Time  Nick        Message
01:02 Fixer       o.O
04:01 GreenDimond is it just me or is minetest.hash_node_position() broken
04:03 GreenDimond on the x axis hashes only change every 10 nodes
04:26 paramat     an issue is a good idea, otherwise this will be forgotten
04:29 paramat     core devs: to sort out #8300 for 5.0.1 as best we can, #8357 can be merged to backport-5 branch together with, and after, the 2 commits linked in the PR. 8357 should also be merged to master branch
04:29 ShadowBot   https://github.com/minetest/minetest/issues/8300 -- Segfault in noise.cpp
04:29 ShadowBot   https://github.com/minetest/minetest/issues/8357 -- num_emerge_threads: Warn of crashes when > 1 by paramat
04:54 paramat     i think a warning is all we can do and the best we can do for 5.0.1. fixing these bugs will take a long time so should be left to 5.1.0. the docs already recommend 1 thread and suggest more might be problematic, so not a big issue
04:57 GreenDimond Disregard earlier hash comments; problem was in a database, not Minetest.
05:19 paramat     good news thanks
06:23 nerzhul     paramat, i think else we will need to wait too long whereas we have a workaround by warning users
06:23 nerzhul     just merge that pr and release 5.0.1
11:07 p_gimeno    @sofar @paramat https://github.com/minetest/minetest/commit/61e5fbab721898a431b15a5a7e24efb58cd80eb4
14:08 rubenwardy  Don't forget to review the other pr
14:08 rubenwardy  #8362
14:08 ShadowBot   rubenwardy: Error: Delimiter not found in "HTTP Error 404: Not Found"
14:08 rubenwardy  #8352
14:08 ShadowBot   https://github.com/minetest/minetest/issues/8352 -- Abort in deactivateFarObjects during unroll caused by Lua Error in a Lua Error
14:08 rubenwardy  Heh
14:08 rubenwardy  The one made by me
14:41 nerzhul     nice catch, what is the real lua error
15:17 sfan5       #8347 would be nice to have too
15:17 ShadowBot   sfan5: Error: Delimiter not found in "Connection timed out."
15:17 nerzhul     #8347
15:17 ShadowBot   https://github.com/minetest/minetest/issues/8347 -- Fix set_hp(0, reason) Lua error by pauloue
15:20 nerzhul     what is the -1 value set to the reference ?
15:21 sfan5       nothing special, it just means a new table will be created next time the reason is pushed
15:21 sfan5       see the other code in the same method)
15:36 rubenwardy  It's a bad fix
15:37 rubenwardy  #8354
15:37 ShadowBot   https://github.com/minetest/minetest/issues/8354 -- Fix cast from const by accessing string data directly by rubenwardy
15:37 rubenwardy  Is the pr I meant to link to
16:16 VanessaE    rotation of node box textures is broken...
16:18 VanessaE    http://i.imgur.com/DyzJJKx.png
16:19 VanessaE    those vertical-placed glowlights didn't look like this when the structure was first built some time ago (under 0.4.x, several months ago)
16:19 VanessaE    (the one on the ground is correct)
16:41 nerzhul     rubenwardy: try .data() instead of .c_str() or using ref + 0 offset
16:41 nerzhul     it should do the trick
16:41 rubenwardy  Is that an actual irr::core::string method?
16:42 nerzhul     oh it's a irrlicht string ? i don't know if it's implemented
16:43 nerzhul     but if it is you should use it instead
16:48 red-045     looking at the online docs they don't allow that
16:49 nerzhul     then it's the only way
16:50 red-045     couldn't you do ret[0] = unicode::BOM_ENCODE_UTF16_BE;?
16:51 red-045     s/BOM_ENCODE_UTF16_BE/BOM_ENCODE_UTF16_BE[0]
16:55 VanessaE    so, no comment on my issue?>
16:56 VanessaE    (as in do I need to file an issue and hope it's fixed some time in the next 6 months, or is it already known?)
17:07 p_gimeno    VanessaE: do newly placed boxes look right?
17:07 p_gimeno    er, glowlights*
17:08 VanessaE    no.
17:08 VanessaE    gonna file an issue.
17:14 VanessaE    #8358
17:14 ShadowBot   https://github.com/minetest/minetest/issues/8358 -- Nodebox facedir no longer rotates textures to follow it.
17:16 p_gimeno    what glowlight is that? the homedecor glowlight works fine for me
17:17 VanessaE    that's a "quarter" glowlight.
17:18 VanessaE    it also affects half- and small cubes.
17:19 VanessaE    http://i.imgur.com/efcxcuo.png
17:20 VanessaE    how old is "older?
17:20 VanessaE    did you have to dye those to make them yellow/
17:20 VanessaE    ?
17:20 p_gimeno    no idea0.4.15
17:20 VanessaE    oh that's old as shit :)
17:21 VanessaE    that's pre-colorwallmounted
17:23 p_gimeno    where's glowlight now?
17:24 VanessaE    https://gitlab.com/VanessaE/homedecor_modpack/blob/master/homedecor/lighting.lua#L52
17:26 VanessaE    (that's why I asked if you had to dye them -- modern lights are white by default)
17:26 p_gimeno    ah it's complaining about basic_materials
17:26 VanessaE    https://gitlab.com/VanessaE/basic_materials
17:30 p_gimeno    yep, reproduced
17:30 VanessaE    ok
17:30 VanessaE    maybe edit your post?
17:33 p_gimeno    when something is a radical change on what was already written, I prefer to write a new post, that way if someone is following only by email, they don't get the wrong idea
17:35 VanessaE    fair enough
18:22 rubenwardy  red-045: you can't use [] directly on the string because then it would update whole characters at a time
18:22 rubenwardy  you want to set bytes, not characters
18:22 rubenwardy  which is why it casts it from wchar to char
18:22 red-045     ah I missed thast
18:23 red-045     that*
18:35 rubenwardy  #8359
18:35 ShadowBot   https://github.com/minetest/minetest/issues/8359 -- HPChange Reason: Fix push after free, and type being overwritten by rubenwardy
18:37 nerzhul     rubenwardy: for the -1 is there any macro instead ?
18:37 rubenwardy  no
18:37 rubenwardy  -1 means the top of the stack
18:37 rubenwardy  -2 is the second from the top
18:37 nerzhul     no no
18:37 rubenwardy  1 is the lowest item on the stack, usually the first parameter
18:37 nerzhul     if (reason.lua_reference != -1)
18:38 nerzhul     this -1 :)
18:38 nerzhul     i worked on CSM i remember the lua part hehe, i talked about this specific thing
18:38 sfan5       rubenwardy: shouldn't that != -1 be >= 0 like in the other function?
18:39 rubenwardy  yeah, just noticed that
18:39 sfan5       also should that newly added if() set the reference to -1 after doing unref or am I misinterpreting something?
18:39 rubenwardy  after the return 0 it goes out of scope anyway
18:40 sfan5       >bool hasLuaReference() const
18:40 sfan5       inline
18:40 rubenwardy  isn't `inline` basically ignored now a days?
18:40 nerzhul     no it's not ignored
18:41 rubenwardy  updated :P
18:41 rubenwardy  https://github.com/minetest/minetest/pull/8359/commits/2b0b31c9a80f4b5f37adbb7e6c3edf5aa77fdd5f
18:41 nerzhul     oh the ref is our own int, i see :)
18:42 nerzhul     i don't like the lifetime of the reason
18:51 rubenwardy  should I also add a docs fix to that PR?
18:51 rubenwardy  heh
18:52 rubenwardy  literally just adding "reason must be a table or nil.  See PlayerHPChangeReason in register_on_player_hpchange"  to set_hp
18:53 rubenwardy  the function should probably also raise a Lua Error if given a non-table
19:04 nerzhul     yes if you can check types it can be nice
19:42 GreenDimond In theory how short can minetest.after() go?
19:43 GreenDimond when I checked a few months ago it seemed like ~0.1 seconds, but benchmarking now it looks like it can be ~0.2 milliseconds
19:43 GreenDimond unless I am benchmarking wrong
19:47 GreenDimond got 0.003 at one test
19:47 GreenDimond it varies
19:49 sofar       GreenDimond: the answer is 1 server tick
19:49 sofar       in other words, the *next* server globalstep lua execution
19:49 GreenDimond mhm
19:49 sofar       since minetest.after could be run in an event handler...
19:50 sofar       it could run almost immediately (since the next globalstep can trigger instantly if the event handler ran right as it was due, or overdue)
19:50 GreenDimond and the default is 0.09 I guess
19:50 sofar       so if your server step is 0.09, then the answer is [0.0, 0.09>
19:51 GreenDimond anything faster than 0.09 is basically instant, no?
21:06 rubenwardy  nerzhul: any other suggestions for #8354   ?
21:07 ShadowBot   https://github.com/minetest/minetest/issues/8354 -- Fix cast from const by accessing string data directly by rubenwardy
21:42 kaeza       It seems the advanced settings list does not update when you install a mod containing a settingtypes.txt file from the CDB.
21:42 kaeza       After a restart it shows the missing settings.
21:43 kaeza       Tested with texmex's item_drop mod.
21:46 kaeza       Is this known or should I open an issue?
21:46 rubenwardy  not known
21:52 kaeza       #8360
21:52 ShadowBot   https://github.com/minetest/minetest/issues/8360 -- Advanced settings list does not update after installing mods.
21:53 rubenwardy  thanks
21:53 kaeza       Np
22:04 rubenwardy  nerzhul, sfan5: would it be ok to force push to backport-5 to clean up the 5.1.0 bump issues?
22:04 sfan5       of course
22:04 sfan5       forcepush rules don't apply to wip/inofficial branches
22:58 paramat     thanks, looks like you've added those 2 'num emerge threads' commits to backport-5, so now #8357 can be added to backport-5 too
22:58 ShadowBot   https://github.com/minetest/minetest/issues/8357 -- num_emerge_threads: Warn of crashes when > 1 by paramat
23:26 nerzhul     paramat no problem for me too on a backport branch