Time  Nick         Message
00:10 ThreePac     hello
12:24 kahrl        Pull Request of the day: #645 Math fractal mapgen with external lib (10+ fractals) -- https://github.com/minetest/minetest/pull/645
12:24 kahrl        I think either this or https://github.com/minetest/minetest/pull/659 should be closed
12:26 kahrl        assuming the technical issues get fixed (unrelated commits, merge commits, code style differences, json in minetest.conf) should the math mapgen be added to core?
17:54 aracnus      Hi! I'm needing a information about translating Minetest. This is the right channel?
17:55 Calinou      yes
17:55 Calinou      there's a weblate, aracnus
17:55 Calinou      http://translate.minetest.ru/widgets/minetest/
17:55 PilzAdam     aracnus, http://translate.minetest.ru/projects/minetest/core/
17:55 aracnus      Yes, I used it to translate Minetest to pt-BR.
17:55 aracnus      But I didn't find any items there.
17:55 Calinou      :-O some lines in french are not translated, gonna translate them.
17:56 aracnus      (tools, stones, weapons, etc...)
17:56 aracnus      The interface is 100% translated, but I would like to translate itens too. Is it possible?
17:56 PilzAdam     aracnus, yes, you cant translate them currently
17:57 Calinou      with a mod, you can do that but it is server-side
17:58 aracnus      That's something. I want to include Minetest on our educational distribution and it's very important that it is translated.
17:58 aracnus      It will be use by children.  :-)
17:58 aracnus      And what module, Calinou?
17:58 PilzAdam     you can directly go into the init.lua of mods and translate it there
17:59 Calinou      no, use intllib
17:59 Calinou      http://forum.minetest.net/viewtopic.php?id=4929
17:59 Calinou      it isn't in any module
17:59 Calinou      you use .txt files to translate
17:59 Calinou      but again, it is server-side
17:59 Calinou      so it only works on singleplayer pretty much
18:00 sfan5        "yes, you cant translate them currently" <- what?
18:00 aracnus      But if I mount a server and other people access it, they'll see the translation, Calinou?
18:00 L4ndM1ne     Calinou why did you kick me?
18:00 L4ndM1ne     lol
18:00 sfan5        aracnus: yes
18:00 Calinou      they will, but they will all see the same language
18:00 Calinou      L4ndM1ne: for being an idiot and insulting people
18:00 Calinou      how about you grow up?
18:01 L4ndM1ne     i said nothing retard
18:01 aracnus      Ok. That's what I want.  :-)
18:01 Calinou      <L4ndM1ne> newphag
18:01 L4ndM1ne     well he is
18:01 Jordach      Calinou, kick him
18:01 L4ndM1ne     and he started being smart first
18:01 aracnus      Thank you for the help, Calinou.  :-)
18:01 L4ndM1ne     jordach stfu albino fat fuck
18:01 Calinou      nope, I can't here
18:01 sfan5        aracnus: you could also use my localisation mod, it uses .po files
18:01 sfan5        and can automatically generate templates for them
18:02 aracnus      Hey, that would be great, sfan5. Where can I find it?
18:02 L4ndM1ne     calinou you really didnt do anyhting
18:02 sfan5        aracnus: https://dl.dropboxusercontent.com/u/30267315/Minetest/localisationmod-mt.tar
18:02 L4ndM1ne     im running on that HMA lol
18:02 L4ndM1ne     i can get back in in 1 min
18:02 L4ndM1ne     but i choose not too
18:02 PilzAdam     thexyz, can you throw me +o?
18:03 aracnus      Downloading it now. Thank you sfan5.
18:03 celeron55    ...too slow 8)
18:03 celeron55    only got myself identified by now
18:04 celeron55    thexyz: do you have privileges to give +o to people here?
18:04 celeron55    in chanserv
18:04 thexyz       yes
18:04 Calinou      you don't auto-identify? you can put your nickserv pass as server pass :P
18:04 celeron55    you could give to some active ones
18:04 thexyz       i've given +ov to PilzAdam and Calinou
18:04 thexyz       anybody else?
18:04 celeron55    Calinou: i don't see the point; i like anonymity
18:04 sfan5        aracnus: to use template-generating you need to create a <modfolder>/po directory and set a variable in init.lua, but make sure to set it to false after you're done
18:05 sfan5        aracnus: if you have questions ask me
18:05 sfan5        celeron55: /msg ChanServer access #minetest-dev list
18:05 sfan5        -er
18:07 aracnus      sfan5: can I contact you by e-mail if I have questions? Or do you prefer here on IRC?
18:08 sfan5        aracnus: I can respond on IRC instantly
18:08 sfan5        IRC's better
18:09 aracnus      Ok. I'll make some tests here and tomorrow maybe I have some results.
18:09 rubenwardy   thexyz: #minetest-mods
18:09 Jordach      why does anyone use minetest-mods
18:10 PilzAdam     rubenwardy, minetest-mods is not registered IIRC
18:10 thexyz       right
18:10 rubenwardy   ah
18:10 rubenwardy   so no kickban
18:11 PilzAdam     /ignore should do the job
18:12 * sfan5      just tested http://www.exploit-db.com/exploits/25444/ -> kernel crashed
18:12 Calinou      lol
18:13 sfan5        I shouldn't trust code that has "void sheep(uint32_t off) {" in it
18:13 hmmmm        ummm
18:13 hmmmm        do you realize what the point of that code is...
18:13 hmmmm        why would you run it
18:15 sfan5        because ....
18:15 sfan5        that code is a local root exploit
18:15 PilzAdam     you should always run code that you find in the internet
18:15 hmmmm        are you not running amd64?
18:16 * sfan5      runs Linux 3.8.0-19-generic #30-Ubuntu SMP Wed May 1 16:35:23 UTC 2013 x86_64 x86_64 x86_64 GNU/Linux
18:17 * Calinou    runs Linux calinou-desktop 3.8.0-19-generic #30-Ubuntu SMP Wed May 1 16:35:23 UTC 2013 x86_64 x86_64 x86_64 GNU/Linux
18:17 Calinou      heh, it's a bit different
18:18 sfan5        its exactly the same, you just didn't remove your hostname from the output
18:18 Calinou      ah
18:42 aracnus      sfan5: just downloaded your mod and untared it on mods/minetest. What do I do next?
18:42 sfan5        install all mods you want to have translated
18:42 sfan5        create a "po" folder in every mod folder
18:43 sfan5        set the appropiate language in init.lua of mods/minetest/localisation
18:43 sfan5        oh and make all mods depend on "localisation"
18:43 sfan5        also add "localisation" to games/common/mods/default/depends.txt
18:44 PilzAdam     common mods cant depend on normally installed mods
18:45 aracnus      Ok. How can I make all mods depend on "localisation"? (sorry but I'm very newbie on lua and minetest internals)
18:45 rubenwardy   sfan5: You could add localisation to common_mods in game.conf, and it will be loaded first.
18:45 rubenwardy   cant you?
18:45 sfan5        rubenwardy: you should be able to
18:45 sfan5        <rubenwardy> sfan5: You could add localisation to common_mods in game.conf, and it will be loaded first.
18:45 sfan5        ^ try this aracnus
18:46 rubenwardy   You have to put  "localisation" into games/common
18:46 rubenwardy   and "localisation" must be the first in the list of common mods
18:46 rubenwardy   ie:
18:46 rubenwardy   common_mods = localisation, default, bucket
18:47 aracnus      Ok, I'll try it (but, in fact, I have any mods installed yet). One more question, sfan5, I saw a "translations.txt" file on your package. I'm guessing it is the "default" items of the game. Where do I put it? On mods/minetest? And it is automatically regenerated?
18:48 sfan5        the translations.txt file is not important for you unless you want german translations
18:49 aracnus      No, I don't want it.  :-)  So, how do I generate default (not mod) items translation?
18:49 sfan5        have you did what rubenwardy said?
18:49 sfan5        *done
18:50 aracnus      Just a minute. I think I did something wrong here.
18:50 sfan5        ok
18:55 aracnus      All right. Did it. I have to put this set on every "game", right? (build, minimal, survival, minetest_game)
18:56 aracnus      (if I want to generate any of them, of course)
18:57 PilzAdam     aracnus, sfan5, can you move this discussion to a different channel?
18:57 sfan5        ok
18:57 aracnus      ok
20:18 sapier       https://github.com/minetest/minetest/pull/724 formspec based mainmenu feature complete and ready for final review
20:19 sfan5        thats a lot of commits
20:19 sapier       I've done small changes most of time ... some are even bugfixes only
20:20 sapier       as I mentioned I'm gonna squelch it once all remaining bugs are fixed
20:20 PilzAdam     sapier, doc of tables should be in table format
20:21 PilzAdam     i.e. {field1, field2} instead of table.field1; table.field2
20:22 sapier       sorry don't understand what do you mean?
20:22 PilzAdam     this https://github.com/minetest/minetest/pull/724/files#L2R26 should be like this:https://github.com/minetest/minetest/blob/master/doc/lua_api.txt#L1882
20:23 sapier       ok I'll change
20:23 PilzAdam     it should generally be more like lua-api.txt
20:24 sapier       I tried to use same style as lua-api.txt ;-)
20:25 sapier       what do you think is not lua-api.txt like except the tables descriptions?
20:25 PilzAdam     lemme change it to what I think it should be
20:26 sapier       of course do what you like it to be :-)
20:27 sapier       but plz don't forget to test functionality  ;-)
20:28 PilzAdam     game.buttonhandler is not described at all
20:28 sapier       it is
20:28 sapier       line 19 ... but event_handler is missing
20:30 PilzAdam     sapier, is selected_world =  0 needed for client mode?
20:31 sapier       I think so yes
20:31 sapier       indices are lua style
20:31 PilzAdam     it would be better if just adress ~= "" specifies if its a client or not
20:32 PilzAdam     are there some kind of default values for gamedata?
20:32 sapier       i did change as less as possible in main.cpp
20:32 sapier       no there aren't
20:34 sapier       I've added the missing event documentation
20:34 PilzAdam     is game.set_clouds(true/false) for the menu clouds?
20:34 sapier       yes
20:35 PilzAdam     why not just pass the gamedata to game.start()?
20:36 sapier       cause I would have to check what to set or not set in game.start then
20:37 sapier       main.cpp already does some plausibility checking I didn't want to add same thing over and over again
20:37 PilzAdam     whats the format of game.get_game()[gamemods_path]?
20:39 sapier       get_game returns a detailed game description as descriped right below function description
20:39 PilzAdam     is gamemods_path just a string?
20:40 sapier       oh you mean the field ... yes it's a string but I don't have any idea what this string means
20:40 sapier       it's just everything known about the game
20:46 PilzAdam     whats the format of get_worlds()?
20:47 sapier       just a string list to be used in world listboxes
20:47 PilzAdam     is it ["name"] = gameid?
20:47 sapier       no
20:48 sapier       it returns exactly what you see in textlist
20:48 sapier       at least atm
20:48 PilzAdam     "returns list of worldnames including gameid"
20:48 sapier       I think about removing get_game and get_world and replace get_worlds by a table of detail informations
20:48 MinerOfLands you know that is useless calinou
20:48 MinerOfLands i can just come back
20:48 MinerOfLands :D
20:48 sapier       yes "worldname <gameid>"
20:49 MinerOfLands Oh, come on. Everyone has thoughts like this; it's only fucked up if you actually do them. Like thinking up the perfect crime or the perfect murder. What's the most deliciously atrocious way way to torture or break someone? Take me: I imagine a father and his teenage daughter, preferably around age 14. Abduct and lock them in separate, sound-proofe
20:49 MinerOfLands d rooms. Tell the father that either he has sexual intercourse with his daughter - forcibly if necessary - or I torture her to death slowly in front of his eyes. I have given him sufficient pharmaceutical aid that his 'faculties' are unimpaired by his emotional state.
20:49 sapier       [ ] are not allowed for formspecs ;-)
20:50 sapier       this is one of the differences to current mainmenu I can't fix
20:54 PilzAdam     sapier, whats a listbox?
20:55 sapier       hmm the irrlicht name of a "textlist" field ... naming should be consistent ture
20:55 sapier       I'll fix this
20:58 PilzAdam     is game.buttonhandler() called when a button is pressed?
20:58 sapier       yes
20:59 PilzAdam     without params?
20:59 sapier       fields as in formspec handler but no form name
21:10 PilzAdam     sapier, https://github.com/PilzAdam/minetest/commits/next_gen_main_menu
21:10 PilzAdam     ummm, I suggest to not look at the diff but the actual file :-)
21:14 sapier       I've fixed the last "listbox" occurence and pushed it
21:15 sapier       thx for your help
21:22 PilzAdam     sapier, I already get merge confilcts when rebasing against master :-(
21:23 sapier       I'm not surprised rebasing multi commits did never work for me
21:23 sapier       first this needs to be squelched to a single commit
21:24 sapier       there are some reverted changes within this commit line too
21:24 PilzAdam     it acutally doesnt matter how many commits there are
21:24 PilzAdam     can you rebase it? Ill just fetch your branch again
21:24 sapier       don't know what exactly is the reason it's just what I realized
21:25 sapier       I'd prefere to have the fixes added prior squelching
21:25 PilzAdam     well, if you rebase it now then we wont push any commits to master that effect the code you want to change
21:26 PilzAdam     so a second rebase is not needed
21:26 sapier       it's almost midnight ;-) wouldn't be a good idea to start now
21:27 PilzAdam     its best working time now!
21:27 sapier       not if you have to work tomorrow ;-P
21:27 PilzAdam     hah! /me just finished his Abitur today
21:28 sapier       congratulations ;-)
21:28 PilzAdam     I have like 7 months free now :D
21:28 sapier       ok there's a full game.cpp to be rewritten ;-)
21:29 sapier       and of course doxygen documentation is missing almost everywhere ;-)
21:30 PilzAdam     oh god, that will just bloat the code
21:30 PilzAdam     its perfectly understandable as it is currently
21:30 sapier       you're kidding?
21:31 PilzAdam     no
21:32 sapier       sorry pilzadam you may have a lot to learn about documentation yet ;-)
21:33 sapier       documentation may seem to be additional work only ... but once you switch developers it's the only thing preventing the new ones from rewriting everything doing same mistakes again
21:38 hmmmm        doxygen comments bloat up the code
21:38 sapier       I don't deny that
21:38 hmmmm        i'd rather put all the comments in a separate place, or at least just document the things that really actually need documenting that aren't completely obvious
21:38 sapier       that's why they only are added to headers
21:38 sapier       lol
21:39 sapier       separate documentation ...
21:39 sapier       creating this is wasted work
21:39 hmmmm        i'm not opposed to it if you put it in the headers only
21:39 sapier       that's what I wanted to suggest
21:40 sapier       full doxygen function and parameter description in headers
21:40 PilzAdam     /* Sets the x value @param int the x value to be set @return nothing, since it only sets the x value */ void setx(int x);
21:41 sapier       usefull descriptions pilzadam
21:42 sapier       and yes for some functions this is redundant but those are normaly < 1/10th of all functions ... help provided by other 9 compensates additional work of those
21:43 PilzAdam     docuemntation in a different place would be actually quite useful
21:43 sapier       no it isn't
21:43 PilzAdam     like a web interface or so
21:43 sapier       separate documentation is always out of date
21:43 sapier       and out of date documentation is useless
21:44 PilzAdam     http://docs.oracle.com/javase/7/docs/api/
21:44 sapier       is done with something at least doxygenlike
21:44 PilzAdam     yea, but why need the docs in code _and_ in a web interface?
21:44 sapier       and java api is quite stable ... a thing you can't tell about minetest api
21:45 sapier       web doc is just rendered from code doc
21:45 PilzAdam     you just say people are too lazy/stupid to keep the doc up to date
21:45 sapier       they are but doxygen will tell you what's missing
21:46 sapier       you don't get any assistance if your doc is separated from code
21:48 sapier       it's just a suggestion ... currently there's no usable documentation at all once doxygen would be added doc could be automaticaly updated on each stable release
21:51 ShadowNinja  Doxygen documentation would be nice. Seeing the comment reminds you to update it.
21:51 kahrl        personally I noticed I don't gain much from doxygen-like comments
21:51 kahrl        that's not to say they shouldn't be added
21:51 kahrl        I gain more from stuff like this: http://c55.me/minetest2/wiki/doku.php?id=code:object_storage
21:51 sapier       yes and could be autogenerated on each commit to showing commits not following conventions
21:52 sapier       kahrl thats usefull too ... just it will be outdated and useless very soon ... and you can add comments like that wihin doxygen too
21:52 sapier       and of course noone will ever find this without knowing it's there
21:53 kahrl        well doxygen comments are more local than that I think
21:53 sapier       everyone can create the doc on the fly and stable versions doc could be added to minetest website
21:54 sapier       but of course I know about the amount of work to be done to get even near to finishing that task
21:54 kahrl        why stable version? most people who work on the core will start from current git
21:55 sapier       it's quite difficult to keep it actual if you don't want to update every day you have to define some intervals
21:55 kahrl        just a cronjob that generates it daily
21:55 kahrl        or would that add too much load?
21:55 sapier       could be done too of course
21:56 sapier       don't know how much load this would cause for full minetest documentation
21:57 sapier       but if it's already checked for each pull request it shouldn't be a problem to update on commit too
21:57 ShadowNinja  Every commit with a GitHub web hook and a limit.
21:57 ShadowNinja  So a miximum of say once a day.
21:58 sapier       yes but thats details if we really want to have this be aware that adding documentation (header only) is about 1 month of work (just a rough guess)
21:59 kahrl        I would say add documentation while rewriting old stuff. Plus some documentation sprints perhaps
22:00 sapier       forcing everyone adding new functions to add proper documentation prior merge ...
22:00 sapier       hmm I think I have to do some work :-)
22:00 kahrl        don't force it
22:00 sapier       if it's not mandatory it's not done
22:00 kahrl        doesn't make sense to doxygenize everything
22:01 sapier       doesn't make sense to have any doxygen docu if it's not complete
22:01 sapier       you'll lose lots of it's benefits
22:01 kahrl        I dunno
22:01 kahrl        there's a lot of simple function with arguments with self documenting names
22:02 kahrl        functions*
22:02 sapier       I know you can drive doxygen to insane levels but some minimum yet complete documentation is reducing overhead and is quite feasable
22:03 sapier       I don't know how often I've been anoyed by functions like setX() not telling me what that damn X is
22:03 PilzAdam     <sapier> doesn't make sense to have any doxygen docu if it's not complete <- this is exactly why I dont want to start with it
22:03 PilzAdam     we will have exactly this: /* Sets the x value @param int the x value to be set @return nothing, since it only sets the x value */ void setx(int x);
22:04 sapier       pilzadam it's just a matter of decision waste time over and over and over again searching what some function is doing or using a big bunch of time to fix it once for all times
22:04 kahrl        if you force people they will document stuff in the way PilzAdam showed
22:04 kahrl        and they will say "sets X" and not say what X is
22:04 sapier       yes but at least they will
22:05 kahrl        why is that good?
22:05 sapier       it's up to those ppl adding the commits to ensure a reasonable documentation ... if core developers don't support doxygenizing code it'll fail
22:06 ShadowNinja  Can't doxygen list those functions without having a comment?
22:06 sapier       no
22:06 ShadowNinja  Hmmm, that should be added.
22:06 kahrl        I support doxygenizing complex functions but not trivial ones
22:06 sapier       at least it will show an error for each of it ... having so much trivial errors you can't use the error log to find the real ones
22:07 ShadowNinja  Self-explanatory ones shouldn't require comments, but I think they should be listed.
22:07 sapier       it's same as memory leaks ... I had to fix most of it (lots of trivial ones) to finde the ones beeing realy critical
22:07 sapier       you often don't realize what "self-explanatory" is and what not
22:08 kahrl        if I don't realize I won't write good comments...
22:08 sapier       the often cited setX() for example
22:08 sapier       it's not that usefull you may think
22:09 ShadowNinja  Well that would be obvious in the context of a object in 3D space.
22:09 sapier       but at least you'll have to think about it increasing chance you'll realize by 100%
22:09 sapier       no it wouldn't shadowninja
22:09 ShadowNinja  (Although setPos() would be better)
22:10 sapier       setX may be X part of acceleration, velocity, coordinate, extent
22:10 ShadowNinja  Hmmm, well yes.
22:10 sapier       it doesn't tell if x is absolute or relative
22:11 ShadowNinja  Although IMO the function should just be renamed is that obscure.
22:11 sapier       once you think about it you realize there often is some information that specifies more exact what a function does
22:12 sapier       sometimes you prefere short function names to avoid having lines like:
22:12 kahrl        then someone changes it from absolute to relative and forgets to update the doxygen :P
22:13 sapier       bsdfbsdg.dsdds().hsadifghg.hsdigfhsaighaighdg.(gsadgusdu.(),ajbfjasd());
22:13 sapier       kahrl of course you can always do mistakes ;-)
22:14 ShadowNinja  Yes, some functions need documentation, but I don't think it should be mandatory.
22:14 sapier       still it's just a suggestion if you don't like it I won't bother adding doxygen comments
22:14 sapier       I still won't waste time if there's no chance to get some complete documentation anytime in future
22:15 kahrl        well I'm not really a core dev, just happen to have push access
22:15 kahrl        so my opinion doesn't really count
22:15 sapier       I don't even have push access ;-)
22:15 kahrl        maybe there should a poll for all core devs?
22:16 sapier       good idea
22:16 kahrl        s/should/should be/
22:16 sapier       hopefully developers don't only see the additional work required to do so
22:17 PilzAdam     sapier, no, we also see the file bloat
22:17 PilzAdam     ;-)
22:17 sapier       if you stopp writing 1k liner files there won't be a bloat ;-)
22:17 kahrl        actually I'm listed as a core dev so I guess I am ;)
22:18 sapier       and headers with >200 lines are crazy  adding 200 additional lines won't drive them even near to 1k
22:19 kahrl        huh
22:19 kahrl        C++ headers get larger than 200 lines quickly
22:19 sapier       but I know it's a lot of additional work and requires developers to kick their own asses to do it ... it's same to me
22:20 sapier       without documentation they don't ... if they do you're design most likely requires some rework
22:21 kahrl        pathfinder.h is an example :P
22:21 sapier       it's with complete documentation ;-)
22:22 kahrl        ok fair enough :)
22:22 sapier       and even with full documentation it's < 400 lines
22:22 sapier       and 100 lines are debug code too ;-)
22:24 sapier       and if you have a look at this file you may even realize worth of parameter documentation
22:24 kahrl        but for example craftdef.h is 400 lines without documentation already
22:25 kahrl        and I don't really see how it would be sanely made shorter
22:25 PilzAdam     remove all "\n"? :-p
22:26 sapier       craftdef.h is a collection of everything craft related ... you could argue about if this even should be all in same file
22:26 kahrl        /** default constructor */  /** copy constructor */  /** assignment operator */ <-- I don't see the point of those
22:26 sapier       makes sense once you have multiple constructors
22:26 kahrl        but you can see what they are from the prototype
22:27 sapier       no you can't
22:27 kahrl        err
22:27 sapier       e.g I don't have any idea what a IGameDef is
22:27 kahrl        what does that have to do with anything?
22:27 sapier       I just know it's a element for guiFormSpecMenu required for almost nothing
22:28 kahrl        if you have a class called Class and you see Class(const Class&); you know it's a default constructor
22:28 kahrl        copy*
22:28 sapier       there are about 3 formspec fields requireing this so it seams to be optional in most cases ... adding a second constructor not requireing this parameter would show this oprionality
22:29 sapier       it's just a matter of completeness
22:29 kahrl        this is a case where I feel making it complete just adds bloat
22:30 sapier       yes this is sometimes anoying ... but if you do something everytime chances you'll forget it will be less than if you happen to do it every once and then
22:30 kahrl        I guess we'll have to agree to disagree about this
22:30 sapier       have a look at my guiFormSpec cleanup ... especialy the parser functions ... header setting speaking names for parameters is completely redundant
22:30 sapier       still it's way better to understand
22:31 sapier       I even accept some performance penalty to do so as formspecs aren't generated thousands of times per second
22:32 sapier       but lets ask others too what they think about it