Time Nick Message 01:08 rubenwardy we need to fix clang-format and then apply it to all files, to prevent code style comments in reviews 01:14 DS-minetest why does my compiler tell me that IReferenceCounted is no base of IGUIStaticText? 01:16 rubenwardy I can confirm that it is 01:16 rubenwardy maybe clean and rebuild? 01:18 DS-minetest oops, i should've read further, the compiler told me that there was a forward declaration somewhere else 01:42 rubenwardy so, I have a disgusting hack to fix scroll containers lol 01:42 rubenwardy https://github.com/minetest/minetest/pull/9101#issuecomment-596156123 01:44 DS-minetest does scrolling with the slider work? 01:46 rubenwardy it does 01:46 rubenwardy I only tried the main scrollbar, not the sub one 01:46 rubenwardy more thoughts are needed on this I think 01:46 rubenwardy this certainly shouldn't be rushed for 5.2.0 01:47 rubenwardy I'm only fine with that hack if it is fixed later, but I find that most temporary hacks become permanent 01:47 DS-minetest have you tried making the scrollbar a child of the scroll_container? 01:48 rubenwardy I didn't 01:48 rubenwardy was that were the problem was? 01:48 DS-minetest I guess 01:49 DS-minetest your solution probably doesn't make the scrollbar not work anymore because it doesn't call OnEvent on it 01:49 rubenwardy it does? 01:52 DS-minetest I've called OnEvent on the new hovered element (which I think is normally done), you only call OnEvent on all children and the scrollbar normally isn't a child of the scroll_container 01:52 rubenwardy the scrollbar doesn't need the mouse move event if it's not inside the scroll container 01:53 rubenwardy the mouse move event means that the mouse has moved relative to the element in this case, which the scrollbar will not have 01:53 DS-minetest indeed 01:53 rubenwardy the scrollbar does receive the scrollbar in the top of that method 01:53 rubenwardy *scroll event 01:53 DS-minetest hm? 01:54 rubenwardy see if (!m_scrollbar->OnEvent(event)) { return false; } 01:55 DS-minetest oh, right that's the scroll event 01:55 DS-minetest so that when the scroll_conatainer gets a scroll event, the scrollbar gets it 01:55 DS-minetest but thats not mouse movement 01:56 DS-minetest I've just tested making the scrollbar a child of the scroll_container 01:56 rubenwardy the scrollbar shouldn't get the mouse move event though if it's outside the container 01:56 DS-minetest that doesn't work fine anyway 01:56 rubenwardy I'm not sure what happens if it's inside 01:56 rubenwardy ok 01:56 DS-minetest so, I'll apply your hack (but a bit modified) 01:57 rubenwardy well, we do need to find out why the enter/leave doesn't feel 01:57 rubenwardy *fire 01:57 DS-minetest it does not? 01:58 rubenwardy I guess the inventory list doesn't listen to it, and it doesn't fire when going between slots 02:00 DS-minetest making each item rect an own element is probably no good idea btw., too many elements 02:01 rubenwardy it's not that many elements 02:03 rubenwardy if it makes the code more resilient, it may be a good idea. Could end up with just more complexity 02:04 DS-minetest hm 02:29 DS-minetest oh, I think now that the real reason why yours works is probably because you send the event on mouse wheel scrolling and not when the scroll container changes 12:08 Andrey01 @appguru_aka_LMD: could you please elaborate what`s mainmenu file that gets serverlist using HTTP request? I seem can not find it 12:21 Krock will push https://krock-works.uk.to/u/patches/0001-Workaround-for-get_player_information.patch in 15 minutes 12:21 Krock fix for #9352. confirmed working by paramat 12:21 ShadowBot https://github.com/minetest/minetest/issues/9352 -- minetest.get_player_information sometimes return nil (running at mod load, in singleplayer) 12:42 Krock <<< 13:30 kilbith >>> 13:30 kilbith conflict resolved 14:56 Krock will merge #9472 in 15 minutes 14:56 ShadowBot https://github.com/minetest/minetest/issues/9472 -- Fixes typo, missing stars in #9469 and #9468 by Jordach 14:59 Krock sfan5: does the Win32/64 build script still work? According to https://forum.minetest.net/viewtopic.php?p=368547#p368547 it doesn't bundle the right libraries 15:11 Krock <<< 15:17 Jordach ta 15:17 Jordach hopefully this makes people happy 15:26 dzho one can dream 15:26 dzho ;-) 16:48 DS-minetest oh god, guihypertext has serious problems with dynamic memory allocation, eg. m_tags holds dynamically allocated pointers but also a pointer to m_root_tag which is a member of the object, this is horrible 17:03 Krock which reminds me of .. #9392 17:03 ShadowBot https://github.com/minetest/minetest/issues/9392 -- guiHyperText: Fix blinky cursor on link hover by SmallJoker 17:07 Krock DS-minetest: when do the pointers get freed now? 17:07 Krock everything's handled by release() and grab() from what I can tel 17:09 sfan5 Krock: libraries not being bundled was always a bug 17:09 sfan5 however the error in that post is because i compile my libraries with a newer compiler than the buildbot uses 17:09 DS-minetest the smart pointers manage the raw pointers, the raw pointers get drop()ed when the smart pointers go out of scope 17:09 sfan5 so if you use the buildbot's lib{stdc++,gcc,...} they will be missing stuff 17:09 Krock sfan5: so that means they mixed the library dll verions 17:10 DS-minetest those irr_ptr<>s work like shared pointers 17:10 sfan5 Krock: this is not the users fault 17:10 DS-minetest see also https://en.wikipedia.org/wiki/Smart_pointer and https://en.wikipedia.org/wiki/Resource_acquisition_is_initialization 17:10 sfan5 anyway this doesn't impact the official windows builds 17:11 sfan5 but I can look at solving it by giving travis a newer mingw version 17:11 Krock DS-minetest: will check. thanks. 17:12 rubenwardy +1000 for smart pointers 17:12 Krock would be nice to have working artifacts from GitLab, but gladly this is not a blocker 17:13 Krock rubenwardy: basically what Irrlicht should already offer in their code 17:13 sfan5 the gitlab artifacts currently shouldn't be used for releases currently anyway 17:13 rubenwardy yeah, but we have to work around irrlicht 17:13 Krock sfan5: not for releases but for testing 17:13 sfan5 ...since my release builds are built with gc64 luajit and users expect that 17:13 sfan5 sure just saying 17:13 rubenwardy in my personal projects, I use C++17 with all the nice memory management features and more 17:13 rubenwardy releases should be built using CI 17:13 rubenwardy that's continuous delivery 17:41 kilbith why don't nerzhul manifests on IRC before merging 17:47 Krock there's no announcement rule for regular PRs 17:48 Krock but it's always a nice thing to do 18:12 sfan5 huh? 18:13 Krock sfan5: due to the last two commits which were merged by nerzhul 18:13 Krock s/due to/it's about/ 19:15 sfan5 hm travis still runs with ubuntu 14.04 19:15 sfan5 guess I'll fix that when I make the PR to use a newer mingw compiler 19:17 Krock 20.04 will be out soon so... 19:29 rubenwardy I recently found it easier to switch to Ubuntu 19.10 to get MinGW 9, than to install it myself using some ppa 19:31 sfan5 building mingw yourself isn't that hard if you've done it once or twice 19:31 sfan5 in fact I'm doing that right now to package it in a .7z for travis 19:51 rubenwardy it may not be hard but still takes longer 20:12 DS-minetest (exit 20:41 sfan5 there we go #488 20:41 ShadowBot https://github.com/minetest/minetest/issues/488 -- Fix EmergeManager::getBlockSeed() by ShadowNinja 20:41 sfan5 #99488 20:41 ShadowBot sfan5: Error: Delimiter not found in "HTTP Error 404: Not Found" 20:41 sfan5 agh 20:41 sfan5 #9488 20:41 ShadowBot https://github.com/minetest/minetest/issues/9488 -- Update Travis infrastructure and buildbot compiler by sfan5 20:50 Krock looks about right 21:02 sfan5 does anyone ever look at the output of clang-tidy? 21:02 sfan5 it's very slow so it might make sense to disable 21:04 sfan5 [PASS] testWrapDegrees_0_360_v3f - 132976ms 21:04 sfan5 wtf 21:39 rubenwardy clang-tidy throws errors on some things