Time Nick Message 07:02 MTDiscord What's going on with COpenGLDriver.cpp vs. OpenGL/Driver.cpp in irrlicht? 08:38 sfan5 i think one is the opengl1/2 driver, one the opengl3 08:38 sfan5 or ask @numberZero 10:03 sfan5 the answer to who decided to include irrlicht.h in porting.h btw is: sapier in 2014 10:04 nrz the good old sapier 😄 10:12 rubenwardy sounds about right 10:13 rubenwardy who did code review, or was subsystems still in effect then 10:14 celeron55 those were some wild times 10:14 rubenwardy looks like it was just added so that porting could get window size and density... definitely the wrong place for that 10:15 rubenwardy #1198 10:15 ShadowBot https://github.com/minetest/minetest/issues/1198 -- Add support for DPI based HUD scaling by sapier 10:15 rubenwardy can't see any reviews 10:16 rubenwardy lmao 10:16 rubenwardy if noone complains I'm gonna merge #1198 soon, especially the hud indention fixes are a nightmare to rebase 10:16 rubenwardy https://irc.minetest.net/minetest-dev/2014-04-27#i_3670927 10:16 ShadowBot https://github.com/minetest/minetest/issues/1198 -- Add support for DPI based HUD scaling by sapier 10:20 nrz window size and density must be in irrlicht directly no ? 10:20 nrz it's not MT engine responsibility to check give porting for that 10:21 rubenwardy yeah, it's not something for porting.h because it's literally just reading from the irrlicht object, it's not using platform specific APIs 17:12 Sokomine what's the reason for minetest.get_content_id getting broken? why not return nil if the node doesn't exist? always checking for existence of the node now means that i have to change quite a bit of my code and write a wrapper almost everywhere 17:13 sfan5 what exactly was broken and when? 17:15 Sokomine some of my mods break because minetest.get_content_id( unregistered_node_name ) now leads to a crash instead of returning nil (-1 or 0 would be fine by me as well) 17:16 Sokomine got told that my moresnow mod doesn't work in more recent relases. and that's right. i tend to use get_content_id a lot. mg_villages/handle_schematics is probably safe because i have wrappers there already iirc 17:20 sfan5 that was changed 4 years ago (apparently without regard for compatibility issues) https://github.com/minetest/minetest/pull/9458 17:24 ROllerozxa it makes sense for get_content_id to error out if you try to retrieve the content ID of an unregistered node, rather than simply silently failing and e.g. returning -1, because it may mean you forgot to depend on the mod that registers the nodes 17:25 ROllerozxa you should always check for the presence of mods with minetest.get_modpath before you try to retrieve the content ID of a node in another mod if you're adding it as an optional dependency, and degrading gracefully when the mod does not exist 17:27 Sokomine completely diffrent use case then. my mods need to interact with a lot of diffrent mods and i can't be certain what is installed. bad change for me as i have to implement a wrapper function 17:28 Sokomine before, it was possible to get the content id - and if it's nil then the node obviously doesn't exist 17:30 Sokomine i've been working lately mostly on the npc (yl_speak_up mod) so havn't done much with my other mods 17:33 Sokomine thanks for locating the issue, sfan5! 18:24 rubenwardy Time to add get_content_id_or_nil then I guess? 18:38 sfan5 @appguru do you have athoughts on rolling release for MTG yet? 18:38 sfan5 also congrats on becoming a coredev (apparently) 18:40 Sokomine that'd be nice. i wrote such a wrapper myself 18:51 MTDiscord sfan5: Thanks. I don't see any downsides to rolling releases for MTG. Should I set up CDB automation? 18:53 sfan5 we need to make sure every pushed commit is production-ready but sounds doable 18:53 sfan5 question is should mtg still have version numbers? 18:54 MTDiscord I don't really see a use for them. Commit hashes should suffice. 18:55 sfan5 👍 18:55 MTDiscord Our pushed commits should already be production-ready, and if they aren't, our non-rolling-release hasn't really helped us with that; it has just delayed the release of buggy code as well as the associated fixes. 18:55 ROllerozxa luatic: do you have access to MTG on CDB? 18:55 MTDiscord I do 18:55 ROllerozxa ah yeah you're listed as co-maintainer there 18:55 sfan5 well sometimes it happens that a regression happens but you make a note to fix it later (usually before release) 18:56 sfan5 but if every commit is a release that doesn't work 18:57 MTDiscord I've set up automatic releases for CDB now 18:59 MTDiscord merging #14227 in 5m 18:59 ShadowBot https://github.com/minetest/minetest/issues/14227 -- Code style: Use non-static member initialisation instead of member initialiser lists for shader uniform names by HybridDog 19:05 MTDiscord btw sfan5, I've fixed the wrong indent in #13477. anything else? 19:05 ShadowBot https://github.com/minetest/minetest/issues/13477 -- Reduce `minetest.after` time complexity and provide ordering guarantee by appgurueu 20:02 Sokomine in general, mt seems to be very stable - even with latest git. on the other hand less involved users go by version number and update way less often