group - but I can't see the relationship to what the map generates - high values and low values - where it is cold
18:19 Krock yes, np_biome is kinda the heat
18:19 Krock https://github.com/minetest/minetest/blob/master/src/mapgen/mapgen_v6.cpp#L449
18:41 entuland joined #minetest
18:54 Ruslan1 hello
18:56 calcul0n joined #minetest
19:00 kaeza Greetings.
19:19 Nametest joined #minetest
19:26 puzzlecube joined #minetest
19:48 puzzlecube joined #minetest
19:53 pauloue joined #minetest
19:54 JDCodeIt And the 40% of the "blend" added to the biome noise is of the perlin noise of the world seed i.e. seeddiff = 0?
19:54 Markow joined #minetest
19:59 JDCodeIt I mean 1/40th not 40%...
20:24 Miner_48er joined #minetest
20:26 sec^nd joined #minetest
20:50 MinetestBot [git] Wuzzy2 -> minetest/minetest: Update many things in Lua API docs (#8236) 1875194 https://git.io/fh5yB (2019-02-17T20:49:51Z)
21:48 nowhere_man joined #minetest
22:42 sveta joined #minetest
23:23 JDCodeIt Hi All, digging into the get_perlin function, I think the lua_api.txt is a little misleading about the word "scale" and should be instead shown as "spread" `minetest.get_perlin(seeddiff, octaves, persistence, scale)`
23:44 argyle77 joined #minetest
23:49 p_gimeno joined #minetest

| Channels | #minetest index | Today | | Google Search | Plaintext