Time  Nick         Message
00:43 paramat      plol first guess at code for player-settable heat/humidity noises https://github.com/paramat/minetest/commits/npbiome
02:39 paramat      so .. yes i am serious about allowing multiple biome systems: it's okay to have multiple biome APIs alongside hardcoded systems alongside specialist mapgens with their own internal biome systems
02:39 paramat      so .. yes i am serious about allowing multiple biome systems: it's okay to have multiple biome APIs alongside hardcoded systems alongside specialist mapgens with their own internal biome systems
02:40 paramat      oops
02:45 paramat      dev work is much healthier, fun and productive when working on multiple projects. insisting on one means working on one which leads to getting fed up with it and numb to it through over-familiarity. hmmmmm perhaps this happened to you with mgv7 and the biome API
02:45 paramat      it just causes pointless headache to worry about which system to use, feeling forced to trash one system when fed up and inspired to work on a new system
02:46 paramat      so there would potentially be a choice of base terrain mapgen, a choice of which biome API system to use, and a choice of which set of biomes
02:47 paramat      so i feel inspired to tune and work on mgv7 and the current biome API, they are impressive and popular (and i have very high standards). hmmmmm i feel you should feel free to develop new mapgens and new biome systems without feeling you have to make an existing system unofficial, or feeling you have to trash mgv7 =)
04:52 hmmmm        well
04:53 hmmmm        i have an idea
04:53 hmmmm        i'm going to hold a vote on how many people like the current mapgen v7's output.  if it's positive enough, i'll leave it in and just skip straight to v8
06:41 sol_invictus is there any working example using http://dev.minetest.net/Lua_Table_Formspec ?
07:00 hmmmm        haha..... i guess not
07:00 hmmmm        I didn't know we had lua table formspecs
08:07 realbadangel sol_invictus, https://github.com/kaeza/mt-firearms/blob/master/firearmslib/chatcommands.lua#L23
08:08 sol_invictus realbadangel: oh, nice. thx.
08:13 kahrl        realbadangel: I don't think that is related to the wiki page sol_invictus linked
08:16 sol_invictus huh, you are right. same strings, but wropped in fancy tables :)
08:17 sol_invictus I remember I've seen somewhere table formspec in issues or pull requests
08:18 Zeno`        realbadangel, there is a bug! :-o
08:20 Zeno`        karhl, feel like pouring over some massif data? :)
08:21 realbadangel Zeno`, oh no, not again ;)
08:21 Zeno`        realbadangel, it's the lighting in caves!
08:21 Zeno`        terrible
08:21 Zeno`        blocker even
08:21 hmmmm        RealBadAngel, the addMeshTaskWithEdge thing
08:22 hmmmm        you need to regen the mesh on the client to correct the lighting
08:22 hmmmm        the lighting doesn't get updated across blocks I guess
08:22 kahrl        Zeno`: heh no, I've been up for >20 hours
08:22 Zeno`        kahrl, that's a shame
08:22 Zeno`        :)
08:23 realbadangel ive tried to reproduce the issue with torches and everything seems fine for me
08:23 Zeno`        realbadangel, kahrl added an additional comment to (maybe) help you reproduce
08:23 kahrl        A feeling that we share, it's a shame
08:23 Zeno`        kahrl, why do I get the feeling you're being sarcastic?
08:24 realbadangel the way i modified the mesh updates is that max of 4 blocks are updated
08:24 kahrl        nope just making references ;)
08:24 realbadangel before it was 13
08:24 Zeno`        realbadangel, once you reproduce it, you can go back to the good commit and see that it doesn't happen before the change
08:25 realbadangel and 4 should be enough in any case imho, maybe theres some glitch
08:25 kahrl        Zeno`: does it happen at those coordinates for you too?
08:25 Zeno`        kahrl, yes
08:25 kahrl        ok good to know
08:25 Zeno`        I can do it on lots of maps though. Harder to reproduce are the dark areas that can be formed like in the forum post
08:26 Zeno`        I can manage to reproduce the dark areas only by trial and error
08:26 realbadangel now im off to work (just 3 hrs). after that i will take a deeper look on it
08:26 Zeno`        have a nice day RBA
08:26 hmmmm        good day sire
08:26 realbadangel btw, ive prepared a pull to restore finalcolorblend in shaders
08:26 kahrl        yeah, as I mentioned it doesn't happen everywhere but once you found a spot, it's easy to reproduce
08:27 realbadangel will push that later today if there are no objections
08:31 Zeno`        Krock, do you have time to discuss #1907
08:31 ShadowBot    https://github.com/minetest/minetest/issues/1907 -- Add Lua helper functions by SmallJoker
08:31 Krock        Zeno`, yup.
08:31 Zeno`        Because I still find that function confusing
08:32 Zeno`        I'm not 100% sure it's doing what the description says it should be doing
08:33 Krock        How would you describe the function?
08:33 Zeno`        I don't know, that's my problem :)
08:33 kahrl        perhaps 'limit' should be called 'tolerance'?
08:33 Krock        yeah. maybe
08:33 kahrl        that's the term used in numerical computing
08:34 Krock        just negative 'tolerances' will result in a dark corner where it could be 1 and -1
08:34 kahrl        just document that it must be >= 0 and defaults to 0
08:35 Zeno`        hmm
08:35 kahrl        and say that the function returns 0 if the number is within [-tolerance,tolerance]
08:35 Zeno`        the function name is kind of misleading in that case
08:35 kahrl        s/the number/x/
08:36 Zeno`        my comment is wrong btw
08:36 Zeno`        let me amend it
08:36 Zeno`        nvm
08:36 Zeno`        it's right
08:38 kahrl        well... I'm not convinced the limit/tolerance parameter is really needed for this function
08:38 Zeno`        I just don't like "mixing" like that
08:38 Zeno`        maybe it's a personal thing
08:38 kahrl        I mean, if some caller needs it they could as well check if math.abs(x) < tolerance then ...
08:38 Zeno`        a function called sign, imo, should do just that and nothing else
08:39 hmmmm        not sure what the point of math.sign() is
08:39 hmmmm        you can get the same results with "x - limit" if I'm understanding this correctly...
08:39 Krock        hmmmm, you can get the direction of something
08:39 hmmmm        oh... odd, x < -limit
08:39 hmmmm        yeah I'm not understanding what the purpose is
08:40 kahrl        hmmmm: it's this: http://en.wikipedia.org/wiki/Sign_function
08:40 hmmmm        if limit == 0
08:40 kahrl        well, yeah
08:40 kahrl        limit > 0 is just to catch rounding errors mostly
08:41 Krock        'limit' or 'tolerance' is just optimal
08:41 Zeno`        If limit remains then I think it should be subtracted from x at the beginning of the function and do the comparisons as if limit was not even a parameter
08:41 hmmmm        ahh this makes sense then
08:41 Zeno`        to "normalise" x
08:41 Zeno`        then the function would be easier to read (I think)
08:42 kahrl        Zeno`: are you thinking math.sign(x, limit) == math.sign(x - limit, 0)?
08:42 kahrl        because that's what I thought at first too
08:42 Krock        Zeno`, " it should be subtracted from x" -> how about negative values of x?
08:42 Zeno`        well, abs()
08:43 kahrl        and it appears hmmmm did so as well
08:44 Zeno`        I dunno. It just seems strange to me. But if it makes sense to the majority then *shrug*
08:44 kahrl        it's a nonrepresentative but conclusive study that shows the limit parameter violates the POLA ;)
08:44 Zeno`        heh
08:45 Krock        ranamed the limit to tolerance. it might be clearer now
08:46 Krock        *renamed
08:48 Zeno`        so limit is a tolerance and not the "origin"?
08:48 kahrl        yes
08:48 Zeno`        hmm ok
08:49 Zeno`        well, I was reading it completely wrong then
08:50 Krock        I wrote it a bit confusing, yes. I hope it's okay now
08:52 Zeno`        Yeah I understand now. Not convinced about the function name though :P
08:54 Zeno`        If others find it fine, then I'm fine with it as well I guess
08:54 kahrl        what would you suggest?
08:54 Zeno`        I don't know because I've never written a function like that heh
08:55 Zeno`        which is why I'm now distancing myself from it
08:55 Zeno`        well actually
08:56 Zeno`        yeah, it's right, but I guess some documentation above the function might be nice
08:56 Zeno`        So I don't ask the same question in a months time
08:57 Krock        :/
08:57 Zeno`        nah it's fine
08:58 Zeno`        It's Lua's fault and mine, not yours
08:58 Zeno`        It's more understandable now
08:59 Calinou      blaming programming languages :-)
08:59 Zeno`        hah
08:59 Zeno`        can x be rounded to tolerance instead?
08:59 Zeno`        or is that dumb?
08:59 Krock        Why round when 'if ()' is faster?
09:00 Krock        but it would be possible
09:00 * Krock      starts collecting :+1: messages and wonders if they come without begging
09:00 Zeno`        If you were given that function without the function name or any documentation could you at a glance tell what it does?
09:01 Krock        I could because I would simulate input possibilities
09:01 Krock        and then it's almost clear what it does
09:01 Zeno`        Ok, it's all cool then
09:02 Zeno`        I'm not being difficult, I really just didn't understand it this morning when I first read it :)
09:02 Zeno`        Maybe I didn't have enough caffeine
09:03 kahrl        yeah, it's basically like a fifth grade assignment: graph the function ...
09:03 Zeno`        kahrl, which is what I did in my head but when it was called limit I was assuming it was a shift to origin
09:03 Zeno`        which messed up my conclusions
09:03 Zeno`        my fault
09:04 Krock        yeah. limit was misleading
09:05 Zeno`        If kahrl is cool with it, I'm cool with it as well
09:05 kahrl        I am
09:05 Zeno`        The documentation is much clearer as well now also (I just noticed that)
09:06 Krock        *Topic switch*
09:06 * kahrl      merges
09:07 Zeno`        well... I made some nice graphs that I have no idea yet how to interpret, Krock :) Not sure I want to talk about them though
09:08 Krock        Your decision.
09:10 Krock        Okay, the fontengine checks the file endings whether it is 'ttf' or 'TTF'
09:10 Krock        That causes this strange bug
09:11 Zeno`        Krock, http://i.imgur.com/lNgtmHv.jpg     <--- server running for 15 minutes, 1 client (me)
09:12 Zeno`        http://i.imgur.com/93kyNLB.jpg  <--- server running for 30 minutes, 1 client (me). Where the cutoff is, is where I disconnected my client
09:12 Krock        Zeno`, seems like there's a memory leak in addArea(
09:12 Calinou      it should only check .ttf
09:13 Krock        kahrl, I can not find any function to make a string lowercase, what's the function of it?
09:14 kahrl        lua or c++?
09:14 Krock        kahrl, C++
09:14 Krock        Zeno`, I haven't updated my server recently, so I can not reproduce that problem yet.
09:14 kahrl        lowercase()
09:15 Krock        kahrl, lowercase(string)? Okay thanks
09:15 kahrl        Krock: yep
09:16 Zeno`        The problem with addArea() is that I have no idea (yet) where it should be released
09:16 Krock        found th source of that .ttf error
09:16 Krock        Zeno`, maybe a function in that function does not delete unused variables
09:17 kahrl        why does the font engine check file extensions at all?
09:17 kahrl        freetype supports way more than .ttf
09:17 kahrl        see http://www.freetype.org/freetype2/
09:17 Zeno`        Krock, I checked that and it seems ok (after a quick glance)
09:17 Calinou      these font settings are not recognized by the game: http://paste.ubuntu.com/9394369/
09:18 Calinou      kahrl, I agree, font_path setting should not have an extension, it should be detected automatically
09:18 Calinou      by the way; mods should probably not use .png extension either…
09:18 Calinou      (it is already possible, but nearly no mods do it)
09:18 Calinou      it implies the game only supports .png
09:18 Krock        the fail is in fontengine.cpp where it only checks for ".ttf"
09:18 Zeno`        Krock, here's a look at the client: http://i.imgur.com/3yTjjzD.jpg  (the cutoff is where I stopped moving around randomly)
09:18 Calinou      while .jpg, .pcx, .tga, .wal (!) work perfectly
09:19 Krock        Zeno`, I see. The area seems to be too big.
09:19 Zeno`        I'll work on these later I think. Kinda tired
09:19 kahrl        heh, .wal
09:19 kahrl        someone should make a quake themed subgame
09:20 Calinou      that's Quake 2 textures
09:20 sfan5        without having read the rest of the discussion. <Calinou> by the way; mods should probably not use .png extension either… why?
09:20 Krock        how about .gif (not animated)?
09:20 Calinou      sfan5, the game should not imply only .png is supported
09:20 Calinou      also, it is shorter
09:20 Calinou      it'll save thousands of bytes :p
09:21 sfan5        Calinou: but why should mods not use .png?
09:21 Krock        Gonna blame the creator of thos issue
09:21 hmmmm        https://github.com/kwolekr/minetest/commit/bb58d8f3a5b6705af7c35ec41523d542215c2c18
09:21 hmmmm        feedback?
09:21 Calinou      mods should use .png, but the extension should not be present in texture definitions
09:21 Calinou      they can use .jpg if needed too…
09:21 sfan5        oh
09:21 sfan5        thats what you mean
09:21 Calinou      legitimate format for HD packs
09:21 Calinou      like HDX already uses for 512
09:22 Krock        hmmmm, as long it's backwards compatible (caves), I'm fine with it
09:22 hmmmm        ....
09:23 Krock        :3
09:25 Calinou      you removed tabs from some lines, I thought this was bad practice and that tabs should be kept?
09:26 hmmmm        nope... trailing whitespace is bad
09:26 Calinou      only spaces at end of lines are bad
09:26 Calinou      ah
09:26 hmmmm        a blank line with some number of tabs.. that's trailing whitespace
09:26 hmmmm        whitespace trailing the end of a blank line
09:27 Calinou      camera.cpp, line 491: m_view_bobbing_speed = MYMIN(speed.getLength(), 40);
09:27 Calinou      can you change the 40 to 100?
09:28 Calinou      I'll make a PR for this
09:31 Krock        And I'm searching the reason why I can't enable or disable mods anymore
09:32 Krock        10:33:33: ERROR[main]: Error writing configuration file: "E:\Programme\minetest\bin\..\worlds\testy\world.mt"     10:33:33: ERROR[main]: Failed to write world config file
09:33 Calinou      do you have write permissions?
09:33 Calinou      start Minetest as administrator (not recommended) or run it from somewhere else
09:33 Krock        Can Windows XP disallaow on accessing a file?
09:34 Krock        Btw, I can't run it as administrator. Everything starts on this rank
09:34 Krock        Also my previous minetest version worked fine..
09:35 Krock        ^ Calinou
09:39 Calinou      I think Windows XP can restrict file access
09:39 Krock        Yeah it can, just there's no "Run as Administrator" menu entry because it runs it as Administrator by default
09:39 Krock        And as already said, the previous version worked fine
09:39 Krock        With the same paths
09:40 kahrl        where previous=?
09:40 kahrl        can you check if the problem was introduced by f0cd59034c?
09:41 Krock        okay, setting HEAD to f0cd59034c
09:43 Krock        build started
09:46 kahrl        it could also be 175b7a28e5 that caused it
09:46 kahrl        but finish checking f0cd59034c first
09:56 * Krock      removes john_minetest's #"-dev" ending
10:01 Krock        kahrl, the bug already existed before  f0cd59034c
10:02 Krock        so it might be 175b7a28 causing it
10:02 Krock        should I check that one too?
10:11 PilzAdam     sfan5, https://github.com/minetest/minetest_game/pull/367
10:11 PilzAdam     should a setting be added?
10:12 sfan5        PilzAdam: no, PR looks ok
10:13 Zeno`        +1
10:14 Krock        does cacti decay?
10:14 sfan5        no
10:15 Krock        nvm. there's a pine needles definition between. read it too fast
10:19 PilzAdam     sfan5, I guess this doesn't count as "minor fix", so I have to get BlockMen or nore agree to it?
10:20 sfan5        PilzAdam: yes
10:21 PilzAdam     VanessaE, whats the license? https://github.com/minetest/minetest_game/issues/363#issuecomment-65458085
10:22 kilbith      she's sleeping
10:22 Zeno`        merging changes to minetest_game has different rules?
10:23 Zeno`        I so I made a mistake a week or two ago (oops!)
10:24 * Zeno`      runs around in blind panic
10:30 Zeno`        sfan5, PilzAdam ^^^^
10:31 sfan5        what
10:31 Zeno`        merging changes to minetest_game has different rules?
10:31 sfan5        yes
10:31 Zeno`        ok, are they written down? I don't want to make mistakes
10:31 sfan5        they are ofc written down
10:32 Zeno`        http://dev.minetest.net/Git_Guidelines   can't see them
10:32 PilzAdam     https://github.com/minetest/minetest_game/pull/368
10:32 PilzAdam     Zeno`, http://dev.minetest.net/minetest_game_Development
10:33 Zeno`        thanks PilzAdam
10:35 PilzAdam     sfan5, what about https://github.com/minetest/minetest_game/pull/362 ?
10:35 sfan5        PilzAdam: seems ok
10:48 PilzAdam     sfan5, https://github.com/minetest/minetest_game/pull/369
10:50 sfan5        PilzAdam: looks good; that qualifies as small fix
10:50 PilzAdam     can I merge it?
10:50 sfan5        yes
10:52 PilzAdam     done
10:52 PilzAdam     thx :D
11:13 PilzAdam     sfan5, https://github.com/minetest/minetest_game/pull/370
11:21 sfan5        PilzAdam: looks ok
11:22 PilzAdam     is it a "small fix"?
11:22 sfan5        no
11:23 PilzAdam     https://github.com/minetest/minetest_game/issues/337 and https://github.com/minetest/minetest_game/issues/192 seem to be non-issues and could be closed
11:27 kilbith      sfan5: https://github.com/minetest/minetest_game/pull/341
11:27 kilbith      waiting for your vote since weeks...
11:29 PilzAdam     hmm... there don't seem to be any issues left in minetest_game that I can work on
11:29 PilzAdam     are there any issues that are not yet added to the issue tracker?
11:29 sfan5        kilbith: do we have .svgs? i like the 1st and the 5th
11:30 kilbith      nope, gimp can't create that iirc
11:30 sfan5        correct
11:30 kilbith      anyway, i couldn't do that, my laptop is dead
11:30 sfan5        GIMP does not create .svg
11:30 sfan5        Inkscape does however
11:31 kilbith      .svg ease the modification if i understand right ?
11:31 sfan5        yes
11:31 kilbith      hmm, i see
11:31 sfan5        and it allows you to make an image for many resolutions
11:32 kilbith      i guess that it doesnt make a crappy pixelation on the edges when you enlarge the images
11:32 kilbith      unlike the .png or .jpg ones
11:32 PilzAdam     sfan5, there could be a minetest.conf.example in minetest_game, with all available settings; what do you think?
11:33 sfan5        PilzAdam: sounds like a good idea
11:39 kilbith      sfan5: i think the basic header one is most versatile actually (in case if we change the mainmenu, the letters-land and gameland couldn't be always appropriate in the future)
11:40 kilbith      is the most*
11:41 kilbith      even if letters-land & gameland are the most popular (respectively 9 votes each other)
11:42 PilzAdam     sfan5, done: https://github.com/minetest/minetest_game/pull/371
11:59 jin_xi       well, i guess i can look a bit more at particles, but i will need to ask some questions...
12:00 jin_xi       so, the plan is to add a particlesystem scenenode to client when particles are enabled and use that for all digging particles and maybe also for spawning single ones
12:03 jin_xi       or maybe just add/delete single particles with their own scene node
12:03 PilzAdam     sfan5, https://github.com/minetest/minetest_game/pull/372
12:49 Zeno`        jin_xi, if you get rid of those currently leaking particles it'd be great as well :)
12:51 Zeno`        If they weren't used so much I'd be tempted to rip out the current particle spawner system entirely
12:55 jin_xi       hm.
12:58 hintss       xD
13:19 Zeno`        #1909
13:19 ShadowBot    https://github.com/minetest/minetest/issues/1909 -- Increase client performance by Zeno-
13:20 Zeno`        not the way I originally wanted to do it, but it's acceptable until after 0.4.11
13:32 realbadangel Zeno`, commented out #1909
13:32 ShadowBot    https://github.com/minetest/minetest/issues/1909 -- Increase client performance by Zeno-
13:33 realbadangel we have talked bout it, that what you are doing here is just a piece of glue
13:33 Zeno`        realbadangel, yes it's glue
13:33 realbadangel we need global settings class, initialized at startup and accesible for all
13:33 Zeno`        I did implement a whole class
13:34 Zeno`        and got into trouble. So, this is an interim solution
13:34 realbadangel so what are you waitin for? xmas?
13:34 Zeno`        Nah, I've made two pull requests with the "global class" and both got shot down
13:34 Zeno`        so I am doing this in the meantime :)
13:34 realbadangel why?
13:35 realbadangel theres too much glue around already
13:35 realbadangel we need that fixed once and for all
13:35 realbadangel and i will veto all caching attempts starting now
13:36 realbadangel no more fuckin glue
13:37 Zeno`        so, I'd like this glue approved ;)
13:38 realbadangel no way :)
13:38 PilzAdam     realbadangel, hi; when are you going to fix #1705 ?
13:38 ShadowBot    https://github.com/minetest/minetest/issues/1705 -- Regression in smooth day-night transition
13:38 realbadangel gimme the class :P
13:39 Zeno`        realbadangel, I PMd you the previous PR
13:39 Zeno`        I can't merge that at this time
13:39 realbadangel PilzAdam, i have fix rdy on my local branch, will push it today
13:39 Zeno`        THIS change I CAN merge with approval
13:39 PilzAdam     realbadangel, nice
13:39 PilzAdam     and what about #1860 and #1887
13:39 ShadowBot    https://github.com/minetest/minetest/issues/1860 -- Wield item is always fully bright
13:39 ShadowBot    https://github.com/minetest/minetest/issues/1887 -- Lighting artifacts
13:40 realbadangel sory for delays i had problems with my box for a few days
13:40 Zeno`        So I am compromising (until 0.4.11 is out the door). After 0.4.11 is another story ;)
13:40 realbadangel i let my distro to upgrade with no space left on hdd
13:40 realbadangel and had to reinstall everything
13:41 Zeno`        I am bending with the breeze :)
13:41 realbadangel 1860 have fixed, 1887 is undecided - need to look for a reason and a solution
13:41 Zeno`        1877 is easy to fix
13:42 Zeno`        but yo must test it first :P
13:42 realbadangel but not with reversing the code just
13:42 Zeno`        lol, yep
13:42 realbadangel old way worked but was EXPENSIVE
13:42 realbadangel i dont want that
13:42 Zeno`        maybe the shader update will fix it?
13:42 realbadangel i wanna know why it is working on my setup and fails on others
13:43 realbadangel no
13:43 Zeno`        realbadangel, you can reproduce it, I know you can! :D
13:43 realbadangel this is some rounding error i guess
13:43 realbadangel and i cant reproduce it atm
13:44 Zeno`        I predict that you'll be able to
13:44 * Zeno`      wiggles his fingers
13:45 realbadangel i think it may be related to that i let some updates be "not urgent" for adjacent blocks
13:45 realbadangel theres a flag for it
13:46 PilzAdam     realbadangel, #1887 may be a race condiition
13:46 realbadangel if server is jammed with another stuff,  it may just wait for update
13:46 ShadowBot    https://github.com/minetest/minetest/issues/1887 -- Lighting artifacts
13:46 realbadangel and be lost somwhere
13:46 realbadangel PilzAdam, yes, exactly
13:50 Zeno`        I'm not really sure if it's a race condition
13:50 realbadangel it may be it
13:50 Zeno`        But, yeah, it might be
13:50 realbadangel ive played with those flags
13:51 Zeno`        but it doesn't go away
13:51 Zeno`        once it's there it's there
13:51 Zeno`        it's not a 1 or 20 frame thing.. it just stays there
13:51 realbadangel such things happens all the time
13:52 realbadangel and definitely are not caused by my change
13:52 realbadangel looks like some requests are just lost
13:52 Zeno`        the light nodes are not so bad, but the dark ones (which are much harder to reproduce) are not good at all
13:52 Zeno`        i.e. those ones I linked to in the mt forum
13:53 realbadangel it was common before that you had to place something/remove to fix the lights
13:53 Zeno`        I can reproduce them but not as reliably
13:53 Zeno`        realbadangel, how come the issue goes away after the commit by bisect (which I did 4 times) then?
13:54 realbadangel my guess is the flag
13:54 Zeno`        hmm
13:54 realbadangel there are two flags, report to the server and importance
13:55 Zeno`        either way I don't think this is something that has been around for a long time (megaf was mistaken)
13:55 realbadangel it was here before
13:55 realbadangel believe me
13:55 realbadangel im playing mt for much longer than you already ;)
13:55 Zeno`        I can't reproduce before that commit though. And after the commit I can reproduce 99.9% of the time
13:56 Zeno`        I know you're more experienced ;)
13:56 realbadangel i believe i just exposed the old bug
13:56 PilzAdam     realbadangel, I haven't noticed this before, except with lava flowing, and I fixed that a while ago
13:56 Zeno`        quite possibly
13:57 realbadangel updating 13 blocks was workaround to cover the original bug i think
13:57 realbadangel (lets update all of them, maybe it will help)
13:57 Zeno`        ok, well the original bug might need to be looked at again(?)
13:58 Zeno`        realbadangel, did you look at the screenshots on the mt forum?
13:58 realbadangel yes
13:58 Zeno`        k
13:58 realbadangel i noticed something when tryin to reproduce that
13:58 realbadangel i saw something similiar but only for a very short period of time
13:59 realbadangel neighbouring block got updated 0,5-1s after the one with changed node
13:59 Zeno`        btw, I'm not "blaming" you. When I started the bisect I fully expected one of my commits to be the issue (since I changed how ambient occlusion and smooth lighting works)
13:59 realbadangel so i got a "flash" of the same effect
14:00 realbadangel but it got updated
14:00 realbadangel now, we do have some issues with threading stuff
14:00 realbadangel im on very low end CPU
14:01 realbadangel what if threading is still fucked up?
14:02 realbadangel and some request are just deleted when timed out or something?
14:02 Zeno`        that might take some time to debug
14:03 realbadangel lighting update code doesnt look at blocks, it demands request on its own
14:04 Zeno`        well, merge my PR an it'll be twice as fast and you may pick it up easier :D
14:04 realbadangel you mean that with caching settings?
14:04 Zeno`        yep
14:05 realbadangel no fuckin way :)
14:05 Zeno`        main loop is 1.90 to 2.2 times faster
14:05 Zeno`        why not?
14:05 realbadangel because there will be always another place to put some glue
14:05 Zeno`        man
14:05 realbadangel we have to cut it
14:05 PilzAdam     realbadangel, it will be removed directly after 0.4.11
14:06 Zeno`        I showed you the other solution and it's not gonna get merged. After 0.4.11 I will add a proper callback and make a proper class
14:06 realbadangel PilzAdam, adding global settings class wont break current way
14:06 Zeno`        I'm not happy but I'd rather see this than having to wait
14:06 realbadangel all usages can be replaced one by one
14:07 realbadangel and till end they can coexist
14:07 Zeno`        I do agree with you RBA
14:07 Zeno`        but... oh, I dunno :(
14:07 realbadangel just do it
14:07 PilzAdam     realbadangel, according to the 0.4.11 milestone it should have been released 7 days ago
14:08 PilzAdam     minus 7 days is not enough time to implement the global settings class properly ;-)
14:08 Zeno`        RBA, I've promised to work on the "real" solution after 0.4.11 and I intend to keep that promise
14:08 realbadangel PilzAdam, for sure i only know that im alive now and here, in next 5 minutes i can fall off the carpet and kill myself
14:09 realbadangel so theres nothing like the exact date
14:09 realbadangel only wishes ;)
14:11 PilzAdam     lol, I just found a nice bug
14:11 PilzAdam     dragging the mouse out of the menu screen causes a segfault
14:13 Zeno`        PilzAdam, that's a feature
14:13 realbadangel got also a nice one, turn on fly, fly into a corner (with a node under your feet) and continue flying
14:14 realbadangel niiiiice farting sound ;)
14:15 Zeno`        RBA, so are you ok with #1909 ? After 0.4.11 is the time to work on a bigger and better solution (IMO)
14:15 ShadowBot    https://github.com/minetest/minetest/issues/1909 -- Increase client performance by Zeno-
14:16 Zeno`        I've been playing the game with it for an hour. I'll keep playing for a while longer (just to be sure, but even looking at the src code I can't see issues)
14:17 PilzAdam     #1910
14:17 ShadowBot    https://github.com/minetest/minetest/issues/1910 -- Dragging mouse out of mainmenu window causes segfault
14:17 PilzAdam     Zeno`, #1909 seems fine
14:17 ShadowBot    https://github.com/minetest/minetest/issues/1909 -- Increase client performance by Zeno-
14:17 Zeno`        PA, I'll merge when I finish playing my game then :)
14:18 PilzAdam     maybe add a comment that states that this is to be removed soon
14:18 Zeno`        ok. I "kind of" suggested that with the TODO comments but I'll make it clearer
14:25 realbadangel Zeno`, let it be
14:25 realbadangel but this is the last time ;)
14:26 Zeno`        :D
14:58 PilzAdam     Zeno`, https://github.com/minetest/minetest/commit/0e78aa296ee8b552273fdc9d6fabec0fb01d2e4a why does this always display the name of the top texture?
15:00 sol_invictus PilzAdam: would you prefer name of a random texture?
15:00 PilzAdam     I would prefer the name of the texture I am pointing at
15:00 Zeno`        PilzAdam, good question. It's supposed to be the bottom texture. Personally I don't really see the point of displaying the texture name (.png whatever) at all but I kept it the same as rubenwardy's original patch which had many positive votes
15:02 sol_invictus but does pointed thing provides face you are pointed at?
15:02 Zeno`        Perhaps the texture name can be removed?
15:02 PilzAdam     Zeno`, it's a nice feature for texture pack developers
15:02 Zeno`        PilzAdam, hmm ok
15:04 Zeno`        sol_invictus, not really
15:04 Zeno`        because pointed at depends on the player location
15:05 Zeno`        and what texture name should be displayed when the node has multiple ones?
15:07 PilzAdam     for regular cubes it should be easy to get the texture name
15:08 PilzAdam     use pointed_thing.above to get it
15:09 Zeno`        can you make the change and test? It's 1AM here and I want to start unwinding. If your patch does what you think it should and what you think other texture pack devs would expect I will pre-emptively approve it :D
15:10 Zeno`        Mainly because I don't like the current text (for the texture filename at all)
15:33 Krock        kahrl, to the issue from earlier today: Configuration file minetest.conf also fails saving with HEAD 0a5373d4 (one before f0cd5903)
15:35 Zeno`        Krock, does 39162de work?
15:37 Krock        Zeno`, that's after 0a5373d4 and master HEAD also has this problm, so It's very unlikely it would work there
15:39 Krock        Commit 175b7a28 contains changes in "Settings::updateConfigFile()" so that might be the bad one
15:41 Krock        https://github.com/minetest/minetest/commit/175b7a28#diff-448a38aded6e71732e58a00bda2435eaR264  shouldn't this be "if (!updateConfigObject(&is, &os, ""))" ?
15:41 * Krock      does not know much about & and *
15:45 Zeno`        after?
15:45 Krock        after ^= newer than
15:46 Zeno`        0a5373d4 is 3 days ago. 39162de is 8 days ago
15:47 Zeno`        so it's before, not after :/
15:47 Krock        oops. copypasted wrong hash
15:47 Krock        I'll try that one
15:48 Krock        I'm almost sure 39162de works
15:50 PilzAdam     Zeno`, meh, this is more complicated than I thought
15:50 PilzAdam     e.g. facedir
15:51 Zeno`        PilzAdam, I know ;)
15:51 Krock        Zeno`, well, I don't need to check. My last build from Nov. 27 6a43b3a worked fine and there was no settings change
15:51 Krock        * between 39162de and 6a43b3a was no setting change
15:52 Krock        175b7a2 must be the bad commit
15:53 Zeno`        Krock, which is what I was trying to determine and the same commit I suspected ;)
15:53 * Krock      throws 175b7a2 at kahrl
15:54 PilzAdam     realbadangel, are there any helper functions to get the tile the player is pointing at?
15:54 Krock        ^ think of bandwith and how fast it changes
15:56 Zeno`        is there an open issue for this, Krock?
15:57 Krock        Zeno`, if none answers to this, I'll open one..
15:57 Krock        seems like a issue is required
15:57 Zeno`        I think you should open one
15:57 Zeno`        I haven't noticed an issue but obviously something is wrong
15:58 Zeno`        and if it's MSVC only you're likely to be the only one to find it in a hurry
15:58 Zeno`        find/experience
15:58 Zeno`        I dunno if it *is* MSVC only, but... you're the one with the data at hand ;)
16:03 Zeno`        Krock, what's the error you get?
16:03 Krock        Zeno`, in the issue.
16:03 Zeno`        oh ok
16:03 * Zeno`      looks
16:11 Krock        maybe found it
16:11 Krock        Zeno`, what do you think about adding "is.close();" at https://github.com/minetest/minetest/blob/master/src/settings.cpp#L342 ?
16:13 Zeno`        That function doesn't open is, so with a quick glance I'd say it would be bad
16:17 Zeno`        Quite possibly the calling function should close it though
16:19 Zeno`        i.e. https://github.com/minetest/minetest/blob/master/src/settings.cpp#L347
16:19 Zeno`        I'll have to look closer tomorrow. I'm too tired :) You'll probably have solved it by then
16:21 Zeno`        insert at line 356 a close (*maybe*)
16:25 Krock        Zeno`, compiling https://github.com/SmallJoker/minetest/commit/24ba930 now
16:27 Zeno`        if that doesn't work just change line 355 to just is.close() and see. But I am so tired I cannot think
16:27 Zeno`        night all
16:28 Krock        night.
16:42 sfan5        I got travis to build for windows: https://travis-ci.org/sfan5/minetest/jobs/43199022
16:45 Krock        -> @1913
16:45 Krock        * #1913
16:45 ShadowBot    https://github.com/minetest/minetest/issues/1913 -- Close settings files and save them correctly by SmallJoker
16:45 Krock        sfan5, cross compile :)
16:45 sfan5        correct
16:46 Krock        sfan5, would you agree with the fix 1913?
16:46 Krock        It could be labeled as blocker
16:47 sfan5        maybe
16:47 Zeno`        Krock, did you try testing with line 355 just as is.close(); ?
16:48 Krock        Zeno`, no I haven't. is.bad() is an error anyway.
16:50 Zeno`        I think it should be closed even if there is not an error though. (although perhaps updateConfigObject() may always return false if there is an error... but relying on that is possibly questionable)
16:51 Zeno`        Anyway, just a suggestion ;)
16:51 Zeno`        I'm not really concentrating anyway heh
16:51 Krock        Ah. It can not throw an error in this case because a file has been opened
16:52 Krock        (Looking at http://www.cplusplus.com/reference/fstream/ifstream/close/ )
16:52 Zeno`        Yeah, so the conditional is probably not required?
16:53 Krock        Yup. removing the if(!is.bad())
17:15 Zeno`        #1931 seems good
17:15 ShadowBot    Zeno`: Error: Delemiter not found in "HTTP Error 404: Not Found"
17:15 Zeno`        err
17:15 Zeno`        #1913 :)
17:15 ShadowBot    https://github.com/minetest/minetest/issues/1913 -- Close settings files and save them correctly by SmallJoker
17:17 Zeno`        apart from whitespace errors :P
17:43 Krock        twoelk, could you donate me a +1 for 1913?
17:53 PilzAdam     sfan5, https://github.com/minetest/minetest_game/pull/373
18:06 Krock        Hmm holding the right mouse button and moving it over craft slots does not work anymore :/
18:06 Amaz         Double click.
18:06 Amaz         So click once, then click again and hold.
18:07 Krock        After the left click, I right-click into the craft field and move it over the slots...
18:07 Krock        actually, with 2x right-click it works
18:07 Krock        I do not like this feature
18:45 sfan5        https://github.com/minetest/minetest/pull/1914
18:45 sfan5        comments please
18:45 Krock        Can't be a bad thing, checking the compability for windows
18:49 PilzAdam     sfan5, does it work?
18:50 hmmmm        soo
18:50 sfan5        PilzAdam: did you even look at the travis statis
18:50 sfan5        status*
18:50 hmmmm        https://github.com/kwolekr/minetest/commit/bb58d8f3a5b6705af7c35ec41523d542215c2c18 ?
18:50 hmmmm        looking for approval
18:50 sfan5        PilzAdam: https://travis-ci.org/minetest/minetest/builds/43206937
18:50 sfan5        hmmmm: looks good
18:51 hmmmm        lol did you actually look at it
18:51 sfan5        yes
18:52 sfan5        i actually looked at it when you first posted it
18:52 sfan5        but forgot to say something
18:52 hmmmm        oh thanks
18:52 PilzAdam     sfan5, seems good then
18:52 sfan5        I'll merge #1914 after hmmmm merged his commit then
18:52 ShadowBot    https://github.com/minetest/minetest/issues/1914 -- Build for win32 & win64 on Travis too by sfan5
18:53 hmmmm        k
18:55 hmmmm        hmmm ... paramat isn't around
18:56 hmmmm        remind me to tell him: I am adding a "clear all biomes" API, and selectable biome groups can be done by the mod itself or within lua
18:56 hmmmm        this is the simplest solution
18:58 Krock        Ah, please don't forget my settings save fix
18:59 hmmmm        ??
19:01 hmmmm        oh man
19:01 hmmmm        does this save failure happen for everybody?
19:02 hmmmm        oh shooooot
19:02 hmmmm        the file is never closed for reading ever since i split up the read and write
19:03 Krock        hmmmm, it did not cause fails on linux because it can deal with it
19:03 Krock        Windoze was not happy about that
19:11 casimir      What is the itemgroup "float" for? It seems to be used nowhere.
19:12 kahrl        if it is used nowhere, where did you stumble upon it?
19:12 Krock        casimir, I see you stalking base_game :P  It prevents falling nodes from replacing others
19:14 casimir      Actually I was not looking at base_game. I wanted to clean up falling, and there is a check for that group.
19:15 kahrl        oh default:snow
19:15 Krock        https://github.com/SmallJoker/base_game/commit/6d690de it was just annoying
19:15 casimir      Could it be from the weather/finite liquid system?
19:16 Krock        yes, that's possible
19:33 PilzAdam     nore, https://github.com/minetest/minetest_game/pull/362 sfan5 already agreed to it
19:33 nore         agreed too
19:33 PilzAdam     #367 too
19:33 ShadowBot    https://github.com/minetest/minetest/issues/367 -- Add selectionbox_color / crosshair_color/alpha by EXio4
19:34 PilzAdam     * https://github.com/minetest/minetest_game/pull/367
19:34 nore         both look fine
19:34 PilzAdam     can I merge them?
19:35 nore         yep
19:39 PilzAdam     done
19:39 PilzAdam     what about https://github.com/minetest/minetest_game/pull/371 ?
19:41 kaeza        Re: 1910, the regression is pretty recent, as my last build from about a week ago works fine
19:43 kaeza        oh didn't notice there's already a possible fix
19:45 PilzAdam     nore, do you have time to look at the whole list? https://github.com/minetest/minetest_game/pulls/PilzAdam
19:46 nore         do you have a screenshot for #368?
19:46 ShadowBot    https://github.com/minetest/minetest/issues/368 -- New tree generator. by RealBadAngel
19:46 nore         ok for #370
19:46 ShadowBot    https://github.com/minetest/minetest/issues/370 -- Fix attachment bug: the parent going faster than the camera by Zeg9
19:46 kahrl        kaeza: re #1910, when I wrote that possible fix, I thought the bug was introduced by Zefram_Fysh's commit in August
19:46 ShadowBot    https://github.com/minetest/minetest/issues/1910 -- Dragging mouse out of mainmenu window causes segfault
19:46 nore         371 too
19:47 kahrl        d1d6a97b
19:47 nore         372 looks good
19:48 PilzAdam     sfan5, your votes on 370, 371, 372 and 373?
19:48 PilzAdam     nore, I'll take a screenshot, gimme a second
19:48 nore         and for 373, not tested, but the code looks ok
19:48 sfan5        PilzAdam: github says 370 needs a rebase; 371 +
19:49 sfan5        372: ok;
19:49 nore         sfan5, btw, about pinetrees: what do you think?
19:49 sfan5        373 ok too
19:49 sfan5        nore: link?
19:49 PilzAdam     sfan5, yes, but the rebase is trivial in game_api.txt
19:49 nore         https://github.com/minetest/minetest_game/pull/366
19:49 sfan5        370 is ok too
19:50 PilzAdam     so I can merge 370, 371, and 372 as both of you agree to it?
19:50 nore         I find them good (but perhaps it would be better without the roots)
19:50 nore         PilzAdam, yep
19:51 PilzAdam     as for 373 I'd like to hear what Tesseract thinks, since he usually spots some errors in Lua
19:51 sfan5        nore: the pines are good
19:53 PilzAdam     nore, https://github.com/minetest/minetest_game/pull/368 screenshot added
19:53 PilzAdam     as I said there, I don't like the textures
19:53 nore         well, it looks fine to me
19:54 nore         sfan5, your thoughts?
19:54 sfan5        nore: about 368?
19:54 nore         yep
19:56 sfan5        sfan5: i agree with PilzAdam on the textures
19:56 sfan5        uh
19:56 sfan5        nore: see above
19:56 paramat      hmmmmm "clear all biomes", cool thanks
19:57 paramat      i can remove the pine roots no problem, just let me know
19:58 Tesseract    PilzAdam: I'm not sure, but this might allow you to add more than stack_max items if you add more than once: https://github.com/minetest/minetest_game/pull/373/files#diff-508f0902c96b1b655002ce43da8db3d9R72
20:01 Tesseract    PilzAdam: This and a similar loop below look ugly and create a new table on every run that's never modified.  https://github.com/minetest/minetest_game/pull/373/files#diff-508f0902c96b1b655002ce43da8db3d9R190
20:02 Tesseract    Seems unnecessary too, won't get_float return 0.0 for non-existing values?
20:03 Tesseract    If it returns nil just add "or 0" after the three getter statements.
20:04 PilzAdam     Tesseract, no, it doesn't allow more than stack_max (just tested; also that code was just copied from the old version)
20:04 Tesseract    The inventory should be initialized in on_construct, but I guess that's to allow placing with vmanip.
20:05 PilzAdam     these ugly loops where just copied, too
20:05 PilzAdam     I remove them
20:05 PilzAdam     I intentionally moved the the inventory definition into the ABM
20:08 Tesseract    PilzAdam: Should this run if there's more fuel?: https://github.com/minetest/minetest_game/pull/373/files#diff-508f0902c96b1b655002ce43da8db3d9R301
20:09 Tesseract    Also see commit comments.  There are also missing spaces after commas.  Otherwise it looks good.
20:09 PilzAdam     if there is more fuel then it doesn't jump into the else
20:15 PilzAdam     updated
20:15 PilzAdam     where are missing space before commas? or could someone tell me a regexp to find them?
20:16 Tesseract    PilzAdam: In vim ",[^ ]" should work.
20:17 Tesseract    Lua has %S and Python has \S for the "[^ ]" I think.
20:17 PilzAdam     thanks
20:18 PilzAdam     updated
20:19 PilzAdam     sfan5, https://github.com/minetest/minetest_game/pull/373 ?
20:19 sfan5        PilzAdam: looks good
20:20 PilzAdam     should I wait until someone actually tests it or can I merge it?
20:28 kahrl        wtf, IGUIScrollBar
20:28 kahrl        Y U NO eat double click events?
20:29 sfan5        wait
20:29 kahrl        (this is what causes #1806)
20:29 ShadowBot    https://github.com/minetest/minetest/issues/1806 -- Double-clicking the scroll bar of the server list joins a server
20:29 paramat      nore sfan5 i will remove the pine roots, because i will probably be adding optional snow biomes to mgv6 (on your request) and these trees will be used for that so should be rootless like the other mgv6 trees, for consistency
20:30 kahrl        before GUITable was added, double clicks were detected in a haphazard way, which apparently was more reliable than the irrlicht event system
20:44 kahrl        pushing this trivial-ish fix for 1567/1806 in an hour: https://gist.github.com/kahrl/eb3e016716c5b48fa380
21:11 hmmmm        kahrl:  looks good to me
21:16 PilzAdam     sfan5, https://github.com/minetest/minetest_game/pull/373#issuecomment-65913976 enough testers?
21:21 realbadangel #1916, waiting for travis
21:21 ShadowBot    https://github.com/minetest/minetest/issues/1916 -- Restore finalColorBlend implementation in shaders. by RealBadAngel
21:21 jin_xi       hmmmm: thanks for adding deco notifications, working nicely
21:28 PilzAdam     RealBadAngel, testing...
21:30 RealBadAngel PilzAdam, im testing the build too
21:30 RealBadAngel should be ok, but more testing wont hurt
21:31 PilzAdam     why does git only fetch with 30 KiB/s?
21:31 PilzAdam     and why is your repo so big?
21:31 PilzAdam     woah, 56 branches
21:32 PilzAdam     you do realize that you can delete old branches via the github web interface, right? ;-)
21:32 kahrl        I believe menu_music is the big branch
21:32 RealBadAngel i wont delete them rather, thx to them i could easily revert such things like finalColorBlend for example
21:33 RealBadAngel and keep there my unfinished stuff
21:33 PilzAdam     what a beautiful sunset <3
21:34 kaeza        I need some help here. An user of my `imageloader` mod is getting "ERROR[ServerThread]: ERROR: An unhandled exception occurred: basic_string::_S_construct null not valid". The mod used to work a while ago, but as far as I was able to trace it, it dies in `place_schematic`.
21:34 kaeza        has there been any change to Lua schematic handling lately?
21:34 PilzAdam     RealBadAngel, merge it!
21:35 hmmmm        kaeza, are you loading the schematic from a file or from a table directly from lua?
21:36 kaeza        hmmmm, directly from Lua
21:36 kaeza        The data seems to conform to the spec in lua_api.txt: `{ size = { y = 1, x = 16, z = 16 }, data = { { name = "air" }, [...], { name = "wool:yellow" }, [...] } }`
21:36 RealBadAngel PilzAdam, i just need to check why i removed one if from there before
21:36 RealBadAngel if(day != night)
21:37 RealBadAngel m_daynight_diffs[i][j] = std::make_pair(day, night);
21:37 RealBadAngel i removed that if at some point because of forgotten reason, i need to figure out why i did that
21:38 kaeza        the relevant call is here, for reference: https://github.com/kaeza/minetest-imageloader/blob/master/imageloader.lua#L118
21:38 kaeza        (ignore the fact that I used 0 and not "0"; changed it locally but still crashes)
21:40 hmmmm        where that error happens doesn't seem to be very obvious...
21:56 kaeza        it looks like the user is using mt-0.4.10 (stable, MSVC), so this is old bug that went unnoticed
21:59 kaeza        in my version (0.4.10-dev, pulled just today, MinGW32), it seems to crash as soon as `std::string` is called with `null`. I don't get any traceback to help pinpoint the location :/
22:01 kaeza        wait... I do
22:03 kaeza        hmmmm, does this make any sense? https://gist.github.com/kaeza/2703f9d1a4471b187594
22:29 paramat      i tried lua table schematics a week ago and couldn't get them to work
22:52 hmmmm        ;/
22:52 hmmmm        you should've made a bug report or tell me here
22:54 hmmmm        kaeza:  it hardly makes sense at all... std::string is called with null?  there is no std::string involved with updateContentIds()
22:55 hmmmm        what version of minetest is that?  i mean which commit is that
22:55 hmmmm        want to see what l_mapgen.cpp:657 is
23:09 kaeza        hmmmm, `schem.placeStructure(map, p, 0, (Rotation)rot, force_placement, ndef);`
23:10 kaeza        latest commit, pulled about two hours ago
23:11 kaeza        "Backtrace stopped: previous frame inner to this frame (corrupt stack?)"
23:11 hmmmm        that has to be a corrupt stack
23:11 kaeza        WTH
23:11 hmmmm        this would make no sense otherwise
23:12 hmmmm        paramat:  https://github.com/minetest/minetest/commit/60feb4ad25003b4ed5f93f7fda7f4b8177a0ba51
23:12 kaeza        if you have some time to test for yourself on a more sane OS, `git clone https://github.com/kaeza/minetest-imageloader imageloader`, then do `/loadimage mp.bmp` in game
23:13 paramat      i cant remember if i told you at the time, but yes i failed to make a bug report
23:13 * paramat    looks
23:17 hmmmm        hmm
23:17 hmmmm        kaeza, not related, but i'm getting lots of "assignment to undeclared global" warnings
23:17 hmmmm        well, errors
23:21 paramat      clear biomes api: a little beyond my coding level to judge, however thanks! if it helps here is my branch where i define a set of biomes https://github.com/paramat/minetest_game/commits/mgv5biomes
23:23 kaeza        already fixed locally, but thanks for reporting anyway :P
23:24 hmmmm        paramat, the idea is to implement the "group of biomes" thing within lua, using register_biome and clear_registered_biomes
23:24 hmmmm        also this way if more than one mod adds a biome, they can erase all previous biomes
23:24 paramat      each grass length and flower type has a carefully chosen perlin noise for beautiful results: there ia an ever-changing density and mix of flower colours
23:24 paramat      okay cool
23:25 twoelk       paramat:  That tree root growing you had suggested. Would the root replace the dirt node the sapling grew from?
23:26 paramat      hmmmmm very useful features
23:27 paramat      twoelk, the pine will have no roots, but my idea was to extend the trunk into the ground: 2 for appletree, 3 for pines and jungletree, so yes replacing the dirt
23:28 paramat      however the new tree schems for mgv5/v7 biomes will be different and i would like to have roots on those
23:29 twoelk       ah, so the automated tree factories for example would also have to add a dirt node to grow a new tree
23:30 paramat      yes so mgv6 trees will stay unchanged to avoid breaking worlds and mods
23:30 twoelk       dirt was not used up by plants until now, wonder how this changes some game-plays such as on the Just-test or moontest servers
23:31 paramat      and since the pine will be in mgv6 snow biomes that should have no roots also, i will edit that tonight
23:32 paramat      roots will only be added to new tree schem designs so wont break anything (i hope)
23:33 twoelk       if dirt can get destroyed by tree growth in some mapgens we might need to add something to craft dirt to default as well to balance it out
23:36 paramat      ah i see, holes will need to be filled
23:38 paramat      nand dirt might run out on skyblock maps
23:38 paramat      *and lol
23:39 twoelk       many servers advise the players to replant trees. this might get a little difficult. Maybe the leaves could be utilized in some way
23:39 hmmmm        kaeza:  https://github.com/minetest/minetest/commit/390c0cd69185b3be7f6f49d7f8fe53f901edf5aa
23:40 twoelk       Dirt is actually rather rare on some servers or games
23:44 * paramat    keeps singing this https://www.youtube.com/watch?v=F_6IjeprfEs