Minetest logo

IRC log for #minetest-dev, 2020-05-24

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

All times shown according to UTC.

Time Nick Message
01:01 ANAND joined #minetest-dev
01:10 ANAND appguru has already made a PR while I was asleep :)
01:15 nephele_ joined #minetest-dev
03:03 Lunatrius` joined #minetest-dev
04:21 calcul0n joined #minetest-dev
05:21 ANAND joined #minetest-dev
05:48 ANAND appguru has a PR ready, while I've only familiarised myself with that region of code, and haven't committed a single line yet.
05:49 ANAND I'd rather use my time to help review that PR, than come up with a very similar implementation.
05:49 ANAND Sorry to disappoint you, Krock ;)
06:09 NetherEran joined #minetest-dev
06:31 YuGiOhJCJ joined #minetest-dev
08:00 ShadowNinja joined #minetest-dev
08:17 T4im joined #minetest-dev
08:18 ANAND sfan5: I edited my HP in my minimal world's players.sqlite file - it works now.
08:19 ANAND (edited externally, that is)
08:21 ANAND Aha
08:22 ANAND My old HP was 3
08:22 ANAND Now, see https://github.com/minetest/minetest/blob/10c3002aea784b5f0075f3f3e3ec824b6ba546ba/games/minimal/mods/test/player.lua#L12
08:22 ANAND player:set_hp(3)
08:23 ANAND Since both the values are equal, the HP change callback didn't run
08:25 ANAND The assertion fails because expect, which is set to a non-nil table in L11, is set back to nil only within the callback, which didn't run in this case.
08:29 ANAND This behaviour change was introduced by #8494.
08:29 ShadowBot https://github.com/minetest/minetest/issues/8494 -- PlayerSAO::setHP - Don't call on_hpchange callbacks if HP hasn't changed by ClobberXD
08:29 ANAND This commit caused caused minimal to crash all the time irrespective of player HP, see See #8510.
08:29 ShadowBot https://github.com/minetest/minetest/issues/8510 -- Minimal crash with enabled damage
08:30 ANAND But I certainly did not foresee this.
08:30 ANAND It'd be nice if the player's HP was temporarily set to max_hp, then these callbacks are "enabled", and then the tests are run.
08:31 ANAND On it o7
08:35 ANAND Note: This happens only when the player's HP is exactly 3
08:38 Darcidride joined #minetest-dev
09:07 ANAND Now that I think about it, there's a very remote chance for someone to come across this crash twice in a row. :O
09:10 ANAND #9926
09:10 ShadowBot https://github.com/minetest/minetest/issues/9926 -- minimal: Fix assertion failure when player HP == 3 by ClobberXD
09:12 sfan5 devtest if (hopefully) about to be merged so wait for that
09:12 sfan5 ..before making changes to minimal
09:18 ANAND Oh that's good news. Is it going to happen soon?
09:20 sfan5 good question
09:20 sfan5 I could merge it right now but I wanted to wait for rubenwardy to do it because he initially reviewed it
09:20 sfan5 ANAND: do you plan to add any more corrections to #9917?
09:20 ShadowBot https://github.com/minetest/minetest/issues/9917 -- lua_api.txt: Fix minor typo by ClobberXD
09:23 ANAND sfan5: I'm not planning to add to that PR
09:24 ANAND I did run enchant on the file, but the output was filled with stuff like method names, minetest, abbreviations, etc.
09:25 ANAND Given that our docs are mostly typo-free apart from very rare ones here and there, I didn't bother to sift through the output for valid fixes.
09:32 sfan5 okay
09:36 sfan5 merging #9917, #9919, #9920 in 10 minutes
09:36 ShadowBot https://github.com/minetest/minetest/issues/9917 -- lua_api.txt: Fix minor typo by ClobberXD
09:36 ShadowBot https://github.com/minetest/minetest/issues/9919 -- [NO SQUASH] Emerge fixes by sfan5
09:36 ShadowBot https://github.com/minetest/minetest/issues/9920 -- Fix two bugs in content_cao by sfan5
10:32 ANAND yay
10:36 NetherEran joined #minetest-dev
10:47 appguru joined #minetest-dev
11:00 Beton joined #minetest-dev
11:09 erlehmann joined #minetest-dev
11:50 Fixer joined #minetest-dev
11:59 NetherEran joined #minetest-dev
12:13 Krock will merge #9736 in 10 minutes
12:13 ShadowBot https://github.com/minetest/minetest/issues/9736 -- Add chat_font_size setting by SmallJoker
12:15 nephele can it also follow the os? :)
12:17 Calinou the OS generally doesn't easily expose a font size setting
12:18 Calinou not to mention this is generally a setting you want to tweak manually
12:18 Calinou that said, we *could* increase the default font size by now
12:19 nephele well, Haiku definetely does, i think android does aswell, and i would imagine MacOS does too? not sure about windows and gnu/linux desktop though
12:19 Calinou it exists, it's not easy to access in a cross-platform manner
12:20 Calinou I think increasing the default size will solve 90% of font-related issues :P
12:20 Calinou which is why I don't think we should make the current system more complex
12:20 nephele Eh, i would rather it follow the guidelines my desktop lays out ;)
12:20 Calinou pretty much no game out there does that, though
12:20 Calinou they generally scale 2D elements based on the resolution, so the text is readable at any resolution
12:20 Calinou we could do this but it requires a bit more work
12:20 rubenwardy games shouldn't necessarily follow the desktop's guidelines
12:21 rubenwardy for example, it's rare to have games follow the system theme
12:21 Calinou not even Unreal does that with its desktop-like theme :P
12:21 Calinou (the first one, that is)
12:21 nephele For mintest i would definetely want it to follow the font settings, it isn't different enough (especially the chat) to not do that ;)
12:21 nephele but hey, options
12:21 Calinou a lot of people would rather have Minetest use a pixel font anyway
12:21 nephele maybe i should just try to make a patch
12:21 rubenwardy font size may be an exception because it allows you to detect what the user might find readable
12:22 nephele pixel font? you mean one based on images instead of vectors?
12:22 rubenwardy also, the system's font size will be tied to the font given
12:22 nephele what's the point of using that?
12:22 rubenwardy style
12:23 rubenwardy a vector font and a pixel UI can look weird
12:23 Krock merging
12:23 nephele vector fonts can look the same as a "pixel" fonts if you properly implement glyph hinting
12:23 Calinou yes, but only to an extent
12:24 Calinou pixel-perfect fonts are often designed on a pixel grid, even if they happen to be available in vector formats
12:24 Calinou (see Terminus)
12:24 Calinou disabling antialiasing completely will look horrible with most vector fonts out there :)
12:25 Krock just use comic sans everywhere and it's fine
12:25 nephele the problem with pixel-perfect is that they also depend on the sequence of color segments in your display :D, if you want you can ship severall fonts for that too
12:25 nephele personally i think it's not worth the effort to get one pixel different in some cases
12:25 nephele specially since glyph hinting is designed to not need that
12:26 Calinou well, the font would almost never be displayed at 1× scale anyway
12:26 Calinou so display pixel order wouldn't matter meaningfully
12:27 nephele well, if you don't have it render in native resolution i see no point at all in bitmap fonts :)
12:47 Wuzzy joined #minetest-dev
13:19 mntmn joined #minetest-dev
13:30 mntmn joined #minetest-dev
13:33 rubenwardy merging #9450 #9890 in 10
13:33 ShadowBot https://github.com/minetest/minetest/issues/9450 -- Replacement for Minimal Development Test (PR) by Wuzzy2
13:33 ShadowBot https://github.com/minetest/minetest/issues/9890 -- Add engine version string to ContentDB API request by rubenwardy
13:37 gd joined #minetest-dev
13:45 texmex rubenwardy: 9450, nice one
13:45 Krock 12 minutes have passed
13:46 nephele 5 seconds have passed!
13:46 sfan5 can we have a top list of "most missed merge minutes per developer"
13:46 rubenwardy I set an alarm, but them someone called me :P
13:48 Krock sfan5: who's going to maintain this?
13:49 Krock the idea is funny
13:49 sfan5 could write a script to do that, maybe
13:49 sfan5 won't be me tho
13:50 Krock tricky to find the corresponding commit that followed after the "merging" message
13:50 Krock although the next following commit date could be taken
14:04 Krock sfan5: are the logs available as raw text files?
14:04 sfan5 http://irc.minetest.net/minetest-dev/2020-05-24/text
14:04 Krock like http://irc.minetest.net/minetest-dev/2020-05-24/text  but monthly or os
14:04 Krock *so
14:04 sfan5 oh
14:05 sfan5 unless you find a link somewhere, no
14:05 Krock well then.. also using wget in this case
14:17 nephele rubenwardy: the a9c3a42323 commit seems to break the build on FreeBSD (one to use xdg-open to open uris), it complains about environ beeing undefined (I did not do a completely clean build, but can do that if needed)
14:20 rubenwardy I can't find the FreeBSD header that's defined in
14:20 rubenwardy try adding `extern char **environ;` above line 723 in porting.cpp
14:20 Darcidride joined #minetest-dev
14:20 rubenwardy I don't like that - I found such a case in a random webpage
14:21 rubenwardy https://www.freebsd.org/cgi/man.cgi?query=posix_spawnp&sektion=3&apr
14:21 rubenwardy docs seem to recommend that to?
14:21 rubenwardy +o
14:21 nephele 723 is in if defined android case, i assume you mean before the posix_spawnp call?
14:22 nephele or atleast now is on HEAD
14:22 rubenwardy https://www.freebsd.org/cgi/man.cgi?environ(7)
14:22 rubenwardy yeah, I meant here: https://github.com/minetest/minetest/pull/8592/files#diff-59f99d02fe5b328bb9a57b023497b5edR723
14:23 nephele That gives me CMakeFiles/minetest.dir/porting.cpp.o: In function `porting::openURL(std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&)': \n porting.cpp:(.text+0x16ae): undefined reference to `porting::environ'
14:23 nephele during linking
14:24 sfan5 rubenwardy: for OS APIs avialable from C/C++ programs you usually want the manpages from 2 or 3
14:24 sfan5 however environ(3) doesn't mention the header file either
14:24 Taoki joined #minetest-dev
14:24 nephele i don't have environ(3) :g
14:24 rubenwardy nephele: try moving it to the top of the file instead
14:24 rubenwardy outside of the namespace
14:25 nephele okay, i had put it on line 67 now, that compiled and runs
14:26 rubenwardy I hate porting code :D
14:27 nephele Haiku has open instead of xdg-open too ;)
14:27 rubenwardy argh
14:27 rubenwardy maybe I should try xdg-open then open
14:27 rubenwardy *it
14:28 nephele anyway, thanks runs now, so i can dev a bit on my gamemode to atleast give it a point pre-release thingy :D
14:30 rubenwardy fwiw: on linux, environ is defined in unistd.h
14:32 rubenwardy fix: https://github.com/rubenwardy/minetest/commit/212a1123e0fd930faf2467b09761815b849efcaf
14:33 rubenwardy I wonder whether this is dependent on libc?
14:34 rubenwardy In which case, the better fix may be to detect non-glibc or just always define it
14:34 nephele hmm, ill test it on musl, one moment :)
14:34 gorbachev_pizza joined #minetest-dev
14:35 sfan5 the docker build runs musl, that would have failed already
14:35 rubenwardy ah
14:41 appguru joined #minetest-dev
14:41 appguru Would like to see
14:41 appguru #9907 merged
14:41 ShadowBot https://github.com/minetest/minetest/issues/9907 -- Shaders as media files (server-sent shaders) by appgurueu
14:42 appguru The only concern I could see is security, which is why there's the "enable_shader_security" and "trusted_shader_servers" settings
14:42 appguru (there clearly are other concerns, but I think they are either irrelevant or have been addressed)
14:55 Wuzzy darn it. why do so many tests fail for #9928?
14:55 ShadowBot https://github.com/minetest/minetest/issues/9928 -- Rename “Minimal development test” to “Development Test” by Wuzzy2
14:56 appguru Wuzzy: probably because you haven't edited them :P
14:56 Wuzzy ???
14:57 Wuzzy I don't have access to these tests apparently ?
15:00 appguru Wuzzy: you need to edit the C++ code
15:00 appguru it contains the tests
15:00 Wuzzy where?
15:00 appguru grep - r "testGetModspec"
15:01 appguru test_servermodmanager.cpp apparently
15:01 appguru ayyy
15:01 appguru void TestServerModManager::testGetModspec(){ServerModManager sm(std::string(TEST_WORLDDIR));UASSERTEQ(const ModSpec *, sm.getModSpec("wrongmod"), NULL);UASSERT(sm.getModSpec("default") != NULL);}
15:01 appguru it effectively asserts that there is no "wrongmod" and that there is a "default" mod or something?
15:02 appguru There seem to be quite a few checks checking for "default" in that file
15:03 appguru Wuzzy: iirc your devtest doesn't contain a "default" mod?
15:03 rubenwardy that's failing in my branch too
15:03 Wuzzy yes
15:03 indiana joined #minetest-dev
15:04 rubenwardy master passes though?!
15:04 appguru [FAIL] testGetMods - 1ms
15:04 appguru [FAIL] testGetModspec - 1ms
15:04 appguru [FAIL] testGetModNames - 1ms
15:04 appguru all of those check for "default"
15:04 appguru tbh the tests are flawed
15:05 Wuzzy i agree
15:05 appguru so just edit that tests and check for the name of your devtest core mod
15:06 Wuzzy but then it will fail for MTG ?
15:07 appguru has devtest been merged btw?
15:07 Wuzzy yes
15:07 appguru yay
15:07 appguru I wonder why I haven't noticed
15:08 Wuzzy it was <30min ago
15:09 fluxflux joined #minetest-dev
15:10 sfan5 Wuzzy: the tests only run for minimal
15:10 sfan5 (or whatever is in the repository)
15:26 rubenwardy those tests probably shouldn't rely on minimal though
15:27 rubenwardy using a /tmp/ folder would be better imo
15:28 Wuzzy yay. all tests succeeded
15:28 Wuzzy yes the tests need improvement, but please dont demand this from my rename PR. ?
15:29 rubenwardy also - looks like the GitHub Actions are based on files changed even on master
15:29 rubenwardy the problem with that is if you make multiple commits, only the latest's files changed is used
15:29 rubenwardy meaning that not all the tests will run
15:29 rubenwardy see master
15:29 rubenwardy only the lua tests ran, because the latest commit is lua-only
15:31 rubenwardy sfan5: is this ok for you? https://github.com/rubenwardy/minetest/commit/212a1123e0fd930faf2467b09761815b849efcaf
15:32 sfan5 sure
15:32 rubenwardy cool, I'll merge it in an indefinite amount of time shortly
15:35 Krock https://pastebin.com/raw/4X0sivPR  stats on bottom
15:36 Krock >19:06 sfan5      merging (trivial) game#2651 in 5 minutes
15:36 ShadowBot https://github.com/minetest/minetest_game/issues/2651 -- Small sign changes by HybridDog
15:36 Krock need to confirm that
15:36 rubenwardy maybe average the minutes?
15:36 rubenwardy or use a threshold to show late merges
15:37 Krock no mercy. for former you can divide the minutes by #Msgs
15:38 rubenwardy the former isn't as useful, as it's biased by a few outliers
15:38 Krock well then.. merge more precisely. could do a median value though
15:44 Krock rubenwardy: Ctrl + F5. Median added. Still looks bad for you
15:44 Krock hmm
15:44 rubenwardy it didn't look bad for me to begin with
15:44 rubenwardy the median reveals it better
15:44 Krock lol
15:45 rubenwardy also, median with two samples :D
15:45 rubenwardy guess I should merge more
15:46 rubenwardy minetest dev slows down
15:47 Krock updated for entire April and May so far
15:51 Krock Wuzzy: devtest (minimal) broke the unittest
15:52 Wuzzy which one?
15:52 Krock https://github.com/minetest/minetest/runs/703998935#step:5:204
15:52 Krock basically every reference to "default"
15:53 Wuzzy this is a link to Aux????
15:54 Krock yes!!! and here's another https://github.com/minetest/minetest/runs/704000336#step:5:204
15:55 Wuzzy Where can I see a list of the breakages?
15:57 Wuzzy Krock: will #9928 fix this?
15:57 ShadowBot https://github.com/minetest/minetest/issues/9928 -- Rename “Minimal development test” to “Development Test” by Wuzzy2
16:11 Krock Wuzzy: https://github.com/minetest/minetest/blob/master/src/unittest/test_servermodmanager.cpp#L123-L156
16:11 Wuzzy Is this a no?
16:11 Krock yes, but I like the PR
16:12 Wuzzy are you sure?
16:12 Krock unittest creates a new world called "test_world" using the "minimal" game
16:12 Krock the world config is opened, saved and closed.
16:13 Wuzzy ah, so you didn'T test ?
16:13 Krock afterwards the marked lines check whether mods were loaded correctly
16:13 Krock I don't need to test when the affected lines aren't changed
16:13 Wuzzy What about: https://github.com/minetest/minetest/pull/9928/commits/7bd6cd755b0f1413bf951b1343f0ea59e6c3b586
16:14 Krock looks good
16:14 Wuzzy see?
16:14 Krock so the PR does more than renaming
16:14 Wuzzy I added it afterwards to repair the PR
16:14 Krock updating description...
16:14 Wuzzy ok thx
16:49 NetherEran joined #minetest-dev
16:51 ANAND devtest seems to require #9926 as well
16:51 ShadowBot https://github.com/minetest/minetest/issues/9926 -- minimal: Fix assertion failure when player HP == 3 by ClobberXD
16:51 ANAND I'll update the PR accordingly
17:14 appguru joined #minetest-dev
17:59 AntumD joined #minetest-dev
18:35 NetherEran joined #minetest-dev
18:53 Icedream joined #minetest-dev
19:18 el joined #minetest-dev
19:28 H4mlet joined #minetest-dev
19:37 H4mlet joined #minetest-dev
20:28 proller joined #minetest-dev
21:10 Wuzzy joined #minetest-dev
21:27 calcul0n_ joined #minetest-dev
22:05 appguru joined #minetest-dev
23:16 erlehmann joined #minetest-dev
23:46 realzies joined #minetest-dev

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