Time Nick Message 06:38 Zeno` #2037 should probably be looked at 06:38 ShadowBot https://github.com/minetest/minetest/issues/2037 -- World mod configuring is broken 08:57 Zeno` I think gregorycu is now ignoring me in github as well 08:57 Zeno` anyway, moving on... 08:58 nrzkt go ! 08:58 VanessaE morning, Zeno` 08:58 Zeno` morning. Umm, I think you should be asleep O.o 09:01 VanessaE my sleep pattern is all screwed up 09:01 VanessaE been a little sick 09:07 VanessaE think I'll take your advice and go to bed 09:08 VanessaE night 09:21 Zeno` Just for reference (I've been reading logs). I did NOT in anyway deliberately block any of gregorycu's requests. Yes, I downgraded the priority on one of them and that's all. 09:21 Zeno` ALL 09:22 Zeno` I am not arguing with him, I never have 09:24 Zeno` He, for some unknown reason, thinks I'm "out to get him" but from the start I encouraged his PRs and even merged 99% of them. I taught him how to use git. I rebased his commits. I fixed whitespace errors. I fixed his broken repo. And then all of a sudden, for no reason that I know of, he went crazy, started attacking me, putting me on ignore etc etc. 09:24 Zeno` I treat his PRs exactly the same as I treat the PRs of any other non-core dev 09:24 nrzkt hey Zeno`, no problem :p 09:24 Zeno` I hold no grudge, but this is getting ridiculous :/ 09:25 nrzkt gregorycu is not an easy men, and you don't need to be bothered by taht 09:27 Zeno` What I am bothered by is the wedge he seems to trying to create 09:27 Zeno` e.g. est31 keeps saying things about me when I let this thing drop ages ago (about 20 minutes after it came up) 09:29 Zeno` I'm sick of it TBH. If he wants to "only discuss his PRs on GitHub" then I'm fine with that. But if that's the case there is no reason for him to join this channel 09:29 Zeno` Dinner is served 09:29 Zeno` :) 09:29 Zeno` bbiab 09:33 Zeno` back, it's not ready 09:35 Zeno` and I am willing to ignore him. I do anyway now. But I will not rely on a crutch like /ignore and I won't ignore his questions if he directs them specifically towards me. I just do not like the stuff that seems to be being said behind my back. Over nothing. There is *nothing* I did to cause this amount of angst from him towards me. 09:35 Zeno` Back to PRs 09:35 Zeno` Is nrzkt still here? I've looked at #2321 (but not compiled or tested) and it looks good 09:36 ShadowBot https://github.com/minetest/minetest/issues/2321 -- Performance Improvement: Use a factory which caches result for getFacePositions. by nerzhul 09:36 nrzkt perfect 09:36 sfan5 >FacePositionFactory 09:36 Zeno` Why the word "factory" though? 09:36 sfan5 sounds very java-y 09:36 nrzkt because it build something :p 09:36 nrzkt but we can relabel it ^^ 09:36 sfan5 also don't use m_something in new code 09:37 nrzkt it's not java, and using a real name 09:37 sfan5 why even bother to create a class for it? 09:37 nrzkt because a class is proper than a ugly global variable ? : 09:37 Zeno` It's not really a factory it's... I dunno the term. Lazy Caching? 09:37 nrzkt i can replace Factory by Cache, yes, seems good 09:37 Zeno` sfan5, m_ is no longer discouraged in new code 09:38 Zeno` sfan5, even hmmmm admitted that ;) 09:38 nrzkt there are no rules for m_* 09:38 sfan5 it isn't? 09:38 Zeno` sfan5, no. hmmmm, who championed the removal of m_ in new code changed his mind 09:38 Zeno` he probably updated the wiki at that time as well :) 09:39 sfan5 " The latter is discouraged for public members in newer code" 09:39 sfan5 thats what the wiki says 09:39 Zeno` hm 09:39 nrzkt it's not a public member there 09:39 nrzkt sorry 09:39 Zeno` oh, public members... yes it should not be for public members 09:40 Zeno` private or protected members it's ok though 09:40 sfan5 m_ is ok then 09:42 sfan5 nrzkt: also you should do s/permit/permits/ and s/reduce/reduces/ in numeric.h 09:42 nrzkt :q 09:44 nrzkt right 09:44 nrzkt then i s/Factory/Cache/g and i fix the missing s 10:00 Zeno` yep, seems fine to me 10:02 nrzkt all is fixed, can i merge ? 10:06 nrzkt oh karhl add a nice comment 10:12 Zeno` Yeah... std::vector is generally (I believe) faster than std::list (on most CPUs) 10:17 nrzkt i review, yes here it's better because it's a linear read 10:18 nrzkt then i modify my PR, do you agree for merge Zeno`, the vector is good for you ? and the approach/names are good ? 10:25 Zeno` yes, I like this better now after kahrl's suggestion 10:43 Zeno` where is hmmmm? 10:43 sfan5 not here obviously 10:44 Zeno` haven't seen him in a few days I mean 10:46 nrzkt where is 0.4.12 ? :( 10:46 nrzkt i think we must create a song to invoke MT 0.4.12 10:50 nrzkt sfan5, you can merge #2228 it's good 10:51 ShadowBot https://github.com/minetest/minetest/issues/2228 -- Add LevelDB and redis to Travis Linux builds by sfan5 10:52 sfan5 ok 10:58 nrzkt can someone review #2310 ? 10:58 ShadowBot https://github.com/minetest/minetest/issues/2310 -- Update zlib support by SmallJoker 11:01 twoelk have been category sorting the minetest-dev wiki lately, please take a look for blunders I might have made 11:02 twoelk http://dev.minetest.net/Category:Root-Category should serve as root to access all categorys 11:03 twoelk don't know what to do with pages such as string and table listed here http://dev.minetest.net/Special:UncategorizedPages 11:04 nrzkt oh... there are ... short 11:07 twoelk|2 uhm unstable connection today it seems 11:07 * twoelk|2 is waiting for his other self to time out 11:09 twoelk there 11:15 Zeno` I don't understand 2310 11:16 Zeno` So I guess as long as it works on our other supported platforms and Krock (smalljoker) says it works on Windows, then... well, I can't object 11:16 Zeno` sfan5, have you tested it? 11:17 sfan5 Zeno`: 2228? 11:20 Zeno` #2310 11:20 ShadowBot https://github.com/minetest/minetest/issues/2310 -- Update zlib support by SmallJoker 11:21 sfan5 uh huh 11:21 sfan5 that should break things 11:22 Zeno` can you indicate that in the PR? So somebody doesn't accidentally merge it, hehe? 11:31 sfan5 Zeno`: did that 11:34 Zeno` :D 11:34 * Zeno` does happy dance 11:35 nrzkt Zeno` eats to many oysters 11:38 Zeno` :D 12:47 LazyJ Minetest 0.4.11 stable has had too many problems to trust with public server use; which branch of Minetest is recommended for server use, "master" or the "feature-freeze"? 12:48 nrzkt not master 12:48 nrzkt please use feature-freeze, master shouldn't be followed by users, except for alpha/beta tests 12:49 LazyJ Thanks, nrzkt. ;) I had seen a lot of fixes being applied to "master" which made me wonder which one was more "stable" at the time. 12:50 nrzkt master is the development version at now; before it should be used for your needs but now devel and stabilisation process have been decoupled to improve the devel process without blocking it at each release 12:51 nrzkt the FF branch will receive backports from master until the release happens, but i'm not the maintainer 12:54 LazyJ Ok, that helps clarify things a bit better. 12:57 LazyJ For my singleplayer worlds I've always used the development versions and, of course, run into problems. I understood that they were "bleeding edge". I do like the idea of a split of absolute "bleeding edge" being separated from "relatively-new-but-not-branded-stable-yet". 12:59 LazyJ I want to keep abreast of the new stuff but not so new that it's still crashing and burning as the bugs are being hammered out. 12:59 LazyJ Thanks ;_ 12:59 LazyJ *;) 12:59 nrzkt for SP is don't think this is a problem, but for MP with network this can be, and the protocol will be reworked, then a break will occurs 13:00 * LazyJ groans, "All the times when I want an underscore _ I can never hit the key." 14:49 JakubVanek is there any network protocol freeze date? 14:50 JakubVanek ^ nrzkt 14:50 nrzkt protocol freeze date ? 14:50 nrzkt what would you mean ? 14:52 JakubVanek date after which requests for protocol break will be rejected 14:53 nrzkt at this time, i suggest to use #2327 as a protocol discussion topic. When WIP tag will be removed i think we could say other requests must wait 14:53 ShadowBot https://github.com/minetest/minetest/issues/2327 -- [Patch 3/4] Network rework: Packet fixes & cleanups by nerzhul 14:54 nrzkt this PR will fix protocol design an unify string transmission (not yet finished, i need to unify longString and strings 14:55 JakubVanek ok 14:55 nrzkt if you have improvement suggestions for the protocol, you can tell me. But protocol modifications on master branch requires 2327 to be merged before, with your idea, or after merge 14:55 nrzkt after the protocol break we would have a network layer break too. This is PR 4/4 which change the UDP layer to a TCP layer, but i need to fix large packet sent/receive (like medias). Other packets are working very well 15:03 nrzkt JakubVanek, have you got ideas for protocol improvements ? 15:03 nrzkt oh i see, dps implementation :) 15:03 JakubVanek players' HP should be sent as u16, player object has u16 already. #2261 upgrades damage from u8 to s16 15:03 ShadowBot https://github.com/minetest/minetest/issues/2261 -- damage_per_second negative number implementation by JakubVanek 15:47 Zeno` will apply http://dpaste.com/2XWGDA7 in lieu of #2014 if anyone agrees 15:47 ShadowBot https://github.com/minetest/minetest/issues/2014 -- Added player direction to debug info screen by yamanq 15:48 Zeno` (will update author) 15:49 nrzkt i'm okay 15:49 Zeno` ok, doing it now 15:49 nrzkt stop 15:49 Zeno` halt! 15:49 nrzkt no in fact :p 15:49 Zeno` who goes there? 15:50 nrzkt please move this function out of game.cpp, find a better place 15:50 nrzkt game.cpp isn't MT trash for functions :p 15:50 nrzkt xD 15:50 Zeno` sigh. Let's leave this until tomorrow then ok? I'll fix wrapDegrees_0_360() at the same time 15:51 nrzkt okay then, if you move it after your merge and fix on wrapDegrees you can merge and chose 2014 PR :) 15:51 Zeno` it's not really a public function meant to be re-used so game.cpp is not tooo bad a location 15:51 nrzkt yes, but i could be used by another class ? maybe 15:51 nrzkt i think we must find new home for certain functions :) 15:52 nrzkt then ok let it in game.cpp we will look at this later 15:52 Zeno` I think we also have to look at what is worth making generic and what is not making generic :p 15:52 Zeno` not worth making* 15:52 Zeno` i.e. those strings it returns are not very generic anyway 15:53 Zeno` so, merge to get the PR closed, or not? 15:53 nrzkt yes, you are free 15:54 * VanessaE peeks in 15:55 nrzkt Please hide VanessaE is coming ! 15:56 VanessaE *rubs sleep from eyes* are we having fun yet? 15:57 VanessaE Zeno`: nice, you confused him ;) 15:58 nrzkt xD 15:59 Zeno` geez, he managed to confuse me as well 16:01 * shadowzone looks at her screen sideways 16:02 nrzkt https://github.com/minetest/minetest/pull/2129 Zeno` you agree for a close without merge ? 16:02 VanessaE there, brought my servers all up-to-date 16:03 nrzkt good ! 16:03 Zeno` nrzkt, not exactly. I think it would be good for Android but no other builds 16:04 nrzkt android already has 16:05 nrzkt -#ifdef __ANDROID__ 16:05 nrzkt - player_stepheight += (0.5 * BS); 16:05 nrzkt -#endif 16:05 nrzkt the PR permit to have this option on non android devices, and to toggle it on GUI 16:05 VanessaE ok, can someone PLEASE fix https://github.com/minetest/minetest/issues/1425 now? 16:05 nrzkt no 16:05 nrzkt because of you :p 16:06 VanessaE (and fix the root cause of the fake entities being created in the first place) 16:06 nrzkt ouch 16:06 nrzkt WTF bug 16:06 VanessaE yeah, this has been plagueing my servers for a looooong time now 16:06 VanessaE it needs fixed. 16:06 nrzkt if i understand the bug, player is an entity, and when we remove entity this remove players ? 16:07 VanessaE yep 16:07 VanessaE something like that 16:07 nrzkt ouch it's a very conceptual problem 16:07 VanessaE but there's enough difference between player objects and regular entities that it's trivial to tell them apart when deleting. idk what the actual problem is. 16:08 nrzkt if it's what i think, i'll look at a mangos-like approach 16:09 nrzkt could you find a mean which always trigger the problem ? 16:10 Zeno` I'll add an issue to make sure we don't run out of them 16:10 VanessaE I have no idea how to reliably reproduce the core issue (the fake/ghost entities), but simply producing enough of them in an area to force the engine to delete them should be enough 16:11 VanessaE I think kahrl added a /dummyball command or something like that which will do it 16:11 Zeno` isn't there a PR that fixes that behaviour, VanessaE? 16:11 VanessaE Zeno`: I don't know 16:11 Zeno` hmm 16:11 nrzkt where is the entity class... 16:11 nrzkt LuaEntitySAO ? 16:12 nrzkt there is no real entity ?? 16:12 nrzkt i think we need to have some heritage for entities and players, and also in the future, units 16:15 nrzkt ok i see. Maybe i will find your problem VanessaE, but not today 16:15 VanessaE definitely needs fixed before 0.4.12 release 16:16 nrzkt i don't know if it will be fixed, i haven't so many elements to debug 16:16 VanessaE when I can connect to my creative server and get booted off immediately because of entity clearing, that's a problem 16:16 nrzkt i understand 16:17 nrzkt if you give me a starting LUA call, i could help your 16:18 nrzkt you* 16:18 VanessaE it's not a lua call, it's something the engine does on its own 16:20 nrzkt erf 16:20 nrzkt cleaning area after a moment, maybe, and if there are entities, ... boom ? 16:20 nrzkt is there a ID conflict ? i think 16:20 VanessaE idk 16:21 nrzkt i think there is a conflict between player entities and unit entities, because player peer_id is generated in a way which doesn't look at other entities ID 16:21 VanessaE I know zip about how the ghost entities issue manifests itself nor why the engine deletes players when it tries to clear out the ghosts. 16:26 nrzkt ShadowNinja: #1606, please squash :) 16:26 ShadowBot https://github.com/minetest/minetest/issues/1606 -- Add mod security by ShadowNinja 16:35 Zeno` anyway, my alternative to #2014 (for reference). This can be compiled as C or C++ : http://dpaste.com/06GFGY2 16:35 ShadowBot https://github.com/minetest/minetest/issues/2014 -- Added player direction to debug info screen by yamanq 16:36 Zeno` if if wrapDegrees_0_360() worked as expected then it would be simply return player_direction[yaw / 90]; 16:37 nrzkt i think we must rework on SAO heritage. There are many functions used by other objects which are useless... ItemSAO has method for getInventory() or setBonePosition... 16:38 nrzkt Zeno`, agree you can modify the function, the PR has been merged and improvement was good. Have you fixed the problem on wrapDegrees ? 16:43 nrzkt Zeno`, are you OK with that ? https://github.com/nerzhul/minetest/commit/fea22e60de829b4913ec047d88b6d3143e60cf12 16:43 nrzkt or nore, maybe :p 16:44 nore way better than #defines :) 16:44 nore but perhaps you could remove the deprecated stuff 16:45 nrzkt which stuff ? :) 16:45 nore it's been deprecated since several years (that's 0.3 stuff) 16:45 nrzkt RAT ? :p 16:45 nore object types rat, oerkki1, etc 16:45 nrzkt firefly too 16:45 nore yep 16:46 nore and mobv2 I think 16:46 Calinou mobv2 is dungeon master 16:46 nore even item is deprecated 16:46 nore it is now a lua entity 16:46 nrzkt i do the cleanup. 16:46 nrzkt i don't know for item 16:47 nrzkt Calinou mobv2 could be removed too ? 16:47 Calinou probably 16:48 kilbith "mobv2 is dungeon master" so yeah 16:49 nrzkt then this needs a cleanup 16:53 nrzkt ItemSAO is called by a l_env call, i cannot remove it 16:57 nrzkt #2330 ready to be merged 16:57 ShadowBot https://github.com/minetest/minetest/issues/2330 -- SAO work: ActiveObject types & SAO cleanup by nerzhul 17:31 rubenwardy Or you could officially support http://rubenwardy.com/minetest_modding_book/lua_api.html 17:36 jin_xi imho we'd be better off with some tutorial like example mods. api alone still lets you figure out lots of stuff on your own 17:36 jin_xi literate lua or so 17:37 rubenwardy http://rubenwardy.com/minetest_modding_book/ ? 17:37 jin_xi oops, i thought i was in #minetest, sorry for the noise 17:42 rubenwardy Should I add cinematic mode on F8, and have different camera_smoothing values? #2325 17:42 ShadowBot https://github.com/minetest/minetest/issues/2325 -- Add camera smoothing by rubenwardy 17:43 rubenwardy Ie: cinematic_camera_smoothing when cinematic mode is on, camera_smoothing when cinematic mode is off. 17:52 kilbith that'd be helpful 18:42 VanessaE um, guys.... 18:42 VanessaE looks like the server has another memory leak, and it's a bad one: 18:42 VanessaE http://digitalaudioconcepts.com/vanessa/hobbies/minetest/stats.html 18:42 VanessaE don't ask for valgrind, I only saw it in the stats record just now 18:44 VanessaE screenshot of that page, in case no one can look in the next ~12 hours: http://digitalaudioconcepts.com/vanessa/hobbies/minetest/screenshots/Screenshot%20-%2002162015%20-%2001%3a48%3a04%20PM.png 18:47 rubenwardy Is it purple that we should look at? 18:47 VanessaE second graph, the bluish-purple yeah 18:48 VanessaE while it's marked as "programs", most of that is minetest instances. 18:48 nrzkt what was the tag on which you are ? 18:49 rubenwardy When the graph drops down, is that your server killing Minetest for too much memory? 18:49 VanessaE nrzkt: at the time that graph was rendered, a few commits behind on master branch 18:49 rubenwardy s/server/VM 18:49 VanessaE rubenwardy: nightly restart for taking backups 18:49 rubenwardy Ah. That is an impressive difference from end to start. 18:49 nrzkt yes.... 18:50 VanessaE (ignore that long CPU process from 0300-0430ish... that's just my Monday-morning routine maintenance stuff) 18:52 VanessaE "Creative" saw a big uptick in those ghost entities being deleted, and some activity today, and this apparent memory leak. I presume there's a correlation. 18:52 VanessaE s/today/yesterday/ 18:53 nrzkt have you got graphs for all week ? 18:54 VanessaE no 18:54 VanessaE I only keep the day's stats 18:55 VanessaE my *logs* go back farther, but I doubt they're useful here. 18:55 VanessaE minetest logs that is 18:55 nrzkt did you notice same thing this week ? 18:55 VanessaE I didn't see any problem in yesterday's graph, no, but every once in a while I see the graph spike like this. 18:57 nrzkt could you give me the commit you are 18:57 nrzkt ? 18:57 VanessaE currently, 98d80e20 18:58 nrzkt ok the last then 18:59 VanessaE before that, I think I was a few commits behind (maybe a week or so) + your patch 2/4 19:00 nrzkt i don't see a commit which uses memory without free 19:03 VanessaE that's why I'm guessing this is an older bug that's just rearing its head again 19:03 VanessaE something being caused by those ghost entities maybe 19:03 rubenwardy And the fun begins! 19:03 rubenwardy Finding year old bugs 19:04 * VanessaE shrugs 19:04 nrzkt i dont have time for this now but i try to look at this when i'll be free 19:04 VanessaE I filed a github report, it was marked as a "blocker" for 0.4.10 19:04 VanessaE then unmarked because they wanted to get 0.4.10 out the door 19:04 VanessaE now it sits in limbo 19:04 VanessaE nrzkt: ok 19:15 twoelk ha, is that me leaking memory on VanessaE's creative? 19:16 VanessaE twoelk: eh? 19:16 twoelk was playing near the northern ringroad junction with cheapie yesterday 19:16 * cheapie heard his name 19:16 twoelk all that traficlight stuff 19:17 cheapie That's quite the memory leak. 19:17 twoelk and just now got thrown when flying around my settlement around the white church 19:18 cheapie I've only added something like two traffic lights in the past week or so. 19:18 cheapie (bringing the total on North Street to 10) 19:19 twoelk interstingly I could only see stuff with farview on, otherwise the screen just went grey 19:42 rubenwardy Updated #2325 to add F8 and add that to the key config menu 19:42 ShadowBot https://github.com/minetest/minetest/issues/2325 -- Add camera smoothing by rubenwardy 19:56 est31 kwolekr == hmmm? 19:56 Krock yes 19:57 est31 ok, waiting for him to review #2225 19:57 ShadowBot https://github.com/minetest/minetest/issues/2225 -- Fix for getCraftRecipes implementation is very slow by gregorycu 19:58 est31 the issue that this PR fixes forced U-I to cache the results of that function 19:59 est31 its usage still causes lags, but at startup 20:46 est31 PilzAdam, can you review #2225? 20:46 ShadowBot https://github.com/minetest/minetest/issues/2225 -- Fix for getCraftRecipes implementation is very slow by gregorycu 20:58 est31 crafting is slow too 20:58 est31 sb converting a stack of wood into sticks spends most of their time waiting for the server 20:59 est31 causes 100% CPU spike too 20:59 nrzkt yes inventory is very very high in requests 21:00 est31 main problem is this for loop I think https://github.com/minetest/minetest/blob/70074800a20/src/craftdef.cpp#L904 21:00 est31 can be mitigated when you "hash" the crafting input and then do a lookup 21:00 est31 for crafts that are "free" you have to do normalisation firs 21:00 est31 t 21:01 est31 and group crafts can have their own loop after that 21:21 est31 but I'll focus on fixing #2308 first 21:21 ShadowBot https://github.com/minetest/minetest/issues/2308 -- Server should send other players inventories when its needed 21:22 est31 also, guess gregorycu knows that code better than me 22:49 est31 sfan5, do you maintain/run/host/are affiliated with master serveR? 22:49 est31 server* 23:36 est31 does cavegen and mud reflow areas larger than chunks?