Time Nick Message 10:52 Zeno` comments on https://github.com/Zeno-/minetest/commit/bb0e72b44386cb2d2199d0d1d18c4cfbfb1af2ff please? 10:53 Zeno` note that the glyph size should be checked anyway and also that in later versions of FreeType .width and .rows are unsigned rather than int 10:55 Zeno` This all relates to: https://github.com/minetest/minetest/commit/0904884ad47b5df67a5a26360d3bb35de1e76e5c at the end of the day 10:58 Zeno` made it #2304 10:58 ShadowBot https://github.com/minetest/minetest/issues/2304 -- Suppress CGUITTFONT build warnings by Zeno- 13:21 Zeno` merging #2304 shortly 13:21 ShadowBot https://github.com/minetest/minetest/issues/2304 -- Suppress CGUITTFONT build warnings by Zeno- 13:23 Zeno` geez, there are a lot of warnings in the mingw build 14:58 Zeno` hmmmm 14:58 shadowzone Hello 21:12 est31 anyone to give +1 on #2225 ? 21:12 ShadowBot https://github.com/minetest/minetest/issues/2225 -- Fix for getCraftRecipes implementation is very slow by gregorycu 21:32 est31 wasn't networkprotocol called else in the past? 21:33 est31 but better name 21:40 VanessaE yeah, media_FUBAR ;) 21:40 VanessaE (before it got fixed, that was the joke name a few of us used to differentiate it from http/cURL) 21:41 est31 wasnt it called clientserver? 21:44 Tesseract est31: Yes, clientserver.h 21:45 est31 ok then i remember right 21:45 Tesseract protocol.txt probably has to be updated because of the change. 21:47 est31 Tesseract, is #2301 better now? 21:47 ShadowBot https://github.com/minetest/minetest/issues/2301 -- Fix crash on passing false as value in table to table.copy(t) by est31 21:50 Tesseract est31: Looks good -- although one of the sets of parenthesis 21:52 est31 ok, should do k → v in some cases too 21:53 est31 what do you mean with that? 21:55 est31 better? 22:01 est31 Tesseract, better? do you think this has too much parenthesis now? at the wrong place? 22:16 est31 Tesseract, if its good, please +1 it :) 22:40 est31 so, the player:name inventory location got added without support for it? 22:41 est31 see #minetest discussion 22:41 est31 I just dont see any way this got sent to the client in the past 22:41 est31 there is TOCLIENT_INVENTORY, but that doesnt contain a name 22:42 celeron55 it might have never been implemented 22:43 PilzAdam I remember that I encountered that problem when creating nodetopia 22:43 PilzAdam player inventories are just send to the player who owns it 22:44 est31 and when you try sth like this (thefamilygrog), it fails: http://pastie.org/9945419 22:44 est31 Inside client.cpp it executes a simple m_env.getPlayer 22:44 est31 and then it returns an empty inventory 22:47 PilzAdam nodetopias usecase: https://github.com/PilzAdam/nodetopia/blob/master/mods/backpack/init.lua#L125 22:56 est31 there are different message types for different kinds of TOCLIENT_ inventory messages in network protocol. one for player inventory, one for detached one, dunno how node inventories are done 22:57 gregorycu How D 22:58 est31 guess the same way other metadata is sent 22:58 est31 dunno how that works either 23:00 est31 guess the whole node is update 23:00 est31 d 23:01 est31 shouldn't those messages be unified? at least for detached and player? 23:05 celeron55 the way the inventories are transferred to the client is different in all cases 23:05 celeron55 for node inventories, they are sent with the nodes 23:06 celeron55 for the player, it's sent in a special way (because the client obviously always has to see the inventory of the player it's playing as) 23:07 celeron55 and... detached inventories are sent to everyone (which is to be optimized so that they are sent only to the clients that need the particular ones) 23:07 celeron55 and are stored differently because they don't belong to any object in the game 23:09 celeron55 really a more unified way of handling inventories would first require a more unified structure of the game to begin with, which may or may not make sense but definitely does not exist now 23:09 est31 you could parse the name 23:09 celeron55 it might also imply a different save format and whatever 23:09 celeron55 not a simple matter 23:10 est31 unifying detached and player inventories seems easy 23:11 celeron55 at what level? 23:11 celeron55 you can't store them similarly at least 23:11 est31 at the network level only 23:12 est31 yes, dont propose that 23:12 celeron55 that could be true 23:12 est31 because currently, there is no way to send another player's inventories 23:14 est31 A generic TOCLIENT_NAMED_INVENTORY would solve this. 23:18 est31 Can I start work on that? 23:21 est31 other option would be to have TOCLIENT_OTHER_INVENTORY 23:21 est31 or add a player name parameter to TOCLIENT_INVENTORY 23:21 est31 or to document that the player:name inventory doesn't work for formspecs 23:32 est31 so what is the best way? other ppl's idea? 23:32 est31 ideas* 23:37 gregorycu Not sure what the problem is 23:38 est31 because currently, there is no way to send another player's inventories 23:38 gregorycu Thank you for continuing to push #2225 23:38 ShadowBot https://github.com/minetest/minetest/issues/2225 -- Fix for getCraftRecipes implementation is very slow by gregorycu 23:39 est31 its needed when you want to display the player's formspec 23:39 gregorycu I see 23:40 est31 and there is even a nice lookup in client.cpp, but that doesn't work when the player entry wasn't populated 23:40 est31 so what we should do is populate it 23:42 est31 how far are you with the crafting api consistency improvement? 23:44 gregorycu It's stashed, I'm working on frame rate improvements for shaders at the moment 23:44 est31 ok 23:44 gregorycu I have finished the implementation of remove craft though 23:44 gregorycu And it appears to work, just want to test it some more