Time Nick Message 00:47 MTDiscord can't the first thing the client downloads be a loading screen? 00:49 MTDiscord Yeah but the client loads basically instantly 00:51 MTDiscord Haha, if only 00:52 MTDiscord Oh, you meant on the loading screen, that would be very neat 01:14 MTDiscord wow, I should really read the full conversation before commenting with something that has already been said 15:03 MTDiscord hm. I just updated minetest/irrlicht revisions to master, and I can't start mineclonia anymore. (i did not do any game updates) 15:05 MTDiscord mineclonia\mods\ITEMS\mcl_chests\init.lua:468: bad argument #2 to 'set_string' (string expected, got nil) 15:05 ShadowBot https://github.com/minetest/minetest/issues/2 -- Burned wood 15:06 MTDiscord interesting. That line is: meta:set_string("workaround", nil) -- Done to keep metadata clean 15:08 MTDiscord https://codeberg.org/mineclonia/mineclonia/src/branch/main/mods/ITEMS/mcl_chests/init.lua#L476 15:08 ROllerozxa burned wood... thank you shadowbot very helpful 15:08 ROllerozxa sounds like it's https://github.com/minetest/minetest/issues/14391 15:08 ROllerozxa (a regression) 16:52 sfan5 easy to fix, PRs welcome 17:36 MTDiscord i'm not sure if only l_set_string should accept nil specially, or readParam more generally 17:39 sfan5 are you suggesting it should generally be allowed to pass nil to mean ""? 17:40 MTDiscord it should not accept nil. passing nil to set_string should be deprecated (or an exception) imo. our docs don't state that it is allowed. 17:40 MTDiscord reading a string in general should actually expect a string. stuff like minetest.get_player_by_name(nil) is simply nonsense. 17:41 sfan5 (should return a spooky ghost player object but only if the date is oct 31) 17:45 Krock I can propose a quick patch for this in an hour if nobody else is picking it up until then 17:45 Krock the nil argument for set_string is also used in other mods 17:45 Krock VanessaE also had this issue years ago when it broke the last time 17:45 Krock possibly with pipeworks but don't quote me on that 17:49 MTDiscord I suggest adding a Devtest unittest if you fix this, there are already some metadata API tests 18:56 Krock on it ... 19:15 Krock #14396 19:15 ShadowBot https://github.com/minetest/minetest/issues/14396 -- MetaData: restore undocumented set_string behaviour by SmallJoker 22:26 sfan5 @paradust btw I think irr#276 breaks the webgl driver, how much of a problem is that for you? 22:26 ShadowBot https://github.com/minetest/irrlicht/issues/276 -- GL loader work by sfan5 22:28 MTDiscord i only merge infrequently. breaks it how? 22:29 MTDiscord i could always just keep the driver as is, assuming the rest of irrlicht remains mostly compatible 22:33 sfan5 it deletes the legacy OGLES2 driver along with CWebGL1Driver.cpp 22:34 sfan5 inside the (new) unified driver I only see a TODO comment for WebGL. don't know how much is missing, could be easy, could be hard 22:34 MTDiscord what does the unified driver do? 22:35 MTDiscord It unifies the driver 22:35 MTDiscord i mean, does it support OGLES3 or something? 22:36 sfan5 numzero (re)wrote some parts so the OpenGL3 driver and the OpenGL ES2 driver can share a common code base 22:37 sfan5 unlike the old OpenGL driver (which supports the fixed pipeline and everything) and the old OGLES2 driver (which is totally separate code) or the old OGLES1 driver (also full of legacy stuff, again separate code) 22:38 sfan5 if you don't want to go diving in the code yourself you might be able to ask @numzero 22:39 sfan5 (opengl es 3 is actually just a spec extension of open gl es 2 and it indeed supports that, yes) 22:39 MTDiscord tbh I've never even looked at the WebGL1 driver that much. I'm not even sure why its 1000 lines. In theory WebGL should just be ogles2 or ogles3 compatible 22:40 MTDiscord it just kind of worked mostly, when I used it 22:40 MTDiscord I can give whatever the new GL ES driver is a try 22:41 sfan5 you will have to fix up the code to make emscripten compile but feel free to 22:42 sfan5 effectively this is going to be merged anyway but now you know ahead of time 22:48 MTDiscord yup. thanks for letting me know