Time Nick Message 09:32 paramat nore sfan5 please could you review game#1076 game#1106 ? proposed mese texture updated, screenshots added 09:32 ShadowBot https://github.com/minetest/minetest_game/issues/1076 -- Fire: Ignite tnt, gunpowder, permanent flame above coalblock by paramat 09:32 ShadowBot https://github.com/minetest/minetest_game/issues/1106 -- Default: New mese texture. Add missing texture credits by paramat 10:17 MarcoPoloWhite I don't understand minetest_game 10:17 MarcoPoloWhite Who's allowed to approve and who's not? 10:19 paramat me, sfan, n-ore, shadowninj, pilz-adam 10:24 sfan5 'normal lettering.' 10:25 sfan5 but it doesn't say mese 10:30 paramat ? it's clearly MESE 10:33 sfan5 not really 10:36 paramat imagine light coming from above, letters are defined by highlight and shadow 10:36 paramat as in celeron's official texture 10:37 celeron55 uhm 10:37 celeron55 oh that version 10:38 paramat sorry if i highlighted you 11:17 paramat letters are raised areas of a pure material, so are defined by highlights and shadows only 13:35 VanessaE paramat: don't like my texture anymore? :( 13:39 VanessaE as for 1076, I know I'm jumping into that one late, I would have preferred the idea of placing a torch on top of the coal block and having it turn into flames, considering that a torch is already made from coal anyway. 13:43 VanessaE (consider the effort it takes to light barbecue charcoal) 14:19 paramat Vanessa 1076 started with lighting coalblock with a torch but i was convinced to use flint and steel instead, see discussion 14:20 MarcoPoloWhite I still think the mese should have a coded message in the texture 14:20 MarcoPoloWhite binary perhaps 14:20 MarcoPoloWhite dark pixels as 0 and light as 1 14:21 VanessaE paramat: I read the discussion 14:21 VanessaE hence my comment here (as opposed to adding to the thread) 14:21 VanessaE consider my comment as a +1 for the idea of torch-on-top as one of the other commenters mentioned. 14:29 paramat ok 14:55 paramat #4163 is updated and now ready 14:55 ShadowBot https://github.com/minetest/minetest/issues/4163 -- Sky: Darker, bluer sky and improved horizon haze at night by paramat 21:06 Fixer https://github.com/minetest/minetest/issues/4067 21:08 Fixer https://github.com/minetest/minetest/issues/4151 epic crap bug 21:51 celeron55 that bug is ridiculous; absolutely nothing that people are finding makes any sense 21:56 celeron55 i love this comment though: 21:56 celeron55 "That stack trace turned out to be very useful. The emerge thread appears to be stuck on a mutex and I spent quite a while trying to figure out how that could possibly happen. I finally spotted it and it's not something that's there, it's something that isn't there. The ServerThread is missing! It could possibly be thread 4, but something has happened to it, most likely while it was holding the mutex. I will see if I can figure out ... 21:56 celeron55 ... what could have happened." 21:56 celeron55 it's like from a horror movie for programmers 21:58 celeron55 if you know how minetest works internally, ServerThread disappearing while holding a mutex is horrible and doesn't make sense, altough it does explain everything that happens afterwards 21:58 celeron55 (that's from https://github.com/minetest/minetest/issues/3307) 22:05 celeron55 or, more like "would explain"; somebody first has to come up with a reason how it could do that to begin with 22:06 celeron55 it pretty much has to be a condition that silently kills a thread without stopping the rest of the program 22:13 celeron55 no signal should do that, but maybe due to some reason some signal could in this case? i have no idea 22:22 kahrl does the linux OOM killer kill whole processes or individual threads? 22:27 kahrl well, the docs say processes, but traditionally threads have been implemented as processes (from the kernel point of view) so I'm still not sure 22:37 Fixer celeron55: thats why bug is so crappy, and kills mp gameplay, and you wonder why people think mt is bug-fest %) 22:45 hmmmm there is no way oom killer would kill an individual thread 22:46 hmmmm at that point might as well execute code in a high-memory-use process and forcibly deallocate heap memory 22:53 Fixer celeron55: what is strange about 3307 is... there were no stalls before, he updated to 0.4.14, stalls appeared, he reverted back to working build - still stalling :/ Something happened to database? And this problem was reported 1.5 years ago, it is not exactly related to 0.4.14 22:53 Fixer just coinsidence 22:53 Fixer ESM was having this stall from very beginning on Aug 15 builds 22:54 Fixer OldCoder even earlier 23:43 OldCoder Fixer, actually, 2 years ago... but is it RESEND RELIABLE? 23:44 OldCoder There are multiple network issues 23:44 OldCoder The patch addresses one regrettable pile-up 23:44 Fixer OldCoder: both xanadu and esm have this resend reliable in logs 23:44 OldCoder The patch should correct this; other "stalls" may remain 23:48 Fixer this is worst part :/ 23:54 OldCoder ? 23:54 OldCoder Fixer, shall we prepare a PR? 23:55 Fixer OldCoder: you can, let other devs review it 23:56 OldCoder PM