Time Nick Message 03:29 Zeno` Will merge #2642 shortly 03:29 ShadowBot https://github.com/minetest/minetest/issues/2642 -- Fix crash on startup (Windows) by SmallJoker 04:23 paramat hmmmm does #2640 look okay to you? 04:23 ShadowBot https://github.com/minetest/minetest/issues/2640 -- DecoSchematic: Fix missing trees in rough terrain by paramat 05:15 paramat hi nore, would you like me to push game#310 as it has enough approval, is rebased and squashed? 05:15 ShadowBot https://github.com/minetest/minetest_game/issues/310 -- Allow overriding papyrus and cactus grow functions by HybridDog 05:27 paramat also, i opened an issue game#496 perhaps sometime let me know what you think 05:27 ShadowBot https://github.com/minetest/minetest_game/issues/496 -- Make flowers farmable, with seeds, instead of the slow uniform-density spread. 05:36 hmmmm paramat: reviewing 05:36 paramat thanks 05:38 Zeno` ugh, I keep pressing ctrl-shift-v to paste lol 05:38 sfan5 [02:52] [git] ShadowNinja -> minetest/minetest: Only lowercase project name at compile time 3be9787 http://git.io/vf3hN (2015-04-21T20:51:16-04:00) 05:38 sfan5 umm 05:39 sfan5 iirc nobody did say that the commit was ok 05:39 sfan5 (02:52 is 5 hours ago) 05:52 Zeno` https://github.com/paramat/minetest/blob/decoschematic/src/mg_decoration.cpp#L250 should be static const 05:52 Zeno` I dunno if gcc would optimise it to static const but I doubt it 05:53 Zeno` oops, that's not part of the PR (sorry, was looking at the whole file :/) 05:53 paramat heh 06:46 celeron55 i'm going to remove nerzhul from the core team on github if i see a second revert like this due to this reason 06:47 celeron55 does he really commit these things without talking about them anywhere at all? 06:47 celeron55 that's super weird 06:47 hmmmm we have reason to believe he's here in this channel under a different name but passively watching 06:48 celeron55 that's even weirder 06:51 hmmmm paramat: the vm->m_area.contains(p) is now redundant, but it shouldn't make a difference 06:52 hmmmm i should probably add british spelling aliases for decoration flags =] 09:19 nrzkt why reverting: https://github.com/minetest/minetest/commit/e0eec201a18a0741114094b600f765313a838bfb ? 09:19 nrzkt this fix is perfectly correct 09:20 nrzkt est31, nore, an idea ? 09:20 nore "The commit being reverted was unauthorized. There had been no discussion, review, or sign-off prior to submittal to upstream." perhaps? 09:21 nrzkt i read that, but the fix is trivial 09:21 nrzkt and not introduce any bug, unlike hmmmm/kwoelkr commits... 09:22 nrzkt this container doesn't need to be a list, it's a queue usage, using a list decrease the performance 09:22 est31 nrzkt, I think hmmm wants you to join irc, say "pushing this and that in 15 minutes", and then wait that time and then only then push 09:23 est31 if its network related you can push without doing that, because you are network maintainer 09:23 nrzkt sorry for that, it was a trivial fix i do yesterday by looking at the code. BUt the fix is correct, reverting a trivial non bug introducing bug for that is stupid 09:23 nrzkt it's semi related to network because this queue is read by network for sending active object messages... and it's only used for that 09:24 nrzkt and i only act on it, other code parts are not mine :) 09:24 est31 I think you should discuss this with hmmm :) 09:25 nrzkt please look at the commit and approve it :) it's trivial. We have a message queue, but the queue is a std::list... WTF 09:25 nrzkt std::list is slower than std::queue 09:25 nrzkt i only change the container name and related calls 09:25 est31 Why did you append "_queue" to the name? 09:25 nrzkt can you look at https://github.com/minetest/minetest/commit/972d17baea81ffe6d508b291ef97207a4822e1da and approve 09:25 nrzkt to understand directly it's a queue ? 09:26 nrzkt i like append this to container, but if you don't want the suffix i can remove it 09:26 est31 I think its not needed 09:27 est31 Otherwise: yes approved. 09:27 nrzkt okay, then i remove this suffix and repush... 09:35 nrzkt pushing now 09:36 est31 perhaps use another authorß 09:36 est31 ? 09:45 nrzkt oh... it's not correctly configured on this pc 09:47 nrzkt done 12:13 * est pushes http://pastie.org/10107281 in 15 minutes 12:13 est (or more) 12:43 sfan5 * est pushes http://pastie.org/10107281 in 15 minutes 12:43 sfan5 what about backward-compatibility? 12:43 sfan5 oh 12:43 sfan5 this sucks 12:44 sfan5 est is already gone 12:45 sfan5 wait what 12:45 sfan5 where did the rules about the small/medium/big patches go 12:46 sfan5 celeron55: where did the "Rule 1 in practice" section go? 12:58 est31 sfan5, this is protocol v25, which isnt used yet 12:59 est31 even by git versions 12:59 sfan5 oh, i see 12:59 sfan5 go ahead then 13:00 sfan5 but next time don't leave instantly after saying "I'll push this" 13:00 est31 sry had to go 13:00 kilbith est31, pastie in plain text is barely readable, you should use https://gist.github.com/ instead 13:01 est31 next time I say "I am going, but I won't push before I came back and have read the history on irc.minetest.ru" 13:02 est31 kilbith, I didn't use gist before because I thought you need a github account for it 13:03 est31 and I don't want my github account to clutter up 13:03 est31 but I see that you can create anon gists too 13:03 kilbith select the C++ format recognition then 13:04 est31 diff is good enough too I think 13:05 est31 this is cpp https://gist.github.com/anonymous/238d7c83d024ceb851fc 13:05 est31 this is diff https://gist.github.com/anonymous/ead6fcad63f34c72c128 13:05 est31 I like diff more 13:07 sfan5 and I don't want my github account to clutter up 13:07 sfan5 github gists don't go to any place where they're in the way 13:15 celeron55 sfan5: i might have deleted it fully or partly when i was organizing the rules in the wiki 13:15 celeron55 i don't remember what exactly i did to it 13:16 Zeno` https://github.com/minetest/minetest/commit/ccc09abc2d4a712c17de833d52cb5aa91b7bf980#diff-72ff16d0b2fcbffdab8b8bde3ca82590L363 13:16 Zeno` est31, kinda surprised that even compiled 13:17 sfan5 i didn't look in all possible places but i think the "you may push smaller patches after announcing that you'll merge them" rule isn't mentioned anywhere now 13:17 Zeno` (you could have added the spaces around the assignment operator when fixing it :P) 13:17 celeron55 sfan5: feel free to add it to some appropriate place 13:17 Zeno` oh oh... is this because I merged krock's simple fix? 13:18 sfan5 i guess I'll add the first part that talks about small patches of the "rule 1 in practice" section to the git guidelines page again 13:22 sfan5 done 15:38 Zeno` is irrlicht still maintained? 15:39 VanessaE isn't it? 15:39 hmmmm I hope so *shudders* 15:39 hmmmm I don't think nzrkt's STL container changes are that valuable, personally. I don't think we should change that doesn't need to be changed for microoptimization purposes 15:40 hmmmm especially when it introduces a chance of buggy side effects 15:41 hmmmm again, std::list -> std::queue is not a perfect translation. queue by default uses deque which invalidates iterators when a new element is pushed back or one is deleted. 15:41 hmmmm unless you check every interaction of this container first, do not change it 15:44 Zeno` I have a million or so performance profiles here 15:44 Zeno` and based on them... that isn't even a microoptimsation heh 15:44 Zeno` it's not even on the chart 15:46 Zeno` queue algorithms are rarely faster than list (unless the person writing the list algo is an idiot, which I doubt the STL devs are) 15:47 Zeno` i.e. any sane list implementation will keep track of head and tail even for singly linked lists 15:50 Zeno` anyway 15:50 Zeno` the irrlicht sourceforge page does show some activity