Time  Nick       Message
01:00 rubenwardy > Irrlicht element has function setBackgroundColor
01:00 rubenwardy > does not set background
01:03 paramat    > surprising
01:04 paramat    i might merge #8376 later as trivial
01:04 ShadowBot  https://github.com/minetest/minetest/issues/8376 -- Lua_api.txt: Document blockpos, coordinate conversion, map terminology. Fix and improve LVM 'calc_lighting()' warning by paramat
05:02 paramat    merging #8376
05:02 ShadowBot  https://github.com/minetest/minetest/issues/8376 -- Lua_api.txt: Document blockpos, coordinate conversion, map terminology. Fix and improve LVM 'calc_lighting()' warning by paramat
06:56 paramat    nerz, did you announce releasing the 5.0.0 android app on -hub by mistake? http://irc.minetest.net/minetest-hub/2019-03-15#i_5512001 5.0.1 is likely to be this weekend so we could have waited. 5.0.0 has bugs so it would have been better to wait for 5.0.1, that seemed to be the plan
06:59 paramat    was there any discussion or dev agreement to release the android app yesterday? may be i missed it
06:59 paramat    *maybe
07:01 paramat    please can the app page be updated with the details of the big improvements to the app?: new controls, menu scaling etc. this really needs mentioning
07:10 paramat    for #8358 numberzero hasn't responded and this is just a visual texture glitch, so i don't agree to delaying 5.0.1 due to this. could always release 5.0.2 later. so i hope we can release this weekend
07:10 ShadowBot  https://github.com/minetest/minetest/issues/8358 -- Nodebox facedir no longer rotates textures to follow it.
08:00 nerzhul    paramat, i looked at the community returns (both server & clients), and many are complaining they either have a client drop or cannot connect to their favourite servers. Also i looked at the beta crash reports and there is no crash reported it's why i released, because remember you have 2 days delays on the play store on updates sometimes, and i want to unlock the situation, i don't see where the 5.0.1 was announced this w
08:01 nerzhul    it's to prevent a usage drop on android, 3 weeks without update is very long, i though 5.0.1 will be released faster than that
08:03 nerzhul    for 5.0.1 i will try to publish in the beta canal on the release day and production 3 days after
10:43 Krock      p_gimeno: Thanks for the feedback. Didn't notice that the axis weren't correct  ahah
11:01 p_gimeno   np, thanks for the tool, it's proven useful :)
11:11 p_gimeno   Krock: also don't miss this, it seems that shadowbot can't send two messages: http://irc.minetest.net/minetest/2019-03-16#i_5512763
11:24 Krock      p_gimeno: MTB told me that in #-hub
11:26 p_gimeno   ah
11:27 p_gimeno   right, MTB, not SB, sorry
19:34 rubenwardy https://i.rubenwardy.com/3AtI6.jpg
19:35 Krock      expected a real twist instead of abort
20:19 sfan5      5.0.1 tomorrow?
20:19 rubenwardy sgtm
20:21 rubenwardy so, for #8383 a person on the Irrlicht forums recommended only making our own version of GUISkin, then using the pointers to elements to store the custom styles. This would mean that we wouldn't have to copy the implementations of most elements to theme them, and would probably be more stable, but would result in a weird mapping of pointers to style data in the skin, which is probably hacky
20:21 ShadowBot  https://github.com/minetest/minetest/issues/8383 -- Add formspec styling using a style[] tag by rubenwardy
20:21 rubenwardy so, I should probably explain a bit better
20:21 rubenwardy inside of   GUIButton, it calls this      skin->draw3DPaneButtonPressed(this, rect);
20:22 rubenwardy in that PR above, we have to completely copy the GUIButton implementation because subclassing is not possible, and make it call   skin->drawColored3DPanePressed(this, colors, rect); instead
20:22 rubenwardy so the only use of our copied version is to add the setColor() function and call the skin method instead
20:23 rubenwardy so the idea is to not copy the button implementation at all
20:23 rubenwardy but instead make the skin class aware of the per-element theming
20:23 rubenwardy so the skin class would do a lookup on the element passed to it as the first parameter
20:23 rubenwardy this may not work for all types of elements, but could remove a lot of them
20:25 rubenwardy here's a code snippet: https://github.com/minetest/minetest/pull/8383#issuecomment-473487655
20:25 rubenwardy there's also a point to whether it's really worth doing this
20:27 rubenwardy I'm hoping that most things will have a method for it already, and most do
20:27 rubenwardy for example, it's already possible to image GUIButtons, and that's what we use for image_button
21:49 rubenwardy ok, from the sounds of it Irrlicht 1.9 won't happen for a long time
21:50 rubenwardy and CuteAlien, the lead and only dev of Irrlicht, recommend doing it by the way that's already in the PR
22:29 rubenwardy #8383 is now ready for review
22:29 ShadowBot  https://github.com/minetest/minetest/issues/8383 -- Add formspec styling using a style[] tag by rubenwardy
22:29 rubenwardy (although I'm still testing it)