Time  Nick       Message
04:50 hmmmm      sort of wonder where i'd put the new things needed for the new setting type.  i have struct FlagDesc, readFlagString(), writeFlagString(), which are used by Settings::getFlagStr() and Settings::setFlagStr()
04:50 hmmmm      util/string.cpp? hmm
04:58 Exio       FlagDesc? what does it have?
05:07 hmmmm      struct FlagDesc { const char *name; u32 flag; }
05:47 Exio       i'd put it in the util/string.cpp and then ask c55 if he thinks it can be "better" in other file
05:47 Exio       but shouldn't the struct be in the header?
05:54 hmmmm      it is
05:55 Exio       hehe :P
09:10 celeron55  i love how clueless these people are: http://www.reddit.com/r/gamedev/comments/17v40k/how_to_handle_large_numbers_of_objects_such_as/
15:24 thexyz     https://github.com/celeron55/minetest/blob/master/src/mapgen.cpp#L162 — should return a value
15:26 thexyz     breaks msvc2010 build instead
15:27 thexyz     hmmmm: ^
16:47 celeron55  by the way, does anyone have any plan to do a release?
16:48 celeron55  it's been two months without one
16:49 Exio       what is the status of the mapgen? (as "finished" code)
16:49 Jordach    would be nice yes.
16:49 Calinou    +1
16:50 celeron55  Exio: that doesn't matter, and it is probably better to not have the actually new mapgen stuff in it, because it should be stable
16:50 Exio       yes, but i mean, the actual code
16:50 Exio       in any case, it'd be nice
16:50 PilzAdam   IMO first get merge this TTF thing to get some languages to work
16:51 celeron55  generally, when you want to do a release, you *don't* want to include anything in it that is new to the developers - and for minetest, developers are for example everyone on this channel
16:51 Calinou    what about my glitch ladder fix too, can it be merged?
16:52 hmmmm      thexyz, nice catch... i guess change it to void or add return true
16:52 hmmmm      there is no way that should cause it to _not_ compile however
16:53 celeron55  well, just to set things going forward, let's set a feature freeze at 06:00 UTC next sunday, and 0.4.5 will be released sometime next week?
16:53 hmmmm      celeron before doing a release i want to add in the new finished mapgen flag config thing and loglevel configuration entry
16:54 hmmmm      alright sounds good
16:54 thexyz     how about freetype?
16:55 celeron55  if you can make it work well, i guess nobody opposes you
16:56 celeron55  to both
16:57 hmmmm      i have more fixes, additions, and enhancements that i hadn't added to master yet as well
16:57 thexyz     it works pretty well now, the only problem we've discovered is that windows builds no longer work in wine because of >std::locale::global(std::locale(std::locale(""), "C", std::locale::numeric));
16:57 PilzAdam   thexyz, and the font is bigger
16:57 Jordach    thexyz, then let someone with a windows os actually test it
16:57 PilzAdam   Jordach, it works on XP
16:58 celeron55  minetest should work in wine too
16:58 thexyz     I tested it in virtualbox, works flawlessly
16:58 celeron55  at least playably; if you can't use non-ascii characters, that isn't a problem
16:58 thexyz     it just crashes on startup
16:58 Calinou    why would you want it to work in wine anyway, it's not a big problem if wine compat is broken :P
16:58 thexyz     fine, that's not urgent feature
16:58 celeron55  because it is a nice feature to have
16:59 PilzAdam   celeron55, it works with LANG=C
16:59 PilzAdam   (in wine)
16:59 thexyz     oh, yes, it does
16:59 thexyz     no idea why though
16:59 celeron55  whatever; but i don't want to develop a culture of dropping platforms in here - it won't end up well
17:01 celeron55  by the way, wine does produce good backtraces if you give it a build with debug symbols
17:02 celeron55  http://wiki.winehq.org/Backtraces
17:02 thexyz     so, is LANG=C wine bin/minetest.exe fine or should that problem be investigated further?
17:03 celeron55  i would like to know how it crashes
17:04 Exio       thexyz: PilzAdam, what is the "message" when it crashes?locale::facet::_S_create_c_locale name not valid" ?
17:04 Exio       oops, copy fail
17:04 thexyz     yes
17:05 thexyz     oh, by the way, I rewrote everything into STL, any chance for this to be merged?
17:06 celeron55  nobody will unless you do
17:08 celeron55  is there a way to see what locales are supported, in C++?
17:09 celeron55  or, well
17:09 celeron55  actually, just put a try{}catch{errorstream<<"Could not set numeric locale to C"<<std::endl;} in there
17:10 celeron55  that error is an standard library exception anyways
17:11 celeron55  i am not sure whether things will break if that numeric locale can't be set... it might screw up floating point number serialization/deserialization to strings, i dont know
17:11 celeron55  which would show up in the Settings stuff
17:15 thexyz     PilzAdam: can you test it once again?
17:18 PilzAdam   I get "18:18:20: ERROR[main]: Could not set numeric locale to C" but it works
17:22 thexyz     fine
17:24 celeron55  let's call it working in wine then 8)
17:25 celeron55  it's kind of irresponsible to allow it to not work on one platform while having it on all others, but... as it's just a game, better work a bit wrong than not at all
17:26 celeron55  (as long as it's nothing related to server security or such)
17:48 PilzAdam   thexyz, add the build files of cguittfont to .gitignore
17:52 thexyz     PilzAdam: done
18:10 hmmmm      so wait, since celeron is around
18:10 hmmmm      where should i put functions that are being primarily used for Settings stuff but is generic enough to be used elsewhere that just deals with strings
18:10 hmmmm      it's not really a string manipulation function so i am wondering if it really does belong in util/string.cpp
18:11 hmmmm      in my own projects i always have a 'util.c', 'misc.c' or 'fxns.c' that i put things exactly of this nature in
18:21 celeron55  there was utility.h
18:21 celeron55  guess what happened to it?
18:21 celeron55  it grew waaaaaay too large and was split up
18:21 celeron55  don't recreate that mess
18:22 celeron55  that is all i have to say 8)
18:22 celeron55  if your stuff fits in string.h/string.cpp without adding many extra includes in string.h, put it there
19:06 hmmmm      anyway this is unrelated but i feel like ore amount should be based on perlin noise
19:07 VanessaE   I'd say two layers of perlin plus depth.
19:07 hmmmm      it doesn't make sense for every part of some terrain to have an equal distribution of coal and iron.. it just doesn't.  doing this would also add a new interesting dimension in gameplay, so people would be encouraged to explore a bit and find the resources they need
19:08 VanessaE   depth translates to pressure, one perlin translates to "this element was deposited here", while the other acts as a surrounding material density control
19:12 celeron55  the map generator of 0.3 uses (or, at least some version used) 3D noise for something like ground wetness and hardness, and did ore and gravel placement (can't recall what else) with them
19:13 celeron55  it doesn't make sense to go overboard with such stuff, but a bit is good
19:14 celeron55  basically, if you need an ore, there must be a way of doing it with sane amount of work - you need to be able to find out where to mine to be able to stumble upon the stuff with reasonable work
19:14 celeron55  currently, you just need the height
19:15 celeron55  if it's distributed with large perlin noise variation, then there must be some other way other than just digging and digging
19:16 celeron55  eg. some deposits could have gravel on top or underneath of them, or some shape of underground water crack - stuff like that
19:22 celeron55  i'm once again worried that we might go to a bad direction gameplay-wise, but really there isn't really a way to do anything worse than in the current v6, except screwing up balance
19:22 hmmmm      honestly i'm gonna have to say that v6 is boring
19:22 hmmmm      especially this iteration of it
19:23 celeron55  so where's the v5 port to the new framework? 8)
19:23 hmmmm      not present
19:23 celeron55  shouldn't be hard, should it?
19:23 hmmmm      of course not
19:23 celeron55  i'll maybe do it sometime this week... maybe
19:24 hmmmm      i'll put the emerge thread cleanup project on hold for it if you'd like me to put it in
19:25 celeron55  at least i'd like to try v5 in current MT
19:25 hmmmm      well okay
19:25 celeron55  make sure to have it as intact as possible 8)
19:28 hmmmm      i thought you didn't want to have anything to do with minetest anymore
19:35 rubenwardy where is the chat console written and updated
19:35 rubenwardy ?
19:35 PilzAdam   guiChatConsole.cpp?
19:35 rubenwardy yeah, just saw that
19:36 PilzAdam   you are going for some segfault hunting?
19:37 rubenwardy maybe
19:41 rubenwardy I am just placing around, learning the structure
19:51 rubenwardy is v3f equal to irr::core::vector3df?
19:55 celeron55  hmmmm: not as much as before, but i'm fine with being a contributor 8)
19:56 celeron55  and consider me as such; i am not going to take as much responsibility as i historically have
20:02 hmmmm      there https://github.com/kwolekr/minetest/commit/9845209db2b2b60c7fdfc13a4029bd6a536c76cd
20:02 hmmmm      now neither you nor plizadam can complain
20:06 celeron55  good
20:07 sfan5      how do i compile a debug build of irrlicht?
20:09 sfan5      ..solved!
20:10 hmmmm      i'm gonna merge the current outstanding patches without pull requests if nobody objects
20:10 hmmmm      which is everything from the seed generation up  https://github.com/kwolekr/minetest/commits/no_mgv7
20:15 sfan5      if you commit something notify me, the Minetest-tan rssnotify module is currently broken