Time Nick Message 00:07 naxeu_ Hello. I'm trying to compile mintest from source. But I get a error on a c++ file 00:07 naxeu_ command: cmake . -DIRRLICHT_INCLUDE_DIR=$PWD/../irrlicht/include -DIRRLICHT_LIBRARY=$PWD/../irrlicht/lib/Linux/libIrrlicht.a 00:10 naxeu_ stderr log: https://bin.veracry.pt/?8f8cd9474c7e240b#FnjPHj73xXMYWjgHdBtSREaqxQMLH9J7cDaAi9HTrbLV 00:11 MTDiscord try "${PWD}/../irrlicht/" 00:11 MTDiscord try enclosing them in brackets, had similar issues with unusual paths 00:12 MTDiscord wait no 00:12 naxeu_ But the problem isn't on cmake. It is no make 00:12 naxeu_ *no -> on 00:12 MTDiscord $PWD would be invalid as there's no variable available in bash 00:14 MTDiscord you can try the following 00:14 MTDiscord cmake . -DIRRLICHT_INCLUDE_DIR="../irrlicht/include" -DIRRLICHT_LIBRARY="../irrlicht/lib/Linux/libIrrlicht.a" 00:16 naxeu_ I've run from the real path since the root "/" 00:16 naxeu_ No changes 00:16 naxeu_ The error is when I run make 00:17 naxeu_ On client/gameui.cpp, 00:17 naxeu_ irrlicht_changes/static_text.h 00:18 naxeu_ The full version is on this link https://vcr.pt/BZOW8 00:19 naxeu_ @dev_team 00:24 MTDiscord are you using the minetest forked irrlicht or the original irrlicht 00:25 naxeu_ Forked 00:26 MTDiscord hmm 00:26 MTDiscord stable-5 version of MT or master branch 00:27 MTDiscord because it looks like you used the stable branch which currently is incompatible with the MT fork of irrlicht 00:28 MTDiscord @naxeu_#0000 link against libirrlichtMt.a 00:28 MTDiscord naxeu_ ^ 00:29 MTDiscord if the library was missing he'd get a linking error, not a compiling error 00:29 MTDiscord He did enter the wrong library name. 00:29 MTDiscord Linking against the non MT branded library Segfaulted for me, so still a good precaution 00:33 naxeu_ Do I have to put --depth=1 when cloning minetest 00:33 naxeu_ ? 00:34 MTDiscord No, but it can speed go cloning if you aren't doing development (it only shallow clones the current state of the repo iirc) 00:34 naxeu_ Benrob0329: I've also put libirrlichtMt.a as library 00:35 naxeu_ I Get the same error 00:35 MTDiscord a print out of git log would be nice 00:36 naxeu_ I don't have the git log of clonning this repos 00:36 MTDiscord Try a relative path rather than a variable 00:36 naxeu_ I've used the real path. Same problem 00:36 MTDiscord git log is a command. It shows the commit history. 00:37 MTDiscord What version of MT are you trying to compile? 00:37 naxeu_ Version: HEAD detached at 5.4.1 00:38 MTDiscord Don't use IrrlichtMT 00:38 MTDiscord Only 5.5+ use it 00:39 naxeu_ Hmm. Thank you 00:39 naxeu_ There is any stable branch with 5.5? 00:39 MTDiscord not yet, it's under development 00:40 MTDiscord Testing is appreciated though 00:43 naxeu_ I still getting the same error 00:44 naxeu_ I've installed irrlicht form my system 00:44 naxeu_ Run cmake . on minetest/ 00:44 naxeu_ Then make -j$(nproc) 00:45 naxeu_ Log at https://vcr.pt/MEwuO 00:48 naxeu_ My c++ compiler is: 00:48 naxeu_ g++ (Debian 8.3.0-6) 8.3.0 00:48 naxeu_ Any ideas? 00:59 MTDiscord cmake might be holding onto your previous config 01:00 MTDiscord you'll probably want to clean, I think there's a cmake clean command 01:17 naxeu_ Solved! 01:17 naxeu_ There is not a cmake clean command 01:17 naxeu_ So I removed the folder of minetest and irrlichtMT 01:18 naxeu_ And cloned again with deepth=1 ( faster :) ) 01:19 naxeu_ So after that it compiled Minetest 5.5.0-dev-4d0fef8 01:21 naxeu_ I'm using Irrlicht 1.9.0mt2 bundled inside mt I think 01:22 naxeu_ I can also run mt without system's libirrlicht. Awesome :) 01:22 naxeu_ Thank you Benrob0329 17:50 MTDiscord #11184 is trivial and has 2 approvals, anything in particular that's keeping it from getting merged? 17:50 ShadowBot https://github.com/minetest/minetest/issues/11184 -- Also return the ObjectRef from minetest.spawn_falling_node() by benrob0329 17:52 Krock Benrob0329: no need to be nervous 17:52 Krock will merge #11184 and #11181 in 10minutes 17:52 ShadowBot https://github.com/minetest/minetest/issues/11184 -- Also return the ObjectRef from minetest.spawn_falling_node() by benrob0329 17:52 ShadowBot https://github.com/minetest/minetest/issues/11181 -- Attachments: Avoid data loss caused by set_attach() in callbacks by SmallJoker 17:52 MTDiscord I'm not, just trying to be polite while also bumping my own PR so that it's off my TODO :-) 17:56 MTDiscord I'd also like to get IrrlichtMt # 22 off my TODO. :-)) 18:01 Krock merging 18:03 Krock I'd also prefer it we could simply replace `true` with `obj` for simplicity. adding more return values is the most precise but clobbering solution 18:07 MTDiscord Like I said on the PR, we could probably get away with it but it's technically a breaking change 18:07 rubenwardy I don't think it'll cause issues, it's just a matter of how pedantic you want to be 18:10 rubenwardy Given its a minor issue, probably easier to side step and do true, obj 20:34 rubenwardy A useful feature for the serverlist would be to have a way of getting debug information 20:34 rubenwardy so a user would search for their server, and it would show why it was rejected 20:38 nerzhul rubenwardy it's not create more work it's like sfan5 said, reorg, and i'd like to see MT main codebase to be only the engine, not graphics/sound engine :D 20:39 rubenwardy it's moving dependencies around 20:39 nerzhul let mtirrlicht become the technical engine and the feature engine in mt :) 20:39 rubenwardy there's no user benefit and I question the architectural benefit 20:39 nerzhul yep, but also we can just abstract. our sound system requires some refactor, i know it works but currently it has various abstractions 20:39 nerzhul it's purely architectural question 20:39 nerzhul nowadays i'm more on architectural and design fixes 20:40 nerzhul i should repass some linters and analysers on code to find where we can improve the core engine too a bit, but really i'd like to make the cleint evolve. I'm not confident with majority of the code. Irrlicht graphics call are too tied to each handler