Time  Nick      Message
00:13 Unarelith rubenwardy, nerzhul, I rebased 5395 into #7971, but changes requested in the previous PR has not yet been made, I wait your opinion on how this should be done
00:13 ShadowBot https://github.com/minetest/minetest/issues/7971 -- FormSpec: support custom size and spacing for slots in list by Quent42340
10:00 nerzhul   sfan5 are you around ?
10:03 nerzhul   or another coredev to review a doc pr
10:05 nerzhul   merging trivial #7964
10:05 ShadowBot https://github.com/minetest/minetest/issues/7964 -- Make showOverlayMessage strings translatable (take 2) by Wuzzy2
10:20 nerzhul   merging #7768
10:20 ShadowBot https://github.com/minetest/minetest/issues/7768 -- Network: Send IEEE floats by SmallJoker
13:15 p_gimeno  thank you thank you thank you!
13:46 nerzhul   give me money now
13:46 nerzhul   :D
13:47 nerzhul   why are you so enthusiast for this merge ?
14:05 p_gimeno  I have been cringing since I saw that floats were transmitted with just 3 decimal places. It improves precision everywhere, and removes a lot of potential problems, including #5521 which I think can be closed now.
14:05 ShadowBot https://github.com/minetest/minetest/issues/5521 -- Server hangs when exceeding position x/y/z near 229367
14:06 p_gimeno  It looked like something very basic that should have been there since 0.1.
14:15 p_gimeno  or maybe 5521 can't be closed yet, until all read/writeF1000 are changed to read/writeF32, which I believe is in Krock's plans
14:16 p_gimeno  but if the cause was the position, it probably can
18:01 Krock     thanks nerzhul for the F32 spproval
18:01 Krock     very nice.
18:01 p_gimeno  +1
18:03 nerzhul   no problem, i took time to ensure it was correct after the war :p
18:03 nerzhul   it's a very important PR for float precision, it was in my 5.0.0 roadmap and remove some historical bugs, but it's a very sensitive serialization, it's why i take some time :)
18:05 Krock     ikr
18:06 nerzhul   Krock: what about my docker doc pr ?
18:08 nerzhul   maybe i will provide K8S template after that for the big minetest networks on k8s
18:09 Krock     nerzhul: hit F5
18:12 nerzhul   what od you mean ? you want me to add ``` marker ?
18:12 nerzhul   do*
18:12 nerzhul   or add _ _ markers ?
18:13 Krock     single quote `
18:13 nerzhul   yeah i'm bad on this and don't really understand which result do you want :p
18:13 nerzhul   oh
18:13 Krock     or is that not a valid markdown formatting character? could be a github-only thing tho
18:13 Krock     at least in the comments it'll result in inline code formatting
18:14 nerzhul   no it's valid
18:14 nerzhul   it's fixed
18:14 nerzhul   is this okay ?
18:22 nerzhul   i will go to lunch, if it's okay i merge both pr when i come back
18:44 p_gimeno  Krock: do you plan on replacing most/all of *F1000 calls to *F32?
18:47 Krock     p_gimeno: most if possible
18:47 Krock     question is.. when
18:55 celeron55 nice job handling the float serialization so far
18:57 Krock     I think barely any other project cares as much we did here :3
18:58 p_gimeno  Krock: it's quick, find src -type f -exec sed -ri 's/F1000/F32/g' {} \; ;)
18:58 p_gimeno  (kidding)
19:00 Krock     it wouldn't even compile like that lol
19:01 p_gimeno  I think it would be pretty close, though
19:01 p_gimeno  anyway, it obviously needs more attention to detail
19:01 Krock     ofc
19:17 p_gimeno  hmm, the unit tests of 7768 may fail in gcc -ffast-math
19:18 nerzhul   ffast-math is not recommended with mt if i remember
19:18 nerzhul   merging #7969 & #7939
19:18 ShadowBot https://github.com/minetest/minetest/issues/7969 -- Add docker image documentation by nerzhul
19:18 ShadowBot https://github.com/minetest/minetest/issues/7939 -- Add an activeobject manager to hold active objects by nerzhul
19:18 nerzhul   ty Krock for the review
19:18 p_gimeno  nerzhul: it's active in release builds IIUC
19:19 nerzhul   maybe i'm wrong then
19:19 nerzhul   don't hesitate to open a blocker
19:20 Krock     np
19:21 p_gimeno  nerzhul: in fact it's causing a unit test to fail already, see #3943 (bottom)
19:21 ShadowBot https://github.com/minetest/minetest/issues/3943 -- testStreamRead and testBufReader failures
19:27 p_gimeno  and -ffast-math also enables -ffinite-math-only which causes problems with the unit tests of NaNs
19:28 p_gimeno  worst is, g++ reports itself as iec559 compliant but isn't - https://gcc.gnu.org/bugzilla//show_bug.cgi?id=84949
19:34 nerzhul   float powa
19:44 Unarelith p_gimeno, yeah I have this issue when I build in release mode
19:46 nerzhul   use clang ? :D
19:54 Unarelith nerzhul, same issue with clang
19:54 Unarelith :)
19:56 nerzhul   arf
19:57 p_gimeno  [FAIL] testFloatFormat - 0ms
19:57 p_gimeno  yeah, as expected
20:12 p_gimeno  #7973
20:12 ShadowBot https://github.com/minetest/minetest/issues/7973 -- Fix unit test in release mode (consequence of -ffast-math)