Time Nick Message 07:31 nerzhul sfan5 i just redeployed minetest outside of k8s on my rock64 arm board and the udp unreliable packets are received but it seems we have a problem, i should debug more 07:33 nerzhul ah after a /killme i spawned in a better place, it seems i spawned in stone in fact and cannot dig... lol :) 08:19 Krock 2020-04-13 10:05:42: WARNING[ConnectionSend]: con(22/3) Packet quota used up for peer_id=1, was 1009 pkts 08:20 Krock connected to a server, and timing out after a certain time 08:31 Krock will merge #9101 in 15 minutes 08:32 ShadowBot https://github.com/minetest/minetest/issues/9101 -- Add scroll_container formspec element (redo) by Desour 08:33 Krock #8873 too 08:33 ShadowBot https://github.com/minetest/minetest/issues/8873 -- Add default stack size setting by SmallJoker 08:34 Krock unless.. rubenwardy ? 08:36 nerzhul cool 08:47 Krock merging (2) 08:53 Krock done 09:10 nerzhul i'm developing the missing pg auth backend 09:11 nerzhul i will propose a PR today, it should be quite trivial as it's just the same backend than our sqlite backend :p 09:22 Krock will merge game#2627 in 10 minutes 09:22 ShadowBot https://github.com/minetest/minetest_game/issues/2627 -- Make straw stairs usable as fuel by pauloue 09:33 Krock mergign 10:17 nerzhul Krock, can you review the WorldSettings PR please ? :) 10:18 nerzhul it's a more robust interface for our world.mt file in the core engine 10:18 nerzhul i have a PostgreSQL auth backend PR ready but it requires this PR (currently) because i coded it on this branch haha 10:23 nerzhul #9647 10:23 ShadowBot https://github.com/minetest/minetest/issues/9647 -- Add WorldSettings object & env var support by nerzhul 11:00 Krock put on my todo list. 11:17 nerzhul thanks Krock 11:29 Wuzzy Looks like minimal was touched recently again ... 11:30 Wuzzy I would really appreciate if DevTest is merged so the two code bases no longer diverge all the time … 11:30 sfan5 yes, by https://github.com/minetest/minetest/commit/0ac999ded725f8efcd26db284161683e37efeecf 11:31 Wuzzy hmmm i already added my own hypertext[] test in DevTest 11:32 Wuzzy so i dont need to cannibalize this ? 11:32 Wuzzy but the scroll_container[] 11:32 Wuzzy ok this is new 11:32 Wuzzy sfan5: any other touches to minimal recently? 11:32 Wuzzy just making sure i did not miss anything 11:33 sfan5 https://github.com/minetest/minetest/commits/master/games/minimal 11:34 Wuzzy i need to up my git skills, really ? 11:35 sfan5 on the CLI you'd get the same result using `git log games/minimal/` 11:43 Zughy wasn't there a GitLab mirror for Minetest? I'm sure I saw it yesterday but I can't seem to find it on GitLab anymore 11:44 Wuzzy *X-Files music plays* 11:46 sfan5 https://gitlab.com/minetest/minetest/ this one? 11:46 Krock https://gitlab.com/minetest/minetest/pipelines/ 11:46 Krock these ninjas 11:47 Zughy wth, yes. I swear I couldn't find it, not even via the GitLab search bar. Ty guys 11:49 Wuzzy woah this scroll container test looks really broken 11:50 Wuzzy did nobody test scroll containers before merging? ? 11:50 Wuzzy scroll container content bleeds out of my formspec 11:53 Wuzzy hmm maybe i just fail at copying ? 12:10 Wuzzy OK, I just updated DevTest with the latest formspec test changes ^^ 12:11 Wuzzy Should I update the PR, too? 12:12 Wuzzy ( I do not update the DevTest PR without request to avoid interferring with the review process. ) 12:47 DS-minetest !tell Zughy regarding http://irc.minetest.net/minetest-dev/2020-04-12#i_5667501 : the main menu formspec is fixed size: https://github.com/minetest/minetest/blob/27d611fe5561db20b380a16fdc6bcf1fefaf5d39/src/gui/guiEngine.cpp#L179 12:47 ShadowBot DS-minetest: O.K. 12:49 Zughy yup, thank you, in the end I figured it out yesterday :D 12:51 Zughy a question about formspecs tho: if a formspec didn't appear at the very launch of the game, the cursor wouldn't be visible? 12:58 Krock the cursor is only visible in formspecs 12:58 Krock because in-game there's the crosshair that replaces the cursor (since it's center-focussed) 15:55 kilbith btw the hypertext example in formspec.lua is wrong 15:56 kilbith this is `hovercolor` not `actioncolor` 16:18 Krock presumably some old testing code, before it was renamed 16:19 Krock might be interesting to Wuzzy ^ (devtest minimal) 16:21 kilbith possible bug discovered with scroll_container 16:39 nerzhul sfan5: can you be more precise on #9647 ? 16:40 ShadowBot https://github.com/minetest/minetest/issues/9647 -- Add WorldSettings object & env var support by nerzhul 16:41 sfan5 I was thinking instead of .getRedisPort(), .getRedisWhatever(), ... methods there could be a .get("redis_port") which checks getenv("WORLD"_ + name.uppercase()) first and falls back to world_mt.get(name) 16:41 sfan5 so kinda like a settings object 16:41 sfan5 not sure if this matches your intention with the WorldSettings class 16:42 Krock just re-use the Settings. tbh I don't see much benefits from adding another class 16:44 * Krock reads the code 16:45 Krock an enum + LUT could restore the generic intends of Settings. Defaults can be handled using setDefault() which is used for regular settings 16:49 Krock for example: conf->get(WORLD_PLAYER_BACKEND); or conf->set(WORLD_AUTH_BACKEND, backend, save?); if you'd like to get rid of the setting name constants 16:54 Zughy guys, in order to better understand the code (since it's huge), wouldn't be better to comment at least some functions in order to avoid searching high and low? Understanding some parts is giving a headache to me and my friends, I humanly don't understand how you can remember it all (kudos). Nor I wanna stress you every time 16:55 Zughy or if there's something explaining the architecture, idk 16:58 sfan5 - << " Syntax error in texture override \"" << line 16:58 sfan5 + << " Syntax error in texture override \"" 16:58 sfan5 another great suggestion by clang-format 16:59 sfan5 consistent indentation? nah man why not mix tabs and spaces? 16:59 nephele 8tabs and 4 spaces 16:59 nephele prereable in a random order 17:00 rubenwardy clang-format is good if you configure it correctly 17:00 rubenwardy I have it configured to only use tabs for indentation, and it works well 17:00 sfan5 in case it doesn't show correct in irc: the first is "" and the suggestion is " " 17:03 sfan5 rubenwardy: well in this case it's using spaces for alignment, but this is extremely stupid to do at the beginning of the line 17:03 rubenwardy that's from the LLVM codestyle, which is used as the base of our linter 17:04 sfan5 Zughy: do you have an editor that allows you to jump to the function/class/variable definition by clicking on it? helps a lot 17:05 sfan5 and since you mentioned gui stuff: understanding guiFormSpecMenu.cpp definitely isn't an easy task 17:06 Zughy yeah, there should be some addon for Atom to do that 17:06 Zughy and yes, that file is long. 17:08 Zughy that's why I headed to the builtin part in lua without touching a lot in C++, as studying Irrlicht docs too wasn't that funny 17:08 rubenwardy oh look: https://github.com/torvalds/linux/blob/master/.clang-format 17:11 rubenwardy updated #9650 17:11 nerzhul sfan5 it doesn't really matches because the idea was to be specialized for the core needs instead, it's why i didn't exposed the generic get & set methods 17:11 ShadowBot https://github.com/minetest/minetest/issues/9650 -- Log deprecation initial properties by rubenwardy 17:12 rubenwardy tbh, I don't understand #9647 17:12 ShadowBot https://github.com/minetest/minetest/issues/9647 -- Add WorldSettings object & env var support by nerzhul 17:13 rubenwardy well, not all - I can see how having environment variables for postgresql etc connections can be useful 17:14 rubenwardy the boot script used by kubernetes/docker could do that instead, however 17:14 rubenwardy oh good - I thought it allowed anything in world.mt, but it is just connection stuff 17:22 rubenwardy Wuzzy: the hypertext tag can appear in two places - in a hypertext tab (which is nice) and the scroll container tab 17:23 rubenwardy having them in multiple places is nice, to be able to check how they work in different contexts 17:24 rubenwardy devtest is very nice 17:49 nerzhul rubenwardy the goal of the PR is twice: 1. ensure we can load in a unified way our WorldSettings when reading them (Map, Auth, Player DB connections) 2. Add env var native support permitting to have a good way to configure it without adding a wrapper script which does very nice sed craziness on non existent world.mt :D. And bonus, we have a good WorldSettings interface which is c++ friendly and uses the underlying Settings 17:49 nerzhul interface which is generic :) 17:50 nerzhul i will remind this on the PR 17:51 nerzhul remember this object is for us, not for modders it's our internals, and having maintainable class is important. Also i think the generic settings interface is nice, but when you want to look which settings are registered in core without looking our settings types, you just become crazy reading the code with grep & sed... :D 18:01 nerzhul i put infos more properly on github around what is said 18:15 sfan5 merging game#2127 in 10m 18:15 ShadowBot https://github.com/minetest/minetest_game/issues/2127 -- Make default.chest.register_chest() usable for other mods by bell07 18:21 Krock yet no helpful backtrace function in Minetest for nested reader functions sadly.. 18:21 sfan5 "nested reader"? 18:21 Krock probably another reason why field value types aren't error-checked in c_content.cpp 18:21 sfan5 merging #9069 too 18:21 ShadowBot https://github.com/minetest/minetest/issues/9069 -- Remove sound menu and show proper messages if trying to change sound volume when sound if disabled/unsupported by Wuzzy2 18:22 Krock yes, for example the function that reads the "sounds" field within the node definition 18:22 Krock Lua does not error for that field, but for the entire function call 18:23 Krock which is presumably minetest.register_node or its wrapper