Time Nick Message 01:25 hmmmm hrmmmmmmmmmmmmm 01:26 hmmmm what if i change the biome terrain type into a generic cave terrain type and caves are generated in the same way that the nether is, except the scale is much much smaller 01:26 hmmmm i wonder if that'd make things look alright 01:27 hmmmm i'd still have to figure out how to do the tunnels 01:27 hmmmm change the nether biome terrain type* 11:39 sfan5 celeron55: it would be helpful if you could give users on the forum warnings 12:04 thexyz celeron55: celeron56: https://github.com/celeron55/minetest/pull/344 16:03 hmmmm alright, ndef->getId("mapgen_gravel") is returning 127 when i try to register a biome. i currently call register_biome in between all the minetest.register_alias("mapgen_whatever", "default:whatever") and the common functions in games/minimal/mods/default/init.lua) 16:04 hmmmm apparently, that's *too* early 16:04 VanessaE perhaps biomes should be in a separate mod 16:04 VanessaE instead of being part of default 16:04 hmmmm lol... 16:04 hmmmm i don't think so, unless you really want water to be a mod 16:05 hmmmm are oceans optional for you? 16:05 VanessaE well, I mean all the default biomes collected into one. 16:05 VanessaE that said, 16:05 VanessaE think of Cornernote's "skyblock" mod/game. that has no land generator at all, except for several very small islands. 16:05 hmmmm so what i want to know is at what point the nodes i want to get a content_t to are present in the nodedef manager 16:05 VanessaE someone else, i forget who (mauvebic maybe?) was looking at making an all-water world. 16:06 hmmmm hmmm, how do you even do that from a mod.... do they just replace everything with air in mapgen.lua:register_on_generated ? 16:06 VanessaE not sure how he did it. 16:06 VanessaE a custom game, for sure 16:06 hmmmm grotesque 16:07 VanessaE case in point :-) 16:07 hmmmm i don't get it though 16:07 hmmmm gravel is clearly being registered in the same file far above this 16:28 thexyz sfan5: did you have a problem with buildbot? https://github.com/celeron55/minetest/pull/344 16:29 hmmmm WTF.... 16:29 hmmmm i am getting SIGILL when i try to printf m_name_id_mapping_with_aliases.size() in CNodeDefManager::getId() 16:30 hmmmm this is ridiculously difficult to debug 16:30 hmmmm hey, at least breakpoints sorta work, right? :) 16:31 hmmmm it's not the worst 16:33 hmmmm alright, so i got it working..... the size of m_name_id_mapping_with_aliases.size() is 44 at the time i want to create my node 16:35 hmmmm it's like the aliases just aren't being added..... but they _clearly_ are, directly above this call! 16:35 hmmmm i just figured i should try default:graven instead of mapgen_gravel and it works 16:36 VanessaE making graven images in minetest? blasphemy. 16:36 VanessaE :-) 19:03 RevenantX hi all 19:03 RevenantX does minetest using batching for chunks? 19:44 hmmmm a chunk is defined as a batch of blocks 19:44 hmmmm :| 20:32 hmmmm guys, I don't think that the LGPL is working exactly the way it was intended as it is for libraries (minetest is clearly not a library) 20:32 hmmmm moreover, the lua scripts are defined as plug-ins, not libraries, as there is much interaction between the two 20:33 hmmmm since LGPL is the same text as the GPL but with the library exception added, i am pretty sure that the GPL policy on proprietary plugins still follows 20:33 hmmmm if you want proprietary Lua scripts to be allowed, it's probably a better idea to change the LGPL license to GPL, and then add in an exception for distributing with proprietary plugins 20:34 hmmmm otherwise (to me at least) it seems like in this regard the LGPL is acting exactly the same as the GPL would 20:34 celeron55 i perfectly know that 20:34 hmmmm (read http://www.gnu.org/licenses/gpl-faq.html#GPLAndPlugins) 20:34 hmmmm oh 20:34 hmmmm then what 20:34 celeron55 the reason for LGPL is that there is no reason to restrict people to less rights than LGPL gives 20:35 celeron55 i do not like the way the FOSS people restrict people as much as they can 20:36 celeron55 if somebody wants to make some part of minetest as a library and use it in a commercial project, GPL would make it not an option for no reason 20:36 hmmmm what don't you like about MIT or BSD with an added GPL-like clause for protection against closed-sourcing minetest? 20:36 celeron55 s/commercial/non-free/ 20:36 hmmmm i dunno, it just seems weird to me 20:37 celeron55 there are contributors who won't make their code anything more free than LGPL 20:37 hmmmm oh 20:37 celeron55 minetest was originally published under GPL and development got somewhat close to dying, so i took in code from some GPL contributors 20:38 celeron55 it 20:38 celeron55 that is the history of the license 20:38 celeron55 then i managed to get an agreement on LGPL later 20:41 celeron55 that being said, i think being so loose about contributions was kind of a dumb decision; but everybody is dumb before they get wiser 20:41 celeron55 8) 20:43 hmmmm yes, on that topic... some 'contributions' might be unwanted 20:43 hmmmm in #minetest there was a conversation about one guy who was luac'ing his mods 20:43 thexyz he was going to do it 20:44 thexyz just /ignore NekoGloop, everything will be fine 20:44 hmmmm i don't think there's anything in violation with the license with that, but it goes against the spirit of freedom 20:44 thexyz sure 20:44 hmmmm not only that, but there's a potential security risk 20:44 celeron55 i couldn't care less; the only thing i will do is restrict publishing via channels maintained by me 20:44 hmmmm without being to peer review the script, you don't know for sure if it's going to try to exploit some vulnerability in the lua api 20:45 thexyz why is there need to exploit something in lua api? 20:45 hmmmm exactly, which is why i said that if it's not being done already, luacs shouldn't be distributed officially in any sort or form 20:45 hmmmm thexyz, to break out of the sandbox, obviously 20:45 thexyz just >os.execute 20:45 thexyz um.. 20:45 hmmmm you can do that? 20:45 thexyz is there sandbox? 20:45 hmmmm i thought lua was sandboxed 20:45 hmmmm that'd be pretty awful 20:46 hmmmm so there's an even WAY BETTER reason to not allow luac 20:46 celeron55 minetest does not have any sandboxing for lua, by design 20:46 thexyz hmmmm: what would change? 20:47 thexyz btw, celeron55, can you please review this one https://github.com/celeron55/minetest/pull/344 ? 20:47 celeron55 very strict sandboxing will be done on client-side if any code sent by the server will be ran there some day, but currently that is not done 20:47 celeron55 thexyz: not today 20:47 thexyz ok 20:47 thexyz then i'll ask again tomorrow 20:48 hmmmm thexyz, it's allow the community to peer review the mods and know if there's something up with them 20:48 hmmmm even with sandboxing, that's not perfect 20:49 celeron55 having the code of mods exposed is a very good thing in many many ways 20:51 celeron55 (so, yes, moderators should not allow any obfuscated or compiled mods to Mod Releases, unless there is a very good reason for them to be so) 20:52 hmmmm that's not only NekoGloop but also that mauvebic guy 20:52 celeron55 (actually, the only allowed reason is compiling for performance reasons, and even then the source code should be available) 20:55 thexyz luac doesn't give any perfomance increase except startup time 20:56 celeron55 maybe if somebody pays enough to cover reviewing of the source by a trusted person and enough to support the community equally compared to released source code, i could consider allowing it 20:56 hmmmm not really sure if that's because lua is just slow no matter what, or if lua just isn't really eating up that much cpu in the first place.. has anyone investigated that? 20:56 thexyz >Pre-compiling does not imply faster execution because in Lua chunks are always compiled into bytecodes before being executed. luac simply allows those bytecodes to be saved in a file for later execution. 20:57 thexyz according to `man luac` 21:06 hmmmm oh mm i had to ask... how do i read an array in from Lua 21:07 hmmmm for (int i = 1; ; i++) { lua_rawgeti(L, index, i); foo[i - 1] = lua_tonumber(L, -1); lua_pop(L, 1); } 21:07 hmmmm what is the condition for end-of-table here? lua_rawgeti does not return a status code of any sort 21:09 celeron55 use lua_next(), search that in google and in scriptapi.cpp 21:09 hmmmm ok 21:10 celeron55 scriptapi.cpp:1185 for example reads a list 21:11 celeron55 (that specific case reas up to 6 values) 21:12 celeron55 +d 21:36 hmmmm what is the preferred way to log problems with registering things in lua?