Minetest logo

IRC log for #minetest-dev, 2022-07-31

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

All times shown according to UTC.

Time Nick Message
01:00 Zughy[m] is #12131 really that needed for 5.6?
01:00 ShadowBot https://github.com/minetest/minetest/issues/12131 -- Add `minetest.settings` to CSM API and allow CSMs to provide `settingtypes.txt` by AFCMS
02:58 kilbith joined #minetest-dev
03:09 Juri left #minetest-dev
04:00 MTDiscord joined #minetest-dev
05:39 calcul0n joined #minetest-dev
06:02 cranezhou joined #minetest-dev
08:18 behalebabo joined #minetest-dev
09:04 Warr1024 joined #minetest-dev
09:08 Wuzzy joined #minetest-dev
10:31 x2048 joined #minetest-dev
10:43 HuguesRoss joined #minetest-dev
11:13 MTDiscord <ROllerozxa> is it intentional that the hotbar has become bigger sometime in the latest weeks?
11:48 Taoki joined #minetest-dev
11:55 sfan5 no
12:23 sfan5 https://0x0.st/oe8C.png 5.5.1; https://0x0.st/oeKr.png current git
12:24 sfan5 i see that the wieldhand color has changed though, previous looks more correct to me too
12:24 sfan5 maybe @x2048 knows
12:26 sfan5 Zughy[m]: not at all
12:26 Krock isn't that #12578 ?
12:26 ShadowBot https://github.com/minetest/minetest/issues/12578 -- Texture overlay color of items is different as wielditem than in inventory/dropped item
12:27 sfan5 could be, there's no texture overlay involved here though
12:27 Krock it could as well be that overlays are not the actual issue
12:27 sfan5 yes
12:29 Krock does your wieldhand correct with disabled shaders? I suppose it's caused by some HSV-like color shift
12:30 Krock *look correct
12:31 sfan5 https://0x0.st/oeK-.png its a shader darker but not the same color yet
12:32 Krock hmm what could've possibly changed in the meantime if not shaders?
12:34 sfan5 light calculation maybe
12:35 sfan5 gitlab build also crashes when exiting from world, this is bad
12:36 Krock https://github.com/minetest/minetest/commit/604fb2b73
12:36 sfan5 all expected then
12:38 Krock setColor -> applyFacesShading
12:41 Krock from what I can see, only final_color_blend (for lplayer->light_color) performs a blue shift
12:47 MTDiscord <ROllerozxa> forgot to say it occurs on android, might be another issue related to the settings clamp commit
12:49 Krock hmm.. the minimum was taken as 1.0
12:50 Krock oh it's a reading mistake on my side
12:50 Krock > hud_scaling (HUD scaling) float 1.0 0.5 20
12:51 Krock 1.0 is the default, not the minimum-
12:51 MTDiscord <ROllerozxa> ...also, uh, the default ogles1 driver appears to be busted too ?
12:53 rubenwardy idea: add a dialog to Minetest Game when the user first joins in singleplayer saying "Minetest Game is designed to be modded. You can install mods from Content > Browse online content" and maybe something about other games
12:56 sfan5 slightly different idea: present the currently selected game's description on first join or whatever
12:56 sfan5 s/first join/first main menu run/
12:56 rubenwardy Zughy's main menu redesign has space for a game description when in the singlemode page
12:56 rubenwardy -mode +player
12:57 Krock will push https://krock-works.uk.to/u/patches/0001-HUD-Fix-wrong-minimum-scale-since-051181f.patch and merge #12604 in 20 minutes
12:57 ShadowBot https://github.com/minetest/minetest/issues/12604 -- Fix Android blank screen by rollerozxa
12:58 Krock #12618 and #12587 too (2x bugfixes)
12:58 ShadowBot https://github.com/minetest/minetest/issues/12618 -- Apply NetBSD string conversion workaround to OpenBSD as well by jeremyshannon
12:59 ShadowBot https://github.com/minetest/minetest/issues/12587 -- Fix rotation of falling facedir nodes in some cases by grorp
12:59 sfan5 <+sfan5> gitlab build also crashes when exiting from world, this is bad
12:59 sfan5 self-built build doesn't crash when exiting world, probably some subtle toolchain issue
13:02 sfan5 (this is about the win64 build in both cases)
13:04 ROllerozxa rubenwardy: are you able to reproduce that the android version of latest master crashes with OGLES1 when starting a world? (with the screen resolution fix applied ofc)
13:04 MTDiscord <Jonathon> @x2048
13:17 Krock merging
13:18 Krock done
13:49 sfan5 even if we don't release today I guess I could import translations later
14:09 Fixer joined #minetest-dev
14:10 Fixer did some updates to ukrainian translation, there are still some rough parts ofc
14:30 Baytuch joined #minetest-dev
14:44 Yigidoo joined #minetest-dev
14:45 Yigidoo left #minetest-dev
14:49 rubenwardy ROllerozxa: works for me on your screen fix PR
14:50 ROllerozxa alright, must just be that the CI apks are broken then
14:50 rubenwardy or your device. Android devices vary a lot, part of the trouble with supporting it
14:50 rubenwardy I use a Google Pixel 4a
14:58 MTDiscord <TMcSquared> Yo Ruben, ( haven’t checked yet) but is there any momentum towards x2048’s work being integrated?
15:04 MAGNETAR joined #minetest-dev
15:05 MTDiscord <Jonathon> you really should check first instead of asking others to check for you https://github.com/minetest/minetest/pull/12465
15:06 MTDiscord <Jonathon> https://github.com/minetest/minetest/pulls?q=is%3Apr+author%3Ax2048+
15:20 ROllerozxa I went ahead and tried with a self-built apk from latest master instead of CI and it works for me too, so nothing to worry about I guess.
15:25 rubenwardy TMcSquared: x2048 is a core dev.Shadows have been merged, the post processing PR is on the 5.7 milestone
15:27 MTDiscord <TMcSquared> Imma be honest I forgot about that XD
15:28 MAGNETAR joined #minetest-dev
15:29 MAGNETAR joined #minetest-dev
15:29 kilbith joined #minetest-dev
15:30 MAGNETAR joined #minetest-dev
16:46 kilbith hmm, this code assumes that `wchar_t` is 4 bytes: https://github.com/minetest/minetest/blob/master/src/network/networkpacket.cpp#L201-L202
16:47 kilbith whilst most OSes are 2 bytes
16:47 kilbith so obviously my compiler complains as the comparison is always false
16:49 kilbith (compiler = MinGW x86-64)
16:52 kilbith should probably add a `sizeof`
17:01 kilbith joined #minetest-dev
17:10 kilbith joined #minetest-dev
17:23 sfan5 the code will work correctly see NEED_SURROGATE_CODING
17:27 sfan5 will import translations now
17:50 sfan5 the "A new $1 version is available" string really needs an annotation, I have seen two mistranslations so far
17:50 sfan5 or changed to 'version of $1' I guess
17:52 kilbith joined #minetest-dev
18:02 kilbith joined #minetest-dev
18:04 sfan5 doine
18:04 sfan5 done*
18:05 nrz sfan5: $1 is the new version ? if yes i suggest New version $1 is available
18:05 sfan5 $1 is the project name so "Minetest"
18:05 nrz ah ! okay then ignore ?
18:48 proller joined #minetest-dev
18:57 kilbith joined #minetest-dev
19:10 proller joined #minetest-dev
19:16 MTDiscord <x2048> Is there a meeting today?
19:18 HuguesRoss Supposed to be, not sure where the others are
19:18 HuguesRoss Could be we're all waiting for someone else to start it :D
19:19 HuguesRoss Going by the notes, it's mostly just the 5.6 wrap-up work
19:23 rubenwardy oops, was supposed to be 20 minutes ago
19:25 Yad joined #minetest-dev
19:25 Krock sfan5 rubenwardy HuguesRoss @x2048 nrz  let's quickly check the few meeting points
19:26 sfan5 hi
19:26 Krock looks like we're not ready for a 5.6.0 release today. well, I cannot say much - was (and will be) rather busy irl
19:26 Krock hello there :)
19:27 Krock > Who writes the release changelog?
19:27 Krock as for the Wiki, I already wrote most of it, and Wuzzy added MTG and edited a few lines
19:27 HuguesRoss Not me, I'm mostly back to normal but better not to tempt fate with a ton of extra typing
19:30 Krock glad to hear that you're getting well again
19:32 MTDiscord <x2048> As for #12608, I'll provide a comment and are we good to merge?
19:32 ShadowBot https://github.com/minetest/minetest/issues/12608 -- Use legacy image implementation when not using 9-slice image by x2048
19:33 sfan5 sure
19:33 HuguesRoss I don't have an issue as long as it's commented. Krock?
19:33 Krock seems good to me. I tested it briefly and it works well.
19:33 MTDiscord <x2048> Krock: I'd not like to add extra code to GUIAnimatedImage, is that fine with you?
19:33 sfan5 Krock: thanks for writing most of the changelog
19:34 Krock my patch would be more of a long-term lasting change, and is not really necessary at the moment
19:34 MTDiscord <x2048> switched from fancy shaders to master and heads-down with real work
19:36 sfan5 the long term change is to fix the bug ;)
19:37 Krock or rather to define where and where not to apply gui scaling/antialiasing
19:38 Krock what about #7629? It's in the milestone, so is it intended for 5.6.0?
19:38 ShadowBot https://github.com/minetest/minetest/issues/7629 -- Run Minetest update checker on startup by SmallJoker
19:39 sfan5 i thought the plan was to add it in 5.6
19:39 sfan5 so I've been waiting for a second approval
19:39 Krock yes, I wonder about the second approval too. I tested the change, but approving it with only like 50% being my work seems rather odd
19:40 Krock well I guess it does not matter since I could just merge it like this
19:41 Krock if anyone would like to have a look at it, please do that soon. I'll merge it tomorrow otherwise
19:43 Krock testing/reviewing #12607 ....
19:43 ShadowBot https://github.com/minetest/minetest/issues/12607 -- [nosquash] Update credits tab by sfan5
19:44 kilbith joined #minetest-dev
19:45 rubenwardy oh oops again, back
19:45 Krock sfan5: is the contributor order random or somehow based on a timestamp or count?
19:45 MTDiscord <osman.> Hi
19:45 rubenwardy it's based on a Python script using an algorithm I came up with
19:45 rubenwardy contributors are weighted based on an arbitrary heuristic
19:45 MTDiscord <osman.> Hello can anyone see my chat
19:46 Krock @osman. yes
19:46 MTDiscord <GreenXenith> Im going to change the dev bridge rank limit in light of recent user influx, apologies for intrusion
19:46 MTDiscord <osman.> Okay
19:46 rubenwardy (the heuristic basically counts the number of commits, weighted based on size)
19:47 rubenwardy https://github.com/minetest/minetest/blob/master/util/gather_git_credits.py
19:47 Krock that's more sophisticated than I expected
19:47 rubenwardy used to be a gross bashscript I wrote, thank sfan for rewriting it
19:50 MTDiscord <x2048> sfan5: I updated #12608, could you take a look?
19:50 ShadowBot https://github.com/minetest/minetest/issues/12608 -- Use legacy image implementation when not using 9-slice image by x2048
19:54 Krock ^ LGTM
19:54 MTDiscord <x2048> I'll merge it now then.
19:57 Krock then there's #12131
19:57 ShadowBot https://github.com/minetest/minetest/issues/12131 -- Add `minetest.settings` to CSM API and allow CSMs to provide `settingtypes.txt` by AFCMS
19:57 MTDiscord <x2048> merged
19:58 HuguesRoss looks like that one had an issue that sfan5 provided a fix for?
19:58 sfan5 yes
19:58 sfan5 someone just has to re-test it (not me, sorry)
19:59 Krock I'll give it a shot
19:59 HuguesRoss rubenwardy, you gave it a thumbs-up
19:59 HuguesRoss does that mean you tested or just a general +1?
19:59 rubenwardy gave what?
19:59 HuguesRoss To the fix sfan suggested on #12131
19:59 ShadowBot https://github.com/minetest/minetest/issues/12131 -- Add `minetest.settings` to CSM API and allow CSMs to provide `settingtypes.txt` by AFCMS
20:00 rubenwardy oh no, that's just saying it looks about right
20:00 rubenwardy it's not an approval
20:00 HuguesRoss gotcha, just checking
20:02 Krock it seems that cmake still caches the irrlicht version
20:02 Krock > At least IrrlichtMt 1.9.0mt7 is required to build
20:03 Krock > commit ff645cc8764ab3fc79f0e67677a328ba36c35087 (HEAD -> master, tag: 1.9.0mt7, origin/master, origin/HEAD)
20:08 sfan5 do you have it inside lib/irrlichtmt?
20:08 sfan5 not sure how cmake caches imported targets anyway
20:09 Krock no, it's in a separate location
20:09 Krock > Found IrrlichtMt 1.9.0
20:09 Krock no trace of mt7, which is rather strange
20:13 sfan5 https://0x0.st/oeZW.jpg supposed to look like this
20:14 Krock yes, I noticed that the cmake file directory changed at some point. cmake was pointing to an old 2021 file
20:16 MTDiscord <x2048> Noticed a problem in the latest master and latest MCL2, #12621.  Might be blocker if confirmed?
20:16 ShadowBot https://github.com/minetest/minetest/issues/12621 -- Random freezes in Mineclone 2 0.77
20:19 sfan5 I have some easy to replicate stutter when just flying forwards in a fresh MTG world
20:19 sfan5 though that was "normal" since I swapped my gpu for an AMD one recently
20:21 sfan5 (just minor, but very noticable)
20:22 sfan5 https://0x0.st/oeZy.png easily correlates with mesh building
20:23 sfan5 anyway back to the meeting: the most important question is when do we release?
20:23 sfan5 given current issues it sounds like at the earliest wednesday
20:24 sfan5 maybe next weekend otherwise
20:27 appguru joined #minetest-dev
20:27 Krock another depth sorting issue?
20:34 MTDiscord <x2048> Can't say for sure whether it's depth sorting. Does not look like any issue I saw earlier. ¯_(ツ)_/¯
20:35 Krock does transparency_sorting_distance = 0  avoid the issue?
20:37 MTDiscord <x2048> Reg. release, let's aim for Wednesday. Is there anything to be done that is not tracked in the milestone?
20:40 sfan5 don't think so
20:46 Sokomine sfan5: stutter? i still have client hangs when walking along your land main street. it's not as worse as before because due to bugfixes and improvements in mods the amount of entities has been hugely reduced
20:52 sfan5 https://0x0.st/oeZd.png this looks suspicious
20:52 Baytuch joined #minetest-dev
20:55 rubenwardy is anyone seeing messed up node sides? https://rwdy.uk/nR44p.png
20:55 rubenwardy it's possible my graphics is broke
20:57 sfan5 wtm
20:57 sfan5 wfm*
20:59 Krock your graphics is broke
21:02 rubenwardy it broke :(
21:06 sfan5 as it turns out iterating the same data structure 20000 times a second(?) can get pretty slow
21:07 HuguesRoss you don't say
21:08 MTDiscord <x2048> sfan5: are you talking about addBlock->cleanupCache()?
21:08 sfan5 yes
21:16 MTDiscord <x2048> it's a good catch, but how does is explain the stutter?
21:17 sfan5 iterating the cache is apparently pretty expensive
21:17 sfan5 and since it happens on the main thread it can get so bad as to delay the rendering
21:18 sfan5 btw if you have multi second freezes it will be trivial to make sense of them using perf
21:18 Baytuch joined #minetest-dev
21:18 sfan5 perf record -z --call-graph dwarf ./bin/minetest, throw the file into https://github.com/KDAB/hotspot, filter for the main thread and switch to flamegraph view
21:19 MTDiscord <x2048> Thanks!
21:20 sfan5 #12622 btw
21:20 ShadowBot https://github.com/minetest/minetest/issues/12622 -- MeshUpdateQueue fixes by sfan5
21:23 sfan5 back to the last point: why trivial? you can also filter for span of times and if you have a multi second one that's easy
21:37 calcul0n joined #minetest-dev
21:45 MTDiscord <x2048> Krock: transparency_sorting_distance = 0 does not avoid the problem.
22:02 MTDiscord <x2048> sfan5: my perf does not recognize -z
22:12 MTDiscord <x2048> nvm
22:34 panwolfram joined #minetest-dev
23:06 Baytuch joined #minetest-dev
23:13 Baytuch joined #minetest-dev
23:28 Baytuch joined #minetest-dev

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