Time Nick Message 00:04 hmmmm .. 00:04 hmmmm alright guys, seriously, this is the #minetest-dev channel, not #minetest-number-2 00:05 hmmmm i'm fine with having a little fun but you've been having a lot of fun lately i see 00:05 hmmmm in the wrong place. the reason why we're here and not #minetest-delta is to maximize the signal-to-noise ration. 00:05 hmmmm s/ration/ratio/ 00:08 jojoa1997|Tablet is there a way i can disable name changing here 00:08 jojoa1997|Tablet it chnages both of my names 00:08 jojoa1997|Tablet i only meant to change in #minetest 00:09 Exio nicks are per server, not channel 00:10 Exio if you want to do that, you can always /part and read the logs later 00:10 hmmmm also you join/part quite often 00:11 jojoa1997|Tablet um that is either my semi-ok connection at home or my bad connection at school 00:11 jojoa1997|Tablet and with school the best is 3 bars when it will shut off my wifi randomly 00:12 Exio if you have those problems you should try to stop irc-ing and studying stuff, no? :) 00:13 jojoa1997|Tablet i have all 95+ 00:13 jojoa1997|Tablet grades i think i can manage irc and work 00:14 Exio well, you can think that, but you can't stop being spammy and childish, so i don't get it, anyway 00:14 Exio we should move this to #mt 00:16 jojoa1997|Tablet meh if you want to see nick change heaven join #freenode 02:46 hmmmm hmm 02:46 VanessaE hm? 02:46 hmmmm PilzAdam, VanessaE, do you guys think we should flip on dungeons and jungles by default? 02:46 VanessaE dungeons, sure. 02:46 VanessaE jungles, well, you know my opinions thereof ;) 02:46 VanessaE but I guess so. 02:47 hmmmm oh right, your mod 02:47 hmmmm when I get DecorationDef in place, I might backport it to v6 02:47 VanessaE don't bother. 02:48 VanessaE by the time we get to a point where it's ready to backport, it won't be necessary - we'll all be using v7 by then I figure. 02:48 hmmmm i don't know, maybe there will be some crazy people who like v6 better for some reason 02:48 VanessaE perhaps 02:49 VanessaE but anyway, dungeons yes, jungles I suppose so 02:49 hmmmm i know pilz wanted it 02:49 VanessaE a lot of people do actually 02:49 Exio hmmmm: will be there any way to disable trees in lua? 02:49 VanessaE Exio: there already us. 02:49 VanessaE is. 02:49 hmmmm there's no real reason why we don't have it enabled by default either, i just wanted to keep things the same as they were unless they wanted it to be different 02:50 VanessaE oh hell I misread that 02:50 hmmmm exio, erm, if what i was talking about is backported, then yeah. 02:50 VanessaE hmmmm: well think of it this way - dungeons mostly exist underground, so for the majority of players, there won't be any sudden change with those 02:50 VanessaE jungles would be more visible, but a lot of people miss them 02:51 Exio i mean, so VanessaE's mod can "replace" jungle trees if jungles are enabled (and even replace the "whole tree gen") 02:51 Exio without having to do the actual "mapgen_aliases to air" 02:52 hmmmm you can already disable more regular trees from being generated by disabling trees in the config file 02:52 Exio i mean from a mod 02:53 hmmmm well if you're using that mod, you probably know that 02:53 Exio he :P 02:53 hmmmm and you can disable it on your own 02:53 VanessaE hmmmm: I think what exio is getting at is, let's suppose I added a GUI config screen to moretrees. a checkbox (or some equivalent) should be able to disable trees at the mapgen level. 02:53 hmmmm also, there are game-specific minetest.confs now 02:53 VanessaE but there's no reason to make a huge deal of it 02:53 hmmmm yeah, well, that sort of flexibility probably won't happen with the current system of generation 02:54 VanessaE aliasing mapgen trees -> air is easy enough and effective, if slower than it could be 02:54 VanessaE wait, idea: 02:55 VanessaE suppose I aliased mapgen trees to CONTENT_IGNORE instead of air. Could the tree generator be tweaked to watch for that alias and abort in that case? 02:55 VanessaE (as it stands now, if I do that I get "holes" in the map) 02:55 VanessaE (I tried it :) ) 02:56 hmmmm no, it can't be 02:56 VanessaE too bad. 02:56 VanessaE oh well :) 02:56 hmmmm what you're asking for is an official way to perform a hack, while a solution isn't being done because there's no way to do it that isn't hacky 02:57 VanessaE well... yeah, exactly. 02:57 hmmmm i still don't understand why you guys are so infactuated with your trees that you can't divert your attention to the config file and disabling them there 02:57 hmmmm s/diabling/disable/ 02:57 VanessaE hmmmm: because the average unwashed user doesn't know how to edit a config file. 02:58 hmmmm the true solution to this is to have a UI for these settings 02:58 hmmmm which i intend to get to 02:59 Exio well, yes :P 03:07 hmmmm haha oh wow http://ompldr.org/vaTBrcQ 03:07 VanessaE nice 03:07 hmmmm stereotypical "overgrown temple underground with single crack of light showing through" 03:08 hmmmm sapier really picked a great time to do all the cleanups 03:08 VanessaE translation: "I am DYING to push this stuff to master NOW." 03:08 VanessaE :) 03:08 hmmmm what i have now needs to be committed first or else there will be a lot more merge conflicts 03:09 hmmmm i already have to toss what he has for biome.cpp entirely since that whole file has changed 03:09 VanessaE better you should commit your stuff first then 03:09 hmmmm i know, but i'm not ready 03:09 VanessaE I'm sure he understands the need to rebase. 03:09 hmmmm people are going to fawn over this and start using it when it's not finished, then talk about how it sucks because there are no trees or caves 03:11 hmmmm i guess what i can do is comment out the registerMapgen line so nobody can use it without knowing why they can't, removing the comment, then compiling it on their own 03:13 VanessaE that would certainly work 03:13 VanessaE and would require the least amount of code changes to undo 03:19 Exio hmmmm: commit message: "Add mapgen v7 (not enabled)" 03:20 Exio and some commented lines 03:21 hmmmm my minetest folder is 498mb, 6662 files 03:21 hmmmm wow 03:24 VanessaE jzz 03:24 VanessaE jeez 03:24 hmmmm https://github.com/kwolekr/minetest/commit/472311ef6ef4e80b644e5781f374ee45d9ff2478 03:25 VanessaE "scranton.edu"... as in PA? 03:26 hmmmm yeah, i thought i told you I was in PA 03:26 VanessaE I'd forgotten. 03:26 VanessaE been almost 10 years since I was in PA last (scranton, in fact). 03:26 Exio PA? 03:26 * VanessaE continues reading 03:26 hmmmm you were in this shithole? :) hah 03:26 VanessaE yep 03:26 hmmmm exio, a state in the united states 03:26 VanessaE I used to work for Burst Net. :) 03:27 hmmmm aaah 03:27 VanessaE that little hole-in-the-wall ISP next to the kosher butcher shop 03:27 hmmmm i know them 03:27 hmmmm i know of them, rather 03:27 hmmmm i think they've since relocated to Dunmore 03:27 hmmmm (an even crappier municipality nearby) 03:27 Exio hmmmm: ah 03:27 VanessaE ah, I thought they had planned to take over the butcher shop. Did they ever get that piece of shit generator running? 03:28 hmmmm *shrug* 03:28 hmmmm lie i said, only know of them :) 03:28 hmmmm like 03:28 VanessaE anyways, 03:28 hmmmm yeah heh 03:28 hmmmm we're setting a bad precedent 03:28 VanessaE lol 03:28 hmmmm so good thing i push stuff like this to my own fork first, i see some things i need to fix 03:30 hmmmm also, something i noticed.. when people are adding new files to minetest consisting of only their own code, it's only proper for them to put their name in the license comment, not celeron's 03:30 hmmmm right? it's the license of that file, not of the entire project 03:30 VanessaE that's what I would do, sure 03:31 Exio it needs a compatible license with the other "code" ... 03:31 VanessaE until someone comes along and modifies my code, then it's only fair for that person to add their credit too, if they want 03:31 Exio ... wait, it is licensed under the same gpl as the game, so no problem! :P 03:32 Exio + //registerMapgen("v7", new MapgenFactoryV7()); aww 03:37 Exio hmmmm: what needs to get fixed in your actual code? 03:53 hmmmm you mean right away, or how much more needs to get done? 03:54 hmmmm right away, i'm just going to fix a minor whitespace issue, remove a printf i left in there, and the m_gamedef in BiomeDefManager 04:01 Exio hmmmm: in both 04:04 hmmmm of course, there'll be tons of tweaking over some period of time... the biggest things to do right now include getting getGroundLevelAtPoint() to work 100%, moving caves into their own file and using them in my mapgen, then start working on DecorationDef 04:04 hmmmm finally. 04:04 hmmmm i don't know how many times i've said "decorationdef", and now it's finally going to become real 04:05 Exio well, say it other time! 04:07 hmmmm also i need to balance out the biomes 04:09 Exio "balance out"? what does that mean? 04:09 hmmmm i don't know what the appropriate heat/humidity points are for biomes 04:10 Exio ah 04:11 VanessaE hmmmm: note the values used in plants_lib 04:11 VanessaE do I need to be ready for a change in those settings? 04:11 hmmmm you can change those settings if you'd like 04:12 VanessaE well, whatever values you settle on, let's make sure to coordinate with SPlizard also 04:12 hmmmm plants_lib just pseudorandomly places a bunch of different decorations in your own calculated "biomes", right? 04:12 VanessaE yep, pretty much so 04:12 hmmmm which is basically what decoraitondef is 04:13 hmmmm so i'm sorry, but i'm going to have to reduce your mod to a bunch of registrations 04:13 VanessaE lol 04:13 hmmmm like moreores 04:13 Exio hahaha 04:13 Exio lovely, "I'm sorry, You need will need to be clearer and faster :(" 04:13 Exio Your mod will * 04:13 VanessaE I expected some day, someone would surpass plants_lib :) 04:13 hmmmm i kind of feel sorry for everybody though 04:14 hmmmm here they're making these mods and they put a lot of work into it 04:14 hmmmm and then i come along and make it obsolete 04:14 VanessaE haha 04:14 Exio xD 04:14 VanessaE that's the whole point of progress though, isn't it? 04:14 hmmmm dunno 04:14 VanessaE to make slower stuff obsolete in favor of faster, better stuff? 04:15 Exio depends 04:15 jojoa1997|Tablet I say now pilzadam can't complain about me using plant livs 04:15 jojoa1997|Tablet For minitest 04:16 VanessaE jojoa1997|Tablet: eh? 04:16 jojoa1997|Tablet I cant add flowers cause pilzadam doesn't want to depend on plant_libs 04:17 VanessaE oh. 04:17 Exio wait some days 04:17 Exio you will see some nice stuff in the engine repo what you are you going to like 04:17 * Exio points at hmmmm with a laser 04:17 jojoa1997|Tablet Yep 04:17 VanessaE jojoa1997|Tablet: so use the old version of flowers: http://forum.minetest.net/viewtopic.php?id=2624 04:18 jojoa1997|Tablet Nah 04:18 VanessaE it's abm-based and uncontrolled, but it doesn't depend on plants_lib. 04:18 jojoa1997|Tablet I can wait 04:18 VanessaE this is all offtopic though. 04:18 jojoa1997|Tablet Pilzadam hates abms 04:18 jojoa1997|Tablet So... 04:19 VanessaE hmmmm: now all that said, I think plants_lib may still have some use even after v7 is out, as I figure there are still some controls it implements that won't be in v7 04:19 VanessaE (like what? I don't know right offhand) 04:19 jojoa1997|Tablet Will the optifine zoom pukk request be added 04:19 jojoa1997|Tablet VanessaE the ability to define your own biomes 04:20 jojoa1997|Tablet That is what will stay 04:20 Exio i will implement a decent and "less hacky" zoom-thingy when a big pulls gets on upstream 04:20 Exio and, the own biomes are the whole point of the mapgen v7 04:20 VanessaE jojoa1997|Tablet: that part is being implemented in v7, the right way. my biome controls will simply get a lot leaner and faster by letting the engine do most of the work. 04:20 Exio jojoa1997|Tablet: http://dev.minetest.net/Mapgen_V7 04:20 jojoa1997|Tablet Oh 04:20 Exio Mapgen V7 is the first map generator to use the biome infrastructure, arguably its central feature. 04:22 jojoa1997|Tablet Hmmmm you should update that wiki to give us a taster of what will be added 04:22 Exio jojoa1997|Tablet: he is working on some... stuff 04:22 Exio (read that as a big part of the mapgen v7) 04:23 Exio i don't think an highlight would help much 04:23 jojoa1997|Tablet Oh yeah :-D XD 04:23 jojoa1997|Tablet CANT WAIt for reals 04:23 jojoa1997|Tablet realms 04:24 Exio realms? 05:03 hmmmm i think kaeza's patch messed something up 05:03 hmmmm i keep getting an outdated version of the binary when i run it 05:03 VanessaE which patch? 05:04 hmmmm the one where he moved binary builds to a separate folder 05:04 VanessaE ah 05:04 * VanessaE checks her build 05:04 hmmmm indeed it builds in src/bin now 05:05 VanessaE hm, my end seems to be clean 05:05 hmmmm and not bin 05:05 VanessaE but I don't use the feature he introduced 05:05 hmmmm this screws up run in place 05:05 VanessaE oops! 05:05 hmmmm MMMMMMm..... 05:05 VanessaE wait that seems wrong...I don' 05:06 VanessaE don't see in his patch where it would do that 05:06 VanessaE aw shit 05:06 VanessaE it did it on mine too 05:06 hmmmm yeah 05:06 VanessaE good thing I don't built run-in-place 05:07 jojoa1997|Tablet What is run in place 05:07 VanessaE build* 05:07 VanessaE jojoa1997|Tablet: when you prepare a package that can just be unzipped and run immediately without installing. 05:07 jojoa1997|Tablet Ow is it a way to build 05:07 * hmmmm shoots self 05:07 jojoa1997|Tablet Like pilzadams downloads 05:07 hmmmm why did i commit this without thinking about it 05:07 VanessaE hmmmm: rushed maybe? 05:07 VanessaE jojoa1997|Tablet: perhaps, dunno 05:08 hmmmm actually, looking at it, i think this might be my problem 05:09 VanessaE hmmmm: right, I don't see how his patch would cause this. 05:09 hmmmm CMAKE_CURRENT_BINARY_DIR is somehow src, and i guess it used to be that CMAKE_SOURCE_DIR is ..? 05:09 hmmmm or rather . 05:10 hmmmm but where is CURRENT_BINARY defined in the first place 05:10 VanessaE it got saved into CMakeCache.txt... 05:11 hmmmm ahah 05:11 hmmmm thank you 05:12 VanessaE you're welcome :) 05:12 hmmmm actually not so sure if that did the trick 05:12 VanessaE (not that I've found yet where that variable is being plugged into there) 05:13 hmmmm it still builds to src/bin 05:14 hmmmm "Linking CXX executable bin/minetestserver" 05:14 hmmmm so it conveniently omits the src/? 05:14 VanessaE did you clear CMakeCache? It's set in ther,e too 05:14 VanessaE -, 05:14 hmmmm i deleted CMakeCache.txt 05:14 hmmmm hold on 05:14 jojoa1997|Tablet night all 05:14 VanessaE hrm 05:15 jojoa1997|Tablet what 05:15 VanessaE night jojo 05:15 jojoa1997|Tablet hmmmm what do you want 05:15 hmmmm uhh.. nothing... 05:15 VanessaE jojoa1997|Tablet: translated, "wait a sec, this looks odd" 05:15 VanessaE not "please wait for me" 05:16 jojoa1997|Tablet oh ok 05:16 jojoa1997|Tablet too tired to thinbk cya 05:16 jojoa1997|Tablet yfhf 05:31 hmmmm done https://github.com/minetest/minetest/commit/fbf76184f0b8ec715255ed1246e49b575fc84873 05:33 hmmmm well of course it'd work 05:33 hmmmm why would you need to test changing back something to its previous state 05:34 VanessaE never hurts to be doubly sure, paranoia 05:34 VanessaE yup, that's got it back to normal. 05:34 VanessaE (I didn't expect it not to, but I'm crazy like that) 05:35 hmmmm alright - so according to the official cmake documentation, we know that kaeza's patch can't work, since it's predefined to be the source dir 05:36 hmmmm which is going to bring up the question "can it be done right" 05:36 hmmmm i have no idea. i really don't know too much about cmake and i'm not interested in knowing either 05:37 VanessaE well the ultimate reason for the patch is to let one drop the binary into an arbitrary directory... if we know src is always the same, can we not append "../"${SOME_OTHER_VAR} <-- i.e. the relative path of the desired build folder? 05:38 VanessaE it's hacky of course 05:38 hmmmm oh god no 05:39 hmmmm whatever i guess.. if the patch produces a tangible benefit and it works for RUN_IN_PLACE too, then sure, fine, i'll add it 05:39 hmmmm i don't care how it accomplishes that task at this point 05:39 VanessaE didn't think so. first thing I could think of since I understand even less of cmake. 05:49 hmmmm i'm just going to make a bunch of little fixes now 05:50 hmmmm https://github.com/minetest/minetest/commit/5961106f9ffd9e9a2bc23e4bd700f6ce07505c9a 05:50 hmmmm vanessae, what do you think, should i fix the cave shadows? 05:51 VanessaE depends on how complex of a fix it would be 05:51 hmmmm i change a digit by 1 05:51 VanessaE this time of night, I'd tend to avoid.... 05:51 VanessaE well in that case sure 05:52 hmmmm yeah but then it has to compute the lighting of 47 more blocks 05:52 hmmmm 7^2 05:52 hmmmm (chunksize+2)^2 to be more specific 05:52 VanessaE ack 05:52 hmmmm i guess it's the price you pay for having caves above ground 05:53 hmmmm i'll add a setting so v7 doesn't have to suffer as well 05:53 VanessaE ok 05:53 VanessaE considering the improvements you've made to the lighting code though, would that difference really be noticable? 05:53 VanessaE I mean compared to say 0.4.4 05:53 hmmmm no idea 05:54 VanessaE hrm.. 05:54 hmmmm i'd say so, because look at how much area you're adding to be computed.... but most of this is CONTENT_IGNORE, so it doesn't recurse 05:56 VanessaE I just don't know. 05:56 VanessaE part of me says no, don't fix it because of the potential performance regression, part of me says the opposite because people obviously prefer nice-looking maps 05:57 hmmmm i'll profile a bit 05:57 VanessaE yeah, I was going to suggest that 05:57 hmmmm calcLighting() has an extremely variable execution time though 06:15 VanessaE ok, time for bed. 06:15 VanessaE I'm outta gas for the night. 06:27 hmmmm https://github.com/minetest/minetest/commit/96e2931c5c154262566359495d4a0b8515376947 08:17 celeron55 < hmmmm> also, something i noticed.. when people are adding new files to minetest consisting of only their own code, it's only proper for them to put their name in the license comment, not celeron's 08:20 celeron55 not doing that can be about two things: either they are simply lazy copypasters, or they want to give the rights to me - but i think doing only that for giving rights isn't really enough to make it legally effective because of the possibility of the first one... so basically the copyright notices in the files are only for the purposes of "who maybe knows about the actual license stuff" and "in what decade this code was written" 08:23 celeron55 everything has been modified by so many people anyway that the end result isn't practically relicensable anyway, so IMO there's really nothing to be gained with better usage of those; i don't like my name being slapped on code i haven't written though, because it makes people think i am to blame 08:25 celeron55 maybe the lines should be replaced with something like "Copyright (C) 2013 various people, look at the git commit log" 8) 08:27 celeron55 i guess this the reason why FSF recommends assigning copyright to them 08:27 celeron55 +is 08:28 celeron55 them = the FSF itself 09:23 Taoki Oh my, mapgen v7 is in :D 09:23 Taoki Does this mean biomes are defined entirely in Lua now, and the whole alias_dirt / alias_sand / etc. things can be removed? 09:24 sapier I doubt anything can ever be removed for compatibility reasons ;-) 09:24 sapier but I'd be glad this would change :-) 09:25 Taoki That's ok. But like, can the game function without them now? 09:26 sapier I thought it always was working without only requireing it to be able to use old maps 10:22 RealBadAngel damn, it looks like i got luajit to compile properly :) 10:22 RealBadAngel with original luajit cmake files, not some copypasta 10:31 Calinou_ /home/calinou/_VCS/minetest/src/mapgen.h:133:2: erreur: ‘NoiseParams’ does not name a type 10:31 Calinou_ ^ can't compile 10:32 Calinou_ redownloading the minetest git repo, it looks like I've got corrupt files 10:33 RealBadAngel im compiling right now fresh sources, havent seen such error 10:34 * Calinou_ crashes as soon as he uses make -j8 10:34 * RealBadAngel is using -j2 10:34 Calinou_ -j2 is slow 10:34 Calinou_ really 10:34 RealBadAngel i will try -j8 then, im compiling now time after time 10:34 RealBadAngel testing luajit 10:42 Calinou how does one switch to mapgen v7? 10:42 RealBadAngel question: make luajit enabled by default? 10:42 RealBadAngel or enable it with compile option? 10:43 RealBadAngel i mean which one shall be default: old lua or luajit? 10:44 PilzAdam RealBadAngel, I guess we are not going to include LuaJIT src into minetest 10:44 PilzAdam so use it if you find it in the system 10:45 RealBadAngel im using original jit distribution 10:45 Calinou what's the difference between "indev" and "v6"? they seem to be the same 10:45 RealBadAngel modified only not to install libs system wide 10:45 PilzAdam Calinou, farlands and float islands 10:45 RealBadAngel and i got it working 10:46 PilzAdam RealBadAngel, why dont you use the system wide version? 10:46 RealBadAngel because we dont want to fiddle with different and incompatible versions of it 10:47 RealBadAngel if we will be using one version known its good, upgrading to another will be our decision 10:47 Calinou can't someone revert how plantlike nodes look now? it looks horrible sometimes since there's an horizontal symetry 10:47 PilzAdam are there incompatible versions? 10:47 Calinou that might be good for a few nodes, but for plants and apples it looks bad :P 10:47 RealBadAngel which will be just replace jit's folder with new sources and commenting out install options 10:49 RealBadAngel anyway im making jit default, if somebody will want to use old lua will disable it just 10:49 PilzAdam RealBadAngel, are there any LuaJIT version that dont support our Lua version? 10:50 RealBadAngel old ones? compiled with disabled lua compability? 10:51 RealBadAngel you want to fight with such problems? 10:52 PilzAdam I mean we have Lua already in the src tree, we dont need both 10:52 PilzAdam if the LuaJIT version is incompatible or not there then fall back to Lua 10:54 Taoki Anyone know if the MineTest code has a bound function? As in, bound(0, myvalue, 1); 10:54 PilzAdam rangelim(x, min, max) 10:54 Taoki thanks 10:55 RealBadAngel PilzAdam, afaik we DO need both 10:55 PilzAdam why? 10:55 RealBadAngel luajit doesnt work on some exotic platforms as hmm noticed 10:55 RealBadAngel so in such cases users will be able to pick lua 10:56 PilzAdam thats why I want Lua as fallback 10:56 PilzAdam and use the system wide LuaJIT if present 10:59 RealBadAngel forcing players to install dozens of libraries is not a good idea 11:00 PilzAdam ask a package maintainer about that 11:00 PilzAdam they already complain that we have jthread in the src tree 11:00 RealBadAngel instead of click and play its going to be hardcore for newbies 11:00 PilzAdam ehm 11:01 PilzAdam n00bs will download the win version 11:03 RealBadAngel so there are no noobs on other platforms? 11:03 PilzAdam the n00b argument isnt good when it comes to compiling 11:03 PilzAdam also, it is the same thing with gettext 11:05 Calinou it's not hard to paste a command in a terminal 11:06 Calinou don't make luajit a requirement, just recommend it 11:06 Calinou but packaged installs should use luajit 11:06 PilzAdam ^ same as gettext 11:06 Calinou but luajit should be used as default when building if it is available, unlike gettext 11:14 RealBadAngel anybody here who can compile on windows? 11:14 PilzAdam cross compile? 11:15 RealBadAngel propably both, cross and in windows 11:15 RealBadAngel i will upload sources shortly, hold on 11:23 RealBadAngel https://github.com/RealBadAngel/minetest 11:23 RealBadAngel sources are up to date, luajit added, compile option -USE_LUAJIT=0 is to disable luajit and fallback to lua 11:33 RealBadAngel PilzAdam, can you make win build out of this sources? 12:43 RealBadAngel anybody wanna try? https://github.com/RealBadAngel/minetest 12:45 Zeg9 I guess I'll try 12:46 Taoki PilzAdam: More eye candy from me. Please test and merge if ok when you have some time; https://github.com/minetest/minetest/pull/609 12:49 Zeg9 RealBadAngel, fun to see luajit is about 25% of the source files to compile 12:54 Zeg9 RBA: Compiles fine, but when I start a singleplayer game, it has segmentation fault. 12:56 Zeg9 RealBadAngel: 12:56 Zeg9 Program received signal SIGSEGV, Segmentation fault. 12:56 Zeg9 0x00007ffff578e1ed in lj_vm_cpcall () 12:56 Zeg9 from /home/guillaume/Desktop/minetest/src/luajit/liblua.so 12:57 sapier directly on start or on restart? 12:57 Zeg9 On start of a server/singleplayer world 12:57 sapier I assume on first start? 12:58 Zeg9 Well, what do you mean by restart... 12:58 Zeg9 oh. The world was already created 12:58 RealBadAngel wonder why the hell its working fine here :) i will try to make game use system wide library 12:58 sapier login in a world after logging out without closing minetest 12:58 Zeg9 This shouldn't make it crash, as lua is server side (or am I wrong? lol) 12:59 RealBadAngel but propably it wont be any different because jit compiles from the very same sources 12:59 sapier as minetest itself doesn't cleanup anything this may result in almost anything 12:59 Zeg9 I can connect fine to servers, but can't start one, either from menu or standalone server 12:59 sapier to many "anythings in one sentence" 12:59 Zeg9 The standalone doesn't even print something, just segmentation fault 13:00 sapier if you start a server there's even more cleanup to be done after exiting a world 13:01 Zeg9 Wait, I get some more warnings, I don't think it's useful though, it talks about linux-vdso.so.1 13:01 Zeg9 Yeah, it also says that with a "vanilla" server (without jit) 13:03 sapier vdso is library loader 13:13 VanessaE good morning all. 13:19 Zeg9 Hi 13:24 Taoki hi 13:31 PilzAdam Taoki, why do you make it configureable? 13:31 Taoki PilzAdam: It makes sense to. So people can configure intensity and turn it off 13:31 PilzAdam damage screen (red + camera tilt) isnt configureable 13:32 PilzAdam I dont like to make too many things configureable 13:34 Taoki Will remove two irrelevant options, but I'll leave at least one enabled (at least so it matches the normal view bobbing) 13:35 Taoki Don't fork yet. But check if there are other issues too 13:35 PilzAdam I wouldnt make it confifureable at all 13:35 Taoki View bobbing is and some might want to turn it off, so it's fair enough to leave just one setting 13:36 PilzAdam from what I understood its just a tilt of the camera when you hit the ground, right? 13:36 Taoki teah. Similar to view bobbing 13:36 Taoki **yeah 13:36 Taoki People might wanna configure the intensity 13:37 PilzAdam so, it doesnt appear all the time (unlike bobbing when walking) so it doesnt have to be configureable 13:37 PilzAdam making too many things configureable isnt good 13:37 Taoki People might still want to disable it or make it stronger. I'll switch it to only one setting for now since I do think that's fair enough 13:38 PilzAdam the damage screen isnt configureable 13:38 Taoki I didn't implement it, if I did it might have had a single toggle setting :P 13:38 Taoki One setting won't hurt anything though 13:39 Taoki Working on removing the others for now, those are indeed too much 13:42 PilzAdam I say dont add a setting for it. 13:42 Taoki Are there any other issues other than the settings? 13:44 PilzAdam looks good, havent tested it yet 13:44 Taoki ok 13:54 Taoki PilzAdam: Pushes a commit removing the useless settings. Still left the primary switch on however... it feels best to have that and the setting doesn't harm anything 13:55 Taoki Maximum velocity now uses the jump setting, primarily for the physics configuration. That was a smart way to use an existing setting automatically :) 13:58 Taoki PilzAdam: Actually, I can do something else and amend my commit; Have the view bobbing switch control the fall bobbing as well 13:58 Taoki That would make sense actually. If you enable view bobbing you probably don't mind the falling view either 13:58 PilzAdam the view bobbing is for walking 13:59 PilzAdam while the tilt when falling is somehow static 13:59 Taoki Yeah, but this is also view bobbing. 13:59 Taoki Meant for the same purpose and effect, even if achieved by a different trigger 14:00 PilzAdam isnt it just a tilt? 14:00 Taoki It lightly moves the camera down and back up when you fall. Which is what view bobbing does as you walk too 14:01 Taoki This is sort of a 2nd part of the view bobbing system 14:01 PilzAdam it is the same effect as camera tilt on damage, right? 14:01 Taoki Almost. It doesn't rotate the camera to tilt it, just lowers it 14:02 PilzAdam then it has nothing to do with the walking bobbing 14:02 Taoki Though I might add a bit of tilting too now that I'm working at it 14:02 PilzAdam how does it look if you have the damage and the falling tilt? 14:02 Taoki ok. I think it does, but if not I could do something else; Make an option for view tilting, which affects both the falling and damage tilt 14:03 Taoki The effects shouldn't conflict 14:03 Taoki Tilting also doesn't lower the camera, so it's all good 14:04 Taoki PilzAdam: Anyway, I'll make an option for both the damage tilt and this, so it's a single setting that makes sense for multiple things. At least that should be good :) 14:05 PilzAdam no 14:05 PilzAdam you make it even worse 14:05 Taoki It should have a way to be toggled since people might not like the feel with it on 14:06 PilzAdam people might not like minetest, so what? 14:06 Taoki Currently it only has one setting, I think it can stay as such 14:06 Taoki that too :P 14:07 PilzAdam I dont like the idea to make every single new feature configureable 14:07 VanessaE it helps you avoid losing users. 14:08 Taoki IIRC it's best to make something configurable when it makes sense to and people might want it on or off 14:08 VanessaE and G*d knows, we need headcount 14:08 PilzAdam VanessaE, why do we need users? 14:08 VanessaE PilzAdam: because without a userbase, developers eventually get tired of bothering. 14:08 VanessaE "Why do I screw with this if no one plays? fuck it, I quit." and the lie. 14:08 VanessaE like* 14:09 * Taoki agrees with VanessaE. It's not like we're making MineTest to go together to the moon and play it ourselves only :P 14:09 VanessaE this same scenario has been played out countless times in multiple segments of the software market. 14:09 Taoki PilzAdam: Anyway, the last commit feels like the best configuration to me. If you can go with leaving out that one setting, can you test it and see if everything is fine to go? 14:09 VanessaE and it has always ended up the same: the project gets abandoned if no one besides its developers use it. 14:10 sapier as far as I have experienced adding additional settings doesn't improve user acceptance ... but this may not be true for all setting options 14:10 PilzAdam VanessaE, {{cn}} 14:10 Taoki VanessaE: Many FOSS games have huge sissues with staying alive cuz not enough people like them, although they are awesome. MineTest is lucky to have people 14:10 Taoki #minetest has about 100 users at all time. That's something to be very lucky about 14:11 sapier at least any setting has to be documented very well in order to be accepted 14:11 Taoki sapier: It is in minetest.conf.example 14:11 sapier #mgv7_np_terrain = 10, 12, (350, 350, 350), 82341, 5, 0.6 14:11 sapier #mgv7_np_bgroup = 0.5, 0.3125, (350, 350, 350), 5923, 2, 0.6 14:11 sapier #mgv7_np_heat = 25, 50, (500, 500, 500), 35293, 1, 0 14:11 sapier plz tell me what this examples show ;-) 14:12 Taoki Yeah, those have no description indeed 14:12 PilzAdam bbl 14:12 sapier # Offset, scale, spread factor, seed offset, number of octaves, persistence 14:12 sapier is explanation to it ... but doesn't help much more 14:12 Taoki ok 14:12 sapier yes if xou have written this code you know what offser, scale etc mean 14:13 sapier but if not these settings are useless 14:13 sapier I do same mistake very often 14:37 sapier https://github.com/minetest/minetest/pull/613 14:37 sapier just small bugfix to improve modders formspec experience 14:39 VanessaE sapier: something I'd like to see with formspecs is being able to turn off or at least reposition the default footer text 14:39 VanessaE (the "Left click: Move all items[...]" one) 14:41 sapier this is not formspec but inventory ... but I'll have a lok if and how this could be done 15:27 hmmmm guys, i disabled mapgen v7 for a reason 15:27 hmmmm don't use it until it's ready 15:27 hmmmm i need it in there so other patches don't screw it up 15:28 RealBadAngel hi hmmmm 15:28 hmmmm hi 15:29 RealBadAngel i managed to merge jit into main tree but for some weird reason it compiles for some ok, for some segfaults after, for some stops during compiling 15:29 RealBadAngel weirdo, im trying now using system wide one 15:30 RealBadAngel for me it compiles and works flawlessly 15:30 VanessaE for me, his latest attempt appears to build fine, but then segfaults as soon as I try to start a singleplayer world. 15:30 VanessaE prior to that, it refused to build at all. 15:37 sapier hmmmm did you already start to merge the memory leaks? if not I offer splitting it up to different topics ... still everything should be added but maybe review is easier if similar changes are linked together 15:38 sapier "merge memory leaks" :-) I meant memory leak fixes of course ;-) 15:56 hmmmm hrmm 15:56 hmmmm sapier, i didn't start to merge them, just the smaller pull 15:56 sapier ok I'm going to split it 15:57 hmmmm basically my intention was to go through each commit one by one and see what needs fixing and what doesn't 15:57 sapier e.g. gettext cleanup can be separated 15:57 hmmmm the noiseparams are handled in mapgen_v7.cpp the same way 15:57 hmmmm for consistency 15:57 hmmmm so i'll go and fix those all at once 15:57 hmmmm biome.cpp is completely different, so i'll have to handle that separately too 15:58 sapier ok no problem I'll split it in order to reduce work for you 15:58 hmmmm alright thanks 15:58 hmmmm btw i removed l_register_biome_groups 15:58 hmmmm also, Biome doesn't have a NoiseParams anymore 16:00 sapier btw the formspec fix I just posted is quite ugly imho ... I'm not quite sure why formspec was working by now but if this was correct my changes are too :-) 16:01 sapier still I suggest adding it as currently only formspec is refreshed but not formspecname ... rendering new formspec useless if it depends on different name 16:02 sapier would you prefere different commits or even different branches? 16:38 RealBadAngel https://github.com/RealBadAngel/minetest 16:39 RealBadAngel tree with system wide luajit use if aviable 16:39 RealBadAngel anybody wants to try? :) 17:06 Taoki hmmmm: Does mapgen v7 enable the new indev features too? Namely floatlands and the high spikey mountains around world edges 17:07 hmmmm not in the slightest, mapgen v7 is entirely different from any other 17:07 PilzAdam of course not 17:07 Taoki Ahh. I thought eventually that all new mapgen features get merged in one 17:08 Taoki And indev is just the beta mapgen which has features in the next release 17:08 hmmmm that's not possible when these different mapgens take fundamentally different approaches to creating terrain 17:08 hmmmm indev is an extension of v6 17:09 hmmmm it's currently v6, in further development so as to not break compatibility 17:10 hmmmm i should probably explain this somewhere 17:16 RealBadAngel hey, anybody will test jit? 17:16 RealBadAngel i would like to know if it works for others 17:16 PilzAdam not yet, Ill test it later 17:17 Jordach RealBadAngel, testing with mesecons / digilines and worldedit operations 17:17 Taoki hmmmm: Would be nice if something could be done to get the important features together. For instance I'd love your v7 mapgen but without losing floatlands 17:17 Jordach RealBadAngel, is luajit enabled by default or cmake option 17:18 RealBadAngel Jordach, mt will use jit if it is installed 17:18 hmmmm taoki, the plan is for v7 to have its own floatlands realm occuring at ~4000 17:18 RealBadAngel if not, it will use old lua. no config options needed 17:18 Taoki hmmmm: Oh, ok. Not bad then :) 17:19 * Jordach forgot to install apt-get luajit 17:19 Taoki hmmmm: I love v7 cuz it finally gets rid of alias_dirt / alias_sand / etc. and allows defining biomes in Lua, in case I understand correctly. Wanted that for some time :D 17:19 hmmmm what are you talking about, alias_dirt? 17:19 hmmmm you still have aliases to nodes 17:20 Taoki ah 17:20 Taoki Yeah, I thought you could make your own biomes in v7 17:20 Taoki Or so I heard 17:20 hmmmm you could, but i don't get what that has to do with node aliases... 17:21 Taoki I thought node aliases are due to the C++ mapgen holing in viomes 17:21 Taoki Like minetest.register_alias("mapgen_stone", "default:stone") 17:21 Taoki Unless you can make your own of these in Lua, it makes sense then 17:21 hmmmm nope, node aliases never had anything to do with biomes 17:22 hmmmm aliases are there for flexibility 17:22 Taoki Ahhh, I thought they're the C++ hooks which told the mapgen "for this part of the noise you use dirt node" 17:22 Taoki Disregard what I said then, they're good :) 17:23 hmmmm if for whatever reason you need to override what a certain node is, you can do that with aliases, you can also give nodes shorter names or different names, etc. 17:23 Taoki yeah 17:23 Taoki What does tell the C++ mapgen which node to spawn for dirt, which for desert sand, etc? 17:23 Taoki Since nodes are defined in Lua but at least with < v7 mapgen biomes are hard-coded 17:24 hmmmm the Server's INodeDefManager converts a string name of a node or node alias into an actual content ID 17:24 Taoki ok 17:24 hmmmm the v6 'biome' is a really poor excuse for a biome 17:25 Taoki Will mapgen v7 remove that entirely and put even the default biomes in a script? I'd like that personally 17:25 hmmmm my goal isn't to remove 17:25 Taoki As in, no deffault biome in the code, only code to define them in Lua 17:25 hmmmm it's 100% possible to add this new mapgen without messing around with the old one 17:25 hmmmm and that's the plan 17:26 Taoki ok 17:26 hmmmm it's not even the plan 17:26 hmmmm it's already been done 17:26 hmmmm it's been done 3 times in fact 17:26 hmmmm indev, singlenode, and v7 are all their own self-contained mapgens that you can specify to use 17:27 hmmmm none of them interfere with the others 17:27 Taoki If someone wants to use v7 to make a world with only original biomes however, will it offer a way to disable the hard-coded ones? Also, will the hard-coded biomes stay enabled by default (even if not removed for compatibility) 17:27 Taoki nice 17:28 hmmmm perhaps i should explain this in more detail 17:28 hmmmm the whole concept of a biome in V6 and indev is self-contained 17:28 hmmmm there is an infrastructure, however, that mapgens may or may not make use of if they feel so inclined, called BiomeDefManager 17:29 Taoki Ah, I understand better now 17:29 hmmmm there is one BiomeDefManager associated with the EmergeManager, that contains registered biomes and noises for heat and humidity 17:29 hmmmm mapgen v7 is the first mapgen to make use of this 17:29 Taoki fun fun 17:29 hmmmm when you register a biome, it gets added to the BiomeDefManager, which is then reflected in any mapgen that makes use of BiomeDefManager 17:31 hmmmm this componentizes biomes and separates them from the mapgen entirely... when you want to make a change to the way biomes work, you change BiomeDefManager, which would in effect change the way biomes work for all mapgens using this infrastructure 17:31 Taoki That is good 17:31 hmmmm so this is definitely a separate entity from a mapgen 17:32 hmmmm the way BIOMES are decided is the same way (i think) Minecraft does it 17:32 hmmmm the whole idea of having two intersecting noises, temperature and humidity, was there the whole time, and at first I was checking if a biome was within max/min value bounds for these 17:33 hmmmm celeron suggested I find the distance between a temperature/humidity point within space and the idealized temperature/humidity points for each biome, and choose the closest biome as the one that node is 17:33 hmmmm a very good idea 17:34 hmmmm after looking at minecraft's code, they do it the same way 17:34 hmmmm he also suggested that i do the same thing with the amount of variance in the terrain within a biome, but that didn't work out, since i had many more noises interacting with eachother 17:35 hmmmm i stuck with the idea of a minimum/maximum height, there 17:35 Taoki Can't understand how temperature / humidity works or what that means in map terms. I always thought of biomes as defining large areas where hills / caves / etc. have certain shapes and frequencies, and where certain types of nodes are used instead of others 17:35 hmmmm this way i can have for height_min= -1 and height_max = 2, three possible biomes, a sandy beach, a gravely beach, or just a dirt coast 17:35 Taoki So for example to define desert, you'd say "Desert biome has this frequency, uses desert stone and sand in this pattern, and has fewer mountains" 17:36 hmmmm yeah, frequency, not so much 17:36 hmmmm the usage of heat and humidity is much better 17:36 hmmmm with this, you'll never see a desert occuring right next to an icy tundra 17:37 Taoki yeah. Hard to understand exactly what they mean 17:37 hmmmm they mean what they do in real life 17:37 hmmmm you have some very hot places, which could be either very dry or very wet 17:38 Taoki Ah. Heat is biome type then (snowy, green, desert, and anything in between) and humidity... I assume that's the frequency of lakes and water 17:38 hmmmm if you have two biomes, jungles and deserts, this would be sort of like: heat_point = 90 for both a desert and a jungle, but humidity_point=10 for a desert, and say humidity_point=80 for a jungle 17:39 hmmmm frequency of water is only determined by height 17:39 * Taoki nods, makes more sense now 17:39 hmmmm a body of water would be an Ocean biome 17:39 hmmmm it'd be just one biome that occurs at negative range of terrain heights 17:40 hmmmm perhaps it would make more sense if i posted the biome definions i test with 17:41 Taoki Would be fun to see them and try them out :) 17:42 hmmmm http://pastebin.com/4RJVZQ4J 17:43 Taoki Ah, nice. That's more simple and clear and easy than I thought :) 17:43 hmmmm of course there will be more later on. 17:44 hmmmm the plan is to attach a list of decorations (trees, cactuses, grass, etc.) for biomes, have a list of mobs that can spawn there 17:44 hmmmm ores 17:44 hmmmm and i would like some client-side visual on differences in biomes, like in minecraft, how the color of the grass has a different shade 17:44 sapier hmmmm I suggest mobs to decide which biome to spawn in not biome to decide which mobs 17:45 hmmmm that's possible 17:45 Taoki Colorizing nodes will have to wait for hardware lighting. I already coded the feature though 17:45 Taoki But you won't see colors until HW lighting is implemented :P 17:45 Taoki That can be used to change grass color per biome tho 17:45 Taoki sapier: Agreed. It should be possible to get the biome at a certain position and take decisions off that 17:45 hmmmm taoki, I'd like to actually talk to you about that 17:46 hmmmm I have this feature that i wanted to implement, which i call the 'colorlike' node type 17:46 hmmmm it's like facedir, wallmounted, etc. an attribute that uses param2 of nodes 17:47 hmmmm so the thing is, you register along with the node that's colorlike, an RGB list of colors to use for that node 17:47 hmmmm then on drawing time, the node is shaded to the color at the index specified by param2 17:48 hmmmm the original point of this is to eliminate the 50000 different nodes people have in mods that are the same exact thing but in a different color 17:48 hmmmm but it can also be used for colorizing nodes in different biomes 17:48 Taoki Hmm. I had my own plan for that... already in the code but not there yet 17:48 Taoki Which is simply colorizing surfaces with the proper Irrlicht functions for setting surface colors 17:49 Taoki However,t his requires hardware lighting. Which Minetest will eventually need, but I don't know when anyone can do it 17:49 hmmmm http://dev.minetest.net/TODO#Add_colorlike_to_node_param_types 17:49 Taoki Funny thing is, the functions to colorize already are in, and should work. But because there's no real light, you can't see them :P 17:49 sapier btw I don't want anyone to compain about me spaming pull requests :-) all of those fixes are valid ;-P 17:49 hmmmm indeed 17:50 Taoki That approach would prolly be a lot better (colorizing with Irrlicht). But c55 isn't interested in coding light, I have no idea (already tried), and I don't see anyone else who might 17:50 hmmmm hardware lighting is very difficult 17:51 hmmmm if you want to try something, you may, but nobody else would be willing to spend time on it 17:51 Taoki Yeah. But also very necessary :P 17:55 sapier very necessary is cleaning up memory leaks ... adding features is benefit 17:56 sapier at least thats my opinion 17:56 Exio and bugfixing 17:56 Exio but reworking some stuff 17:56 Exio ends in stuff fixed 17:56 Taoki Engine priorities as I see them now are hardware lighting and mesh optimization (to increase performance which is still very bad with long draw distance). Just how I see it that is 17:56 sapier no adding features is adding bugs ... you can't avoid this ;-) 17:57 Taoki IIRC block surfaces are still drawn invidivually 17:57 Exio sapier: touching any code will end with new bugs 17:57 Exio even fixing leaks 17:57 sapier no :-) there's one operation you will fix bugs in general 17:59 Taoki PilzAdam: Feel like finishing with the view falling code now if you have time? In case you can make that compromize to allow one setting to toggle the feature (which I'd really prefer). I tested already and it seems to work fine in all circumstances. 17:59 sapier "del" is bugfixers best friend ;-) if code isn't necessary delete it ... code not present can't have bugs 18:00 Exio there are some ways to fix bugs 18:00 Exio one of them is rewriting the code what is bugged as hell for something cleaner, easy to understand and better than before 18:00 Exio what will add news bugs, as everything 18:00 Exio but at the same time will fix the bugs for what it got rewrote 18:01 hmmmm taoki, indeed those are pretty important... most of us aren't decent enough with or even interested in the 3d graphics part of it 18:01 hmmmm so if you don't do it, it'll probably never get done 18:01 PilzAdam RealBadAngel, tested LuaJIT on Linux, seems to work fine 18:01 sapier true but sometimes it's difficult to decide if code ieams to be complex becaus its wrong or complex because you didn't understand what it's doing 18:01 hmmmm minetest would still live on without these fixes, but it'll just stay crappier than it could've been 18:01 Exio hey, i'm interesed in that! but i don't have any idea of coding anything of what i want :P 18:02 Taoki BTW. My Linux distro doesn't have luajit in its packages. Is that automatically installed with Lua itself? 18:02 RealBadAngel PilzAdam, i noticed it doesnt work when theres no luajit even :) trying to fix it now 18:03 PilzAdam it works for me if no LuaJIT is present... 18:03 hmmmm taoki, if there's no luajit, it should fall back to regular lua 18:03 RealBadAngel Taoki, http://luajit.org/download.html 18:03 hmmmm that's what we're aiming for, but for some reason it seems that people are having trouble doing this 18:03 RealBadAngel unpack, make, make install 18:04 Taoki Probably not for me. If it's not in distro packages it probably means it's not good, popular or maintained enough :P 18:04 RealBadAngel or distro is worth shit ;) 18:04 RealBadAngel debian squeeze also doesnt have jit, or irrlicht 1.8 18:05 RealBadAngel and propably wont have for next few years ;) 18:07 Calinou hmmmm: why not make indev the default mapgen? it works nicely and has nice features :P 18:08 hmmmm because something in development shouldn't be the default 18:08 hmmmm the default should be the stable, working thing that people expect 18:09 * Taoki prefers v7 mapgen to be default. Erm, if it will have floatlands which I heard it will 18:10 hmmmm floatlands are really the dealbreaker!? 18:10 Taoki Not the most important thing, but one of the most awesome new stuff in MineTest if you ask me :) 18:11 Calinou hmmmm: well, it works fine 18:11 Calinou people use arch right? they consider it stable 18:11 Calinou ;) 18:11 Taoki I use openSUSE. And no, no bad words about it L:P 18:11 hmmmm even if they don't exist for some reason, you can add floatlands by making a derived class of MapgenV7 and implementing the virtual function generateExperimental() 18:11 hmmmm in much the same manner proller's indev does 18:12 hmmmm everything is made to be super easy and extensible 18:23 RealBadAngel i love changing the code and wondering why the heck it doesnt work. then realize that i was changing files in one directory, and launching game in another... 18:23 hmmmm that was pretty much all last night before i realized kaeza's changes screwed up run_in_place 18:24 hmmmm :/ 18:24 RealBadAngel looks like i have now detecting luajit workin. 18:24 PilzAdam has anyone a working LuaJIt 2.0 libary for windows hanging arround? 18:24 hmmmm i wouldn'tve caught that as fast as i did if i weren't developing on 0.4.5 and just rebased to 0.4.6 18:27 RealBadAngel PilzAdam, http://scilua.org/get.html 18:29 PilzAdam there is no libary 18:29 RealBadAngel indeed :( 18:30 RealBadAngel and you wanted to use system wide libraries, when folks cant even download it for windows? :) 18:31 RealBadAngel i found 2.0 beta, and some link on piratebay lol 18:31 RealBadAngel http://thepiratebay.se/torrent/7819091/LuaJIT_2.0.0_final_binaries_for_Windows_2012-11-08 18:32 hmmmm that's pretty crazy that you'd have to resort to TPB for compiled versions of some library 18:32 hmmmm but in windowsworld this is the norm because everything is so backwards 18:32 PilzAdam RealBadAngel, I dont gonna download it from this site 18:34 RealBadAngel theres bat file to compile it on windows 18:35 RealBadAngel /src/msvcbuild.bat 18:37 hmmmm i found a minor bug in the ore generation 18:37 sapier hmmmm do you intend to do bigger changes in emerge manager? 18:37 hmmmm sapier, i don't 18:38 PilzAdam salamanderrake, could you please stop spamming with new pull reqeusts? 18:39 sapier ok then I'll issue a pull request on emerge managers destructor too 18:39 RealBadAngel so folks, do we want luajit to be external or bundled? 18:39 PilzAdam external 18:39 hmmmm it's supposed to be external 18:40 hmmmm sapier you have a lot of pull requests :p 18:40 hmmmm what order do i need to do this in? 18:40 RealBadAngel so i think its done, need just more folks to test it 18:40 Taoki PilzAdam: Any news on the view falling code and when we can finish it? 18:41 sapier you wanted separate commits ;-P 18:41 RealBadAngel i will just squash it and pull 18:41 sapier so you can do cherry picking from only cherrys ;-P 18:41 hmmmm alright 18:41 hmmmm so i can do it in any order 18:41 hmmmm hmm 18:42 hmmmm i should probably put off the most recent ore modification then 18:42 sapier btw I'm far from finished ;) 18:43 hmmmm hm 18:43 hmmmm the original ore code had this problem too 18:43 hmmmm volume is calculated with all the chunk min/max, not the calculated bounding box of the ore 18:51 RealBadAngel ok, i pulled luajit related change 18:51 RealBadAngel hope it will work for everyone 18:52 hmmmm i'm expecting problems 18:52 RealBadAngel like? 18:52 hmmmm things we didn't expect to happen... people having rare problems with linking for some odd reason, etc. 18:53 hmmmm you know it always happens 18:53 RealBadAngel :) indeed 18:53 RealBadAngel i thought bundled will work 18:53 RealBadAngel and it did for me and kaeza, others reported strange and weird errors 18:53 RealBadAngel like openal crashes 18:54 hmmmm openal crashes? 18:55 RealBadAngel yup, after compiling bundled jit, openal refused to work 18:55 RealBadAngel completely unrelated :) 18:55 hmmmm are you referring to the crash on startup on windows that has an address in the 0x62-69XXXXX range? 18:55 hmmmm oh, nevermind 18:55 RealBadAngel no, not on windows 18:55 sapier openal isn't quite clean at either I found some problems with valgrind on linux too 18:55 RealBadAngel on linux 18:56 RealBadAngel but only change was bundled jit 18:56 RealBadAngel vanessaE had some weird assertion errors and couldnt even complie 18:57 RealBadAngel but here it worked flawlessly 19:11 hmmmm grr 19:11 hmmmm have to do stuff with family 19:11 hmmmm i'll merge those things when i get back, sorry guys 20:11 salamanderrake umm I'm not PilzAdam 20:11 PilzAdam oh, sorry, I meant sapier 20:11 salamanderrake PilzAdam, I git pulled once for each project and that was it. 20:11 salamanderrake o ok 20:12 salamanderrake I was worried for a min there that someone hacked my account. 20:12 PilzAdam stupid nick completion :-) 20:13 salamanderrake yah I get in trouble with that some times. 20:14 sapier what did you mean PilzAdam? 20:15 PilzAdam you have 19 (!) pull requests for you memory leak fixes 20:15 sapier yes .... hmmm didn't want the big one it has been before 20:16 PilzAdam why do you create a pull request for each one of it? 20:16 sapier and if you have a look at them they're all valid bugs 20:16 PilzAdam just point hmmmm to the branches on your repo 20:16 sapier no just for leaks in different components 20:17 sapier no I hope for some educational effect if ppl see how much memory they've lost 20:17 sapier so I'm not hiding bugs 20:27 sapier ok done splitting big memory/texture/mesh/sound leak patch ... now you can merge or drop those pull requests it's your decision 20:28 sapier I hope i didn't miss something included in the big one 20:28 VanessaE whatever you miss now can be looked for later. 20:28 VanessaE that's what valgrind's for :) 20:29 sapier obviously noone uses valgrind 20:29 sapier but complains about too much memory leaks beeing fixed 20:29 VanessaE heh 20:30 sapier sorry ... you're the last person to be guilty about this point ;-) 20:31 sapier I can't even blame someone for not using valgrind with that many leaks in you couldn't even find a single error between all those minor ones 20:31 VanessaE it's okay, I have no purpose to run that util anyway since I don't regularly code in C/C++ 20:32 Taoki Is there a function to get the position of a node? I searched but can't find it anywhere... 20:33 sapier don't you normaly only have a reference to a node definition? 20:33 sapier so a node doesn't even have a position at all 20:33 sapier entities have positions 20:34 Taoki Ah... I understand. Yeah wanted from entity node definition 20:34 Taoki or ID or what that is 20:34 sapier node definition ... you got it by get_node? ... a little bit misleading this name 20:35 sapier but if you did get_node you had the position? 20:39 proller good idea for crazy mapgen http://www.youtube.com/watch?v=bO9ugnn8DbE 20:46 Taoki Uh... why is it that using on_place and on_dig causes nodes to be impossible to place and dig? 20:48 sapier ? sorry don't understand 20:49 Taoki When using on_place and on_dig in register_item, items can no longer be placed or digged 20:50 sapier oh .... you may have to specify correct return value 20:50 Taoki How does that work? 20:51 sapier return false ... but I'm not quite sure about that but it's worth a try 20:51 sapier should be written in lua_api.txt ... but I'm not sure about that either ;-) sorry 20:52 Taoki Checked and didn't find much info 20:53 sapier minetest.register_on_placenode(func(pos, newnode, placer, oldnode, itemstack)) 20:53 sapier ^ Called when a node has been placed 20:53 sapier ^ If return true no item is taken from itemstack 20:53 Taoki Both returning true and false causes crash 20:54 sapier so you've done anything else wrong ;-) 20:54 Taoki no, not minetest.register_on_placenode 20:54 Taoki on_place and on_dig. They're item definition properties 20:55 sapier ohhh ... you have to call minetest.item_place within your function to get standard behaviour 20:55 sapier passing same/modified by your function parameters to it 20:56 Taoki That is confusing 20:56 sapier if you don't specify anything minetest.item_place is called 20:56 sapier if you specify a specific function you need to call this function from your function 23:45 BlockMen celeron55, and who ever wants know: template for dokuwiki is almost done: https://dl.dropbox.com/s/w7i3639nza56nl1/dokuwiki_prev.png 23:50 RealBadAngel that looks nice