Time Nick Message 03:08 VanessaE ok I'm arguing and arguing here and I'm lost. 03:09 VanessaE http://pastebin.ubuntu.com/7413811/ 03:09 VanessaE why doesn't this code work? The particles appear, I can resize them, change their texture, etc. but they just absolutely REFUSE to move. 03:09 VanessaE it's as if the velocity parameter is being totally ignored. 03:10 VanessaE (I'm trying to adapt snowdrift mod partly to use non-deprecated API calls so I can use the vertical param. that param, btw, also seems to work) 03:17 VanessaE ah HAH 03:17 VanessaE the API 03:17 VanessaE https://github.com/minetest/minetest/blob/b5acec0a3c5701c53854ff7afdf4008863e6e8df/src/script/lua_api/l_particles.cpp#L67 03:17 VanessaE the API says these should be "velocity" and "acceleration". changing them to just "vel" and "acc" as in the C++ code makes things work. 03:19 VanessaE was something that does just a substring match actually wanted here, or is strcmp() intentional? 03:30 VanessaE oh well, bug report filed. 03:30 VanessaE https://github.com/minetest/minetest/issues/1280 03:33 hmmmm oh my god 03:33 hmmmm that code is horrible 03:33 hmmmm who wrote that 03:33 VanessaE not me 03:33 VanessaE :P 03:34 hmmmm if you'd like to fix it, remove that entire loop that goes through the table and in place, put in get*field_default for each field 03:34 hmmmm and then do the same for l_add_particlespawner 03:35 VanessaE no need to "fix" it per se. it'd be easier to just tweak lua_api.txt to match what the C++ calls for 03:37 hmmmm yes, that would be easier, but that shitty code would stay shitty 03:38 VanessaE if I try to fix it, it'll be worse :) 03:38 hmmmm lol 03:45 hmmmm in fact you know what 03:45 hmmmm I noticed that all particlespawner related code is of extremely poor quality 03:45 hmmmm gotta wonder how it made it into the engine in the first place 04:29 VanessaE Suggestion: Some kind of profanity database that mods can consult when implementing filters in family-safe situations 04:29 VanessaE already a mod exists that does this for player names, now someone's about to submit a pull request for an optional, similar filter for signs_lib 04:30 VanessaE might be a good idea to have a central place to look for that sorta thing. less duplication etc. 04:30 VanessaE perhaps with three or four levels of profane-ness. 04:31 VanessaE (well maybe not more than 3) 07:48 kaeza VanessaE, that could be done by a "library" mod (a mod that does not offer any blocks/items, but only an API) 07:49 kaeza and then all mods that want the functionality just (opt)depend on it 08:45 Calinou since you defined statbars in builtin, could you also add empty heart/bubble textures and use them? 08:45 Calinou (to see how much health/bubbles you miss) 12:31 Guest41677 Hi all 14:15 CraigyDavi Experiencing an issue: https://github.com/minetest/minetest/issues/1281 14:55 RealBadAngel hi folks, have anybody checked #1278? 14:55 ShadowBot https://github.com/minetest/minetest/issues/1278 -- Map irrlicht log level to minetest. Allow write irrlicht logs to debug file. by RealBadAngel 15:10 BlockMen RealBadAngel, it does write to minetest log only with lvl4? 15:15 celeron55 i asked freeminer's standing on the namespace issue from xyz, and it is this: 1) freeminer doesn't have plans to break api compatibility, 2) it already has an alias of the "minetest" namespace as "freeminer", 3) it will add an alias for "core" if minetest switches to it, 4) xyz seems to agree that if all forks have their own namespaces and everyone has to support everyone else's namespace, it's stupid 15:28 crazyR_ has anyone got any info on how this is to be used minetest.swap_node is it just pos and node name that are required? 15:36 ShadowNinja crazyR_: minetest.swap_node(pos, {name="node:name", param1=0, param2=0}) 15:36 ShadowNinja crazyR_: The params are optional.. 15:36 ShadowNinja -. 15:36 crazyR_ :D thanks. 15:37 ShadowNinja YW. :-) 15:37 PenguinDad RealBadAngel: #1278 works without problems here 15:37 ShadowBot https://github.com/minetest/minetest/issues/1278 -- Map irrlicht log level to minetest. Allow write irrlicht logs to debug file. by RealBadAngel 15:45 BlockMen celeron55, what are the reasons again to use core instead of engine (except its 2 chars shorter)? 15:46 Anchakor1 it's shorter 15:47 BlockMen Anchakor1, "... (except its 2 chars shorter)" 15:49 RealBadAngel BlockMen, irrlicht log levels are the same as mt ones. So setting mt log level means setting engine one. If mt action level is logged so is irrlicht one 15:49 RealBadAngel theyre just defined in different order 15:50 RealBadAngel thats why mapping 15:51 BlockMen RealBadAngel, ok. then im fine with #1278 15:51 BlockMen (but havent tested) 15:51 ShadowBot https://github.com/minetest/minetest/issues/1278 -- Map irrlicht log level to minetest. Allow write irrlicht logs to debug file. by RealBadAngel 15:51 RealBadAngel code is damn simple for so many features 15:52 RealBadAngel btw i lowered default log level in terminal there 15:53 RealBadAngel so no more spam from player dug something and irrlicht loaded a mesh 15:53 RealBadAngel but you can set action log level back using command line "action" switch 15:54 RealBadAngel side effect of the patch: mt tries twice to load textures with no name given, and third is not valid texture format 15:55 BlockMen RealBadAngel, where? i dont see the changed default value 15:55 RealBadAngel default was action, i set default to error 15:55 BlockMen oh, #750 15:55 BlockMen *L 15:55 ShadowBot https://github.com/minetest/minetest/issues/750 -- Delay deleting replaced textures by kahrl 15:55 BlockMen ah, too late^^ 15:55 RealBadAngel and added action switch which were completely missing 15:58 RealBadAngel so before setting error log level was just impossible 15:59 RealBadAngel anyway, code is damn simple and should be self explanatory, not to mention that solves two issues 16:01 RealBadAngel but the texture errors i mentioned above... some1 should set log level to debug and see where they come from 16:02 RealBadAngel requesting textures named "" is definitely a bug 16:16 celeron55 official decision: 16:16 celeron55 minetest will use the minetest namespace for now and until more forks appear 16:16 celeron55 however, if or when that happens, it will be changed 16:17 celeron55 so if you make a fork, please use the "core" namespace because that's where everything will go then 16:27 proller celeron55, maybe add alias core=minetest ? 16:29 RealBadAngel celeron55, thats wise decision imho 16:29 RealBadAngel and no for the aliases atm 16:32 ShadowNinja celeron55: So, should minetest use 'core' only as a implementation detail? Because using one name allows me to clean up part of the codebase. 16:33 celeron55 proller: if freeminer or some other compatible fork starts primarily using it already, then maybe 16:33 ShadowNinja Otherwise Lua files use local core = minetest or engine and the like, and everything that creates a Lua stack has to create it's table seperately. 16:34 celeron55 ShadowNinja: i'm not sure what makes most sense there... maybe switching it internally to core causes least conflicts in the future 16:51 RealBadAngel ShadowNinja, celeron55 opinions on #1278? 16:51 ShadowBot https://github.com/minetest/minetest/issues/1278 -- Map irrlicht log level to minetest. Allow write irrlicht logs to debug file. by RealBadAngel 16:59 RealBadAngel PilzAdam, btw, since when theres minetest.swap_node aviable can i modify minetest_game screwdriver to use it? 18:00 * RealBadAngel is about to comment chan's activity: 18:00 * RealBadAngel yawns 18:26 celeron55 RealBadAngel: i don't know why to do that or why not to 18:27 celeron55 why do you change the default console log level? it doesn't really seem related and i don't think it causes any problems as how it is now 18:28 RealBadAngel celeron55, look closer, action is default, and higher priority than error 18:28 RealBadAngel where was the way to use error just? 18:29 RealBadAngel most users dont want to be spammed by action messages, and they are now 18:29 RealBadAngel so the servers 18:29 celeron55 can you point me to a source of that fact 18:30 celeron55 i have never seen anyone say anything like that 18:30 RealBadAngel Megaf, requesting way to shut it up, here on #dev 18:30 RealBadAngel in fact i second it 18:31 RealBadAngel there was no way before to set error log level 18:31 RealBadAngel so i made it default, with minimal terminal output 18:32 RealBadAngel and added action switch 18:32 celeron55 there is the option of instead adding --error that would do it so that the default behavior is same as before 18:32 celeron55 i'll wait for opinions of other people 18:32 celeron55 (and, while at it, maybe --quiet too) 18:33 RealBadAngel not really 18:33 RealBadAngel look at irrlicht log levels 18:33 RealBadAngel theyre almost the same 18:33 RealBadAngel and at this very level are hell annoying 18:33 celeron55 i don't care what irrlicht does 18:34 RealBadAngel reporting loading of meshes 18:34 celeron55 if irrlicht is annoying, then map it differently to minetest's levels 18:34 RealBadAngel you dont? so why let you spam the terminal wiht it? 18:34 celeron55 minetest doesn't need to try to dance along with irrlicht's levels 18:34 celeron55 put a gap there at the action level or something; people aren't interested in irrlicht's output anyway 18:34 RealBadAngel we do share same amount of levels, named the same 18:35 RealBadAngel just the order is different 18:35 celeron55 the fact that they have the same names don't mean that they mean the same thing 18:35 RealBadAngel yes they do 18:35 celeron55 or that irrlicht tries to output the same amount of stuff 18:35 RealBadAngel read the docs 18:35 celeron55 you need to match them by the amount of output, not by the names 18:35 celeron55 if irrlicht outputs too much, then you have mapped the levels wrong 18:36 RealBadAngel hold on, sir, you want to make apps generate same ammout of characters? cmon :P 18:36 celeron55 that's what you are complaining about 18:36 celeron55 and the fix is to map the levels differently 18:36 RealBadAngel i mean level of flooding user 18:36 celeron55 not to change minetest's behavior 18:37 RealBadAngel please stop being stubborn, and just test the patch with some log levels 18:37 celeron55 i'm objecting this as it is, assuming what you say is correct 18:37 RealBadAngel and compare it to current behaviour 18:38 RealBadAngel patch is easy enough, you can compare log levels too 18:39 RealBadAngel and the fact they can be controlled with single setting 18:40 RealBadAngel if thats not enough, dont expect anybody to write parser to translate irrlicht logs to mt ones. 18:40 celeron55 i like seeing what is happening in the game by default in the console 18:41 celeron55 i get a constant flood of unnecessary stuff to console with this patch 18:41 RealBadAngel what? 18:41 RealBadAngel you must be fuckin jokin now 18:41 celeron55 and don't see the things that i want to see 18:42 celeron55 unless i add --action to the command line every time i run minetest 18:42 RealBadAngel in error log level, which you have obviously FORGOT to implement to set up 18:42 RealBadAngel you can see only error logs, or print messages 18:43 RealBadAngel in Action, which is default atm, user is flooded with celeron55 placed tome_of_wisdom at (0,0,0) 18:44 celeron55 on the console this doesn't change anything for me except that i have to supply --action now to minetest, and the irrlicht flood is prefixed with "Irrlicht log: " (before it had no prefix) 18:44 celeron55 RealBadAngel: that's not a flood for me, only a single message now and then 18:44 RealBadAngel and meanwhile all the irrlicht engine log entries reporting "i managed to load a mesh, yikes!!!" 18:45 celeron55 why doesn't this hide the useless irrlicht log entries if it's mapping the levels? 18:45 RealBadAngel celeron55, but this is flood for folks out there, including me. and now theres no way to turn it off 18:46 celeron55 i'm fine with being able to turn it off; what i am not fine with is trying to bundle this with the irrlicht log redirect and level mapping 18:46 celeron55 they are unrelated and now this discussion is just messy 18:46 celeron55 pull them apart and we'll talk about them individually 18:47 RealBadAngel a) i cant and i wont make parser for logs, that would be just ridiculous 18:49 RealBadAngel b) i think i can separate it, but its not what sapier wanted me to code 18:49 RealBadAngel he wanted complex and easy solution 18:49 RealBadAngel so i somehow managed to do it 18:51 RealBadAngel if you want default settings be same as before, and let user decide to set another (lower) log level, patch will become more complicated 18:52 RealBadAngel theres no method to delete stream, just to add 18:52 celeron55 separate them anyway, then it's easier to refer to them for discussing about them 18:52 celeron55 currently when i try to talk about the other aspect, you start talking about the other one and it's not fun 18:53 RealBadAngel ok, lets calm down 18:53 RealBadAngel so what we shall do? on which points we do agree? 18:53 RealBadAngel mapping is ok with you? 18:57 celeron55 mapping doesn't seem to do anything as far as i can see 18:57 RealBadAngel it works flawlessy 18:57 celeron55 it flawlessly does nothing? 8) 18:58 RealBadAngel well, thx to it i have found out that engine calls twice for "" named textures 18:59 RealBadAngel set debug_log_level in conf 18:59 RealBadAngel and try 19:00 RealBadAngel btw, not my fault that you have made switches for terminal log levels (console is your idea, btw?) and saving to file log leves 19:01 celeron55 wot 19:02 RealBadAngel command line switches do affect only displaying in terminal (not so called "console" in the game) 19:02 RealBadAngel config setting do affect logging to file 19:03 celeron55 bleh, i'm too tired; figure it out with someone else 19:04 RealBadAngel ;) 19:05 RealBadAngel i will make the patch with action as default and add --error command like switch, will it be ok with you then? 19:05 RealBadAngel simple as that 19:06 celeron55 don't change it yet 19:06 celeron55 (i already said) i want to hear other people's opinions on that 19:07 RealBadAngel ok, thats not a big change, nor require lots of time to code 19:08 RealBadAngel im open for suggestions 19:15 pitriss celeron55: btw: I asked too to be able stop that action spam here in dev:) 19:18 VanessaE personally I see no need for anything but minetest errors and debug info to be dumped to the terminal (e.g. in the case of a mod crash). In a production environment, (read: a multi-player server where there isn't even a controlling terminal normally active to begin with), spewing lots of "player did X" to stdout is a waste of CPU cycles. Similarly, dumping anything but critical errors to the clients' terminals is a waste. 19:19 VanessaE log those things if you want, but dumping them to the terminal is pointless. 19:20 VanessaE I guess that would imply having neither such message active by default, and both --error and --action as command line switches, or having both ON by default, and --no-error/--no-action switches to turn them off. 19:21 pitriss VanessaE: +1 19:22 VanessaE in some programs, log levels are done as bitfields rather than some numeric level, so that individual message categories can be turned on/off. 19:23 VanessaE maybe that is warranted here, supposing the log level were set higher than 4 anyway. 19:24 VanessaE but that would require a total rewrite of the entire logging system I guess 19:25 RealBadAngel but since Irrlicht already is using levels, we are using levels, named the same, just numbered in reverse order... why just make it the very same way? 19:25 RealBadAngel *why not 19:25 VanessaE brb 19:30 VanessaE ok 19:32 RealBadAngel primo: we cannot expect irrlicht engine will follow our fantasies. so core is as is 19:32 RealBadAngel we can play with it or not 19:33 RealBadAngel but having two differnt log systems in one app? 19:37 Calinou could ACTION[ServerThread] be renamed to ACTION:? 19:37 Calinou (shorter) 19:40 RealBadAngel propably it could, but you know which thread sent it 21:21 sapier Calinou renaming "ACTION[ServerThread]" to Action may sound like a reasonable suggestion on first glance, but is rather useless if you look into detail, [ServerThread] is not a hardcoded value but the name of the thread doing the log. Especially on looking for threading issues that information is crucial 21:54 ShadowNinja ServerThread could be shortened to Server. It would be less clear though. 21:55 sapier Well you're free to do so but there are other Threads with even longer names too 23:25 SoniEx2 ShadowNinja, wanna help me over at #lua ? 23:25 SoniEx2 (if you're there...) 23:26 ShadowNinja SoniEx2: What with? 23:27 SoniEx2 ShadowNinja, well idk... bug them about a proper string metatable sandbox? 23:27 SoniEx2 (even tho I think the proper way to do that would be thru the mailing list) 23:28 ShadowNinja SoniEx2: I don't think you're going to get anyone to change that unless you do it yourself. And regular indexing isn't allowed, so sandboxes are O.K. 23:28 ShadowNinja SoniEx2: I'll write you a sane version of what you posted. 23:28 SoniEx2 ShadowNinja, regular indexing? 23:31 SoniEx2 ShadowNinja, what does regular indexing mean? 23:40 ShadowNinja SoniEx2: Here's a sane version that only does __index (probably all you need) (for Lua 5.2, just use setfenv instead of _ENV for Lua 5.1): http://codepad.org/NBMuK1s1 The "Output" part is incorrect as that site apparently uses Lua 5.1, where _ENV is meaningless. 23:41 SoniEx2 ShadowNinja, the output is right 23:41 SoniEx2 the function was declared in a different env 23:41 ShadowNinja SoniEx2: Run it in Lua 5.2 and see for yourself. 23:42 SoniEx2 oh and you forgot to do "local _ENV" 23:42 ShadowNinja Also: -> PM. This is off-topic. 23:42 SoniEx2 ok