Time  Nick        Message
01:11 paramat     nore is it possible you could approve #5744 ? luckily it's in mapgen code i wrote myself so i know what i'm doing here
01:12 ShadowBot   https://github.com/minetest/minetest/issues/5744 -- Caverns: Remove unnecessary liquid excavation by paramat
06:13 nerzhul     hello :) is #5756 okay for a merge ?
06:13 ShadowBot   https://github.com/minetest/minetest/issues/5756 -- CSM: Document forgotten functions by SmallJoker
06:33 Zeno``      does merging using github fix whitespace errors?
06:34 Zeno``      and why is my nick Zeno``? O.o
06:47 Zeno`       https://github.com/minetest/minetest/blob/master/src/threading/event.cpp#L30 is not doing what is should be doing. If USE_CP11_MUTEX *is* defined then notified is never initalised to false
06:51 Zeno`       the easy and obvious (but kinda ugly) fix is: http://dpaste.com/0PJ0Q2C  but perhaps the whole thing needs to be restructured to be a bit cleaner. If I have time tonight I'll do it otherwise anyone else is welcome :)
06:55 Zeno`       This memory is not freed on shutdown: https://github.com/minetest/minetest/blob/master/src/mesh_generator_thread.cpp#L211 (CC celeron55_ celeron55)
06:56 nerzhul     Zeno`, exact :à go ahead :D
06:56 Zeno`       I'll try and look more to see if it's a "real" leak or just not released at shutdown. If I get time tonight (I should) I'll look at it more but c55 already understands the code and I've never even seen it before :D
06:58 nerzhul     if not release at shutdown => when leaving local server then memleak
06:58 Zeno`       yeah, but less critical than one constantly increasing RAM usage (still should be fixed though)
06:58 Zeno`       I'll make a PR for the unitialised var. issue
07:04 Zeno`       nerzhul, #5764
07:04 ShadowBot   https://github.com/minetest/minetest/issues/5764 -- Fix uninitalised variable in event.cpp by Zeno-
07:06 nerzhul     +1
07:06 * VanessaE  still waits for another +1 on game#1732 and game#1731 needs merged too
07:07 ShadowBot   https://github.com/minetest/minetest_game/issues/1732 -- redo new open chests to use all single textures by VanessaE
07:07 ShadowBot   https://github.com/minetest/minetest_game/issues/1731 -- Chests: Revert inventory name by tenplus1
07:38 paramat     paramat is doing
07:39 paramat     i'll do game#1731 in a few mins
07:39 ShadowBot   https://github.com/minetest/minetest_game/issues/1731 -- Chests: Revert inventory name by tenplus1
07:43 paramat     ideally that should have been merged by Krock when he merged 1733
07:46 VanessaE    what about 1732? :(
07:47 paramat     i'm neutral
07:48 paramat     oh dear it has 1.5, you'll have to bug the other devs :]
07:50 VanessaE    :(
07:51 VanessaE    I'm running out of people to pester :P
07:57 paramat     hmm i might add a +0.5
08:10 VanessaE    please do
08:38 paramat     VanessaE what is your opinion, as a server admin, on game#1693 ? i would very much like your input
08:38 ShadowBot   https://github.com/minetest/minetest_game/issues/1693 -- Disable fire by default on servers by Ezhh
08:38 VanessaE    +100000
08:38 VanessaE    'nuff said? :)
08:39 paramat     thanks
08:39 paramat     please add comment
08:39 nerzhul     the problem with fire is that fucking lua which doesn't permit to spread fire because lua developpers don't understand what is a server loop
08:40 VanessaE    paramat: I added my :+1: to the first post.
08:40 nerzhul     fire & tnt recursive calls must be spread over multiple server loops
08:40 nerzhul     (i did that on my MT version 2 years ago, for TNT, it permits to nuke 1000 tnt blocks in ~1min without any lag)
08:40 VanessaE    nerzhul: not to mention you can burn down entire forests and cities with it -- it doesn't have any kind of limits on spread.
08:43 nerzhul     it's logical in fact, and the limit is the chance to burn near nodes :)
08:50 paramat     thanks
08:54 VanessaE    nerzhul: when I say a "limit" I mean, you can burn down an entire 10'000 square meter area easily.
08:54 VanessaE    in-game that would be like burning down the entire Smoky Mountain National Forest
08:54 nerzhul     if there are no firefighter :)
08:54 VanessaE    heh
08:55 nerzhul     maybe add a node which spread water when there is a forest fire :p
08:55 VanessaE    well even with firefighters, the most recent wildfires burned for a good couple of weeks and destroyed millions of dollars worth of property
08:55 VanessaE    (in the SMNF)
09:04 celeron55   where the eff is my voice on #-hub
09:04 celeron55   lol
09:04 VanessaE    right there. :)
09:05 celeron55   looks like i was missing +V
09:38 paramat     nore sfan5 perhaps you could add your opinions to game#1693 ? ShadowNinja are you +1 ?
09:38 ShadowBot   https://github.com/minetest/minetest_game/issues/1693 -- Disable fire by default on servers by Ezhh
09:40 Zeno`       does anyone know if using the squash and merge button on github fixes whitespace issues?
09:40 Zeno`       (i.e. trailing whitespace)
09:41 Zeno`       I guess I'll do it the traditional way just in case.
09:42 Zeno`       will merge #5758 soon
09:42 ShadowBot   https://github.com/minetest/minetest/issues/5758 -- Add option to use neither node highlighting nor outlining by Ezhh
09:52 Zeno`       whitespace errors fixed, will merge in 5 minutes
09:55 Zeno`       paramat I don't think the issue of reducing memory for a local variable is an issue at all
09:56 Zeno`       the lifetime of that object (in the constructor) is so short that if it causes RAM problems then something is seriously wrong
09:56 paramat     ok no problem :]
09:57 Zeno`       pushed :)
09:58 Zeno`       yeah I dunno nerzhul...
09:58 Zeno`       you can fix it if you want :)
10:02 nerzhul     rrRr
10:04 Zeno`       rrrrrr :)
10:05 Zeno`       I'll also merge #5764 in a few minutes since it's a trivial fix
10:05 ShadowBot   https://github.com/minetest/minetest/issues/5764 -- Fix uninitalised variable in event.cpp by Zeno-
10:06 nerzhul     Zeno`, hopefully in two months this ifndef /ifdef C++11 compat code will be removed :p
10:08 Zeno`       yes :)
10:09 Zeno`       all the #ifndef / ifdef is ugly for a start heh
10:10 Zeno`       merging now
10:10 Zeno`       done
10:10 Zeno`       I'm exhausted after writing all the complex code in that PR
10:11 Zeno`       celeron55, did you want me to look at the memory leak I mentioned?
10:12 nerzhul     you should do it Zeno`
10:12 nerzhul     it's very important
10:12 Zeno`       <Zeno`> I'll try and look more to see if it's a "real" leak or just not released at shutdown. If I get time tonight (I should) I'll look at it more but c55 already understands the code and I've never even seen it before :D
10:12 nerzhul     just remove the map elements in the destructor
10:12 Zeno`       yeah if c55 doesn't have time I'll do it after I eat. Just seemed easier if he did it because I didn't even know that file existed :) May as well look at it though
10:13 Zeno`       I bb in an hour or so. If someone fixes it before then, just do it and I'll see it in the logs
10:16 Fixer_      backtrace is now available: https://github.com/minetest/minetest/issues/5661
10:18 nerzhul     okay
10:21 nerzhul     wow the line is so unsecure in terms of code
10:21 nerzhul     if (isspace(line.text.getString()[in_pos + frag_length]))
10:46 paramat     rubenwardy it seems you are +1 for game 1693? if so i will merge it with 1731 very soon
10:47 rubenwardy  game#1693
10:47 ShadowBot   https://github.com/minetest/minetest_game/issues/1693 -- Disable fire by default on servers by Ezhh
11:00 paramat     ok assuming +1, will merge in 15mins
11:02 rubenwardy  I'm not +1, I just don't care
11:04 paramat     ok, thanks, wil not then
11:20 paramat     please could anyone approve #5744 ? it's a change to my own code so i know what i'm doing
11:20 ShadowBot   https://github.com/minetest/minetest/issues/5744 -- Caverns: Remove unnecessary liquid excavation by paramat
11:24 rubenwardy  Who has access to weblate?
11:25 nerzhul     m
11:25 nerzhul     me
11:26 nerzhul     is there any problem ?
11:27 rubenwardy  no, I was just wondering if it would be acceptable to put https://github.com/rubenwardy/mtmods4android as a project in there, even though it's third party
11:27 nerzhul     as you said it's third party, and if you put it in weblate why not, but you should manage it alone as it's your project :)
11:28 rubenwardy  ofc
11:29 rubenwardy  oh
11:29 rubenwardy  wth
11:30 rubenwardy  I literally just received a response from weblate about my weblate request
11:33 rubenwardy  it's now here: https://hosted.weblate.org/projects/minetest/mtmods4android/
11:33 rubenwardy  did you do anything, nerzhul, or is it by chance that they responded just now?
11:35 nerzhul     i'm not weblate
11:36 paramat     i just had a weird crash
11:37 paramat     started a world in head - 2, recompiled to head - 17, on world restart game freezes, window cannot be closed, "ServerError: AsyncErr: World data version mismatch in MapBlock (0,-34,0). World probably saved by a newer version of Minetest"
11:37 paramat     happened twice
11:38 paramat     will open issue
11:38 paramat     i have seen someone else report this too
11:38 nerzhul     kilbith, ^
11:39 Zeno`       err, bisect it?
11:39 paramat     yeah might do that
11:39 Zeno`       it'd help
11:39 kilbith     yep, got the same issue
11:40 nerzhul     it's due to sfan5 changes i think
11:40 sfan5       yes
11:41 sfan5       you can't cross a version-increase boundary backwards
11:42 sfan5       but if you're testing a client bug you can leave the server at the new version and it will work
11:44 paramat     thanks i did not relise we had bumped map format
11:45 sfan5       (this is also why i think making a forum post about this is stupid: because it happens relatively often)
11:45 paramat     which commit was it?
11:47 paramat     nevermind
11:48 paramat     seems to be 'nodemeta'
11:48 nerzhul     it was meta private PR
11:49 paramat     yeah, sorry for the silly question
11:54 paramat     happy to confirm it was just map format bump
12:09 paramat     6 days to freeze
12:10 rubenwardy  I won't be able to do any development until this Friday, as exams
12:11 rubenwardy  so that leaves 1/2 days for #5176
12:11 ShadowBot   https://github.com/minetest/minetest/issues/5176 -- Use JSON serialize to store ItemStack Metadata by rubenwardy
12:16 paramat     game#1573 ?
12:16 ShadowBot   https://github.com/minetest/minetest_game/issues/1573 -- Remove set_inventory_formspec in default, remove delay in sfinv by rubenwardy
12:16 paramat     is that needed for release?
12:19 rubenwardy  no, it's not a blocker
12:22 kilbith     the sky code looks kinda broken: https://lut.im/jerf6xHLe8/Ffi8We1keBWZ8wqa.gif
12:22 kilbith     I'm slightly moving the camera to the right and it turns grey
12:22 Raven262    I have seen this many times, when you look away from the sky it gets gray
12:22 Raven262    Idk why though.
12:24 rubenwardy  that's been like that for a while
12:24 Raven262    like forever xD
12:24 Raven262    I thought that was a feature?
12:24 rubenwardy  XD
12:24 rubenwardy  https://hosted.weblate.org/widgets/minetest/mtmods4android/svg-badge.svg
12:24 rubenwardy  oops
12:24 rubenwardy  not my area though
12:30 VanessaE    paramat or rubenwardy, can someone move my name to "Active contributors" please?
12:31 VanessaE    (assuming 1732 goes in)
12:31 kilbith     not enough commits
12:31 kilbith     or give $50
12:32 kilbith     also, MTG commits are not counted
12:39 VanessaE    well pfft.
12:44 paramat     yeah, if it goes in
12:44 paramat     will merge game#1731 in a moment
12:44 ShadowBot   https://github.com/minetest/minetest_game/issues/1731 -- Chests: Revert inventory name by tenplus1
12:45 kilbith     credits are for the engine, it has been said several times in the past
12:54 paramat     merging
12:59 paramat     merged
13:03 nerzhul     active  contributors is for core not mgt
13:03 nerzhul     mtg*
13:04 paramat     ok i did wonder
13:25 celeron55   Zeno`: feel free to fix it; it's trivial
13:28 Zeno`       celeron55, yeah I will if nobody does in the meantime. But as I said it's not something that's critical so I've been doing other stuff instead
13:41 nerzhul     it's critical as if a user change from one world to another the memory is lost
15:14 red-005     could someone review #5757?
15:14 ShadowBot   https://github.com/minetest/minetest/issues/5757 -- Improve password change menu by red-001
16:31 * VanessaE  peeks in and pokes whoever about game#1732 :P
16:31 ShadowBot   https://github.com/minetest/minetest_game/issues/1732 -- redo new open chests to use all single textures by VanessaE
16:32 DS-minetest > * kilbith uses math.ceil on the approval number
16:36 paramat     will merge game#1693 in a moment
16:36 ShadowBot   https://github.com/minetest/minetest_game/issues/1693 -- Disable fire by default on servers by Ezhh
16:44 paramat     merged
22:15 paramat     Fixer this seems to be working now #5760 could you test?
22:15 ShadowBot   https://github.com/minetest/minetest/issues/5760 -- Particles: Do not send to distant players by paramat
22:37 paramat     how should the distance be set? i think 'active block range' is too short at 48 nodes. we could have a new setting, or maybe use 'max block send distance' (160 nodes) as that determines how much world is visible to a client
22:38 VanessaE    can't you just limit by players' view range?
22:46 KaadmY      Does the server ever get the player's fog distance?
22:47 VanessaE    I should think it gets something vaguely related, as the client can request blocks directly now, can't it?
22:55 paramat     i feel it is best to keep it lightweight and simple, there is still a huge gain from simply not seeing particles beyond, for example, 160 nodes. the player's view range will still stop the particles being visibly seen beyond that view range
22:59 KaadmY      Most particles will probably be small enough so after ~160 nodes they most likely won't be visible anyway
23:03 paramat     player view range is dynamic so could not be cached. so every individual particle spawned (can be up to 100 per second) would need to fetch the currently active view range of every player on a server
23:04 KaadmY      Wouldn't a server-side setting make the most sense?
23:04 VanessaE    paramat: are you sure about the view range effectively hiding them?
23:04 KaadmY      After all, it will still reduce network traffic/bandwidth from what it currently is
23:04 VanessaE    seems fixer had a problem with that as well as them being sent in the first place.
23:04 paramat     but i could be more precise with particle spawners, and use the closest distance out of 'minpos' and 'maxops' instead of the midpoint
23:04 Fixer       paramat: hi, will look at it
23:04 Fixer       i was afk
23:04 VanessaE    speak of the devil :P
23:05 paramat     KaadmY yes that's what i'm suggesting, i'm arguing against getting player view ranges
23:05 KaadmY      Okay :)
23:05 Fixer       paramat: another question is... can you see particles that far? say someone explodes TNT 160 blocks away...
23:05 KaadmY      Since using the individual player view range doesn't give any real advantage over a single variable setting
23:06 paramat     TNT smoke is a large particle, it would be visible at 160 nodes yes
23:06 paramat     that's why 48 nodes is too small
23:07 KaadmY      Could the distance be multiplied by the particle size relative to how large it would be onscreen?
23:07 KaadmY      Roughly, at least
23:07 Fixer       paramat: so there is choice 48 vs 160, I don't think 160 nodes will be bad enough, not sure, why other devs can't opinion on this
23:07 Fixer       160 will probably be fine
23:08 paramat     the hardcoded value is temporary, for testing
23:08 Fixer       maybe overkill for some, who knows
23:08 Fixer       paramat: hmmm, but wait
23:08 paramat     we could have a new setting, or use a chosen existing setting
23:08 KaadmY      Probably a new setting?
23:08 Fixer       paramat: when you explode tnt and run away 160 nodes away from it - blocks will not update from that distance... right?
23:08 KaadmY      Since it's not tied to anything else in specific?
23:09 Fixer       or maybe i'm wrong
23:09 paramat     depends whether it is a spawner or individual particles
23:09 paramat     the distance is checked using the spawner's position, or an individual particle's position
23:10 paramat     (or rather, the midpoint between minpos and maxpos)
23:10 KaadmY      Why not use the nearest of minpos/maxpos?
23:10 KaadmY      It wouldn't be accurate at all, but it'd be better
23:10 paramat     yes i just suggested that above
23:10 Fixer       active_object_send_range_blocks = 3 / active_block_range = 3 / max_block_send_distance = 10
23:10 paramat     i might do that
23:11 paramat     10 mapblocks is better than 3
23:12 paramat     i suggested 'max block send distance' as that determines how far a client sees
23:12 paramat     but a new setting might be good
23:19 paramat     (btw am tired, been up long time)
23:20 VanessaE    /give paramat homedecor:coffee_maker
23:20 VanessaE    :)
23:25 paramat     i use dark choc
23:28 paramat     don't drink tea / coffee :O
23:48 VanessaE    heh