Time  Nick       Message
05:32 VanessaE   what's the deal with farming mod?  wheat seeds are not growing
05:38 sofar      can you do me a favor? load the inspector mod and see if the nodetimer is running on them
05:40 sofar      are they on wet soil? enough light?
05:40 VanessaE   actually it might just be a mod conflict vs. farming_plus
05:40 sofar      heh
05:40 sofar      tenplus1' magic at work
05:40 VanessaE   wet soil, yes.  enough light, yes.  inspector does not mention a timer.
05:40 VanessaE   actually, I think this is pilzadam's old farming_lus
05:40 VanessaE   +p
05:41 sofar      well, I see the timer for my minetest_game
05:41 sofar      250secs remaining
05:43 * sofar    sees cotton and wheat plants appear
05:43 sofar      shrug, works for me
05:44 VanessaE   any chance someone can fork it into minetest-mods and fix?  or is there something that can safely replace it?
05:47 sofar      pilzadam has refused to consider minetest-mods before
05:49 VanessaE   well his mods are adandoned now, aren't they?
05:49 VanessaE   at least farming_plus - he tried to hand that one over to me, and I have commit access I think, but I didn't really want control over it.
05:50 VanessaE   (I don't know the code therein well enough)
05:59 sofar      same with nether
06:00 sofar      I fixed up some stuff, but I'm just going to stop unless he moves it to minetest-mods
07:12 sofar      is there a way to make a formspec textlist scroll to the bottom?
07:18 sofar      hmm I think there is
07:20 sofar      yup :D
09:23 Krock      Great. Now we have three pulls about the mesh update limitation.
10:30 red-002    sfan5 shouldn't #5273 at least be reconsider since it got one approval and so did the previous one?
10:30 ShadowBot  https://github.com/minetest/minetest/issues/5273 -- Extend is_yes() by red-001
10:32 red-002    or are you still insisting that on means false
11:52 Krock      on is off and reds are greens
11:52 Krock      red-002, that pull is somewhat trivial and to not extend this discussion to death... just merge it
11:53 Krock      or better: get a 2nd approval and let it merge
11:54 red-002    considering two different people approved the first and the second pull request that shouldn't be too hard
11:55 red-002    but first it would be nice if someone could reopen the pr
11:55 Krock      sfan5 ^
12:28 sfan5      Krock: "this would cause some discussion and to avoid that we'll just merge it" is the worst approach to PRs i have heard yet
12:28 Krock      thanks :)
12:28 sfan5      red-002: additionally pr merging is not a voting contents where "2 votes and you're in" is a thing
12:30 red-002    doesn't mean you should close PR because you think they are unnecessary
12:31 red-002    especially if people support the PR
12:32 red-002    just vote against the pull request or something
12:32 rubenwardy 2 +1 don't count if there are objections
12:32 rubenwardy you need to resolve the objections
12:39 sfan5      red-002: multiple people disagreed with the approach last time
13:05 red-002    well if you don't want to merge this you should make it clear that minetest.is_yes doesn't support "on"
13:06 rubenwardy ie: list the valid true responses
13:06 red-002    ^
13:07 red-002    is you don't want to fix it at least document it
13:07 red-002    if*
13:08 sfan5      sure send a pr
14:18 nore       #5274, almost finished
14:18 ShadowBot  https://github.com/minetest/minetest/issues/5274 -- [WIP] Clientside translate by Ekdohibs
14:20 Krock      lovely. but escaping looks quite ugly IMO
14:21 Krock      but helper functions should solve that
14:23 nore       Krock: well, the escaping is never seen
14:23 nore       except on *very* old clients
14:23 red-002    don't old clients see it?
14:23 Krock      it's seen in the definition table in mods, nore
14:23 nore       no, they will just strip escape sequence
14:24 nore       Krock: what do you mean?
14:24 Krock      +       description = "\x1bTStone \x1bFarg\x1bE\x1bE",
14:24 red-002    ^
14:26 red-002    so what will old clients see? Just the normal untranslated string?
14:27 Krock      https://github.com/minetest/minetest/pull/5274/files#diff-ad60d65b34e16a3319296bb5d683acd6R2385  <- unused stream
15:05 burli      hi
15:07 Krock      hi burli
15:07 burli      is it possible that find_nodes_in_area doesn't work well in register_on_generated?
15:07 burli      local nodes, count = minetest.find_nodes_in_area(minp, maxp, {"air"})
15:08 burli      I would expect that this returns 512000 in empty chunks with just air.
15:08 burli      But the highest value I get is 53248
15:10 burli      est31, that was the first I tried (MarkuBu @github)
15:11 burli      hi Krock
15:11 Krock      check it for ignore
15:13 burli      Krock, tried it already. Doesn't change anything
15:13 burli      {
15:13 burli      air = 53248,
15:13 burli      ignore = 0
15:13 burli      }
15:15 burli      is it a bug or is it just not possible to run minetest.find_nodes_in_area in register_on_generated?
15:16 Krock      Lua is single-threaded, thus I see no reason why it shouldn't be possible
15:17 burli      is the area to large?
15:21 Krock      tried it to. getting 0 ignore and 0 air after writing to the map in my Lua mapgen
15:23 burli      tried the same with "group:stone". Would expect a few 100k underground, but the highest value is 61894
15:23 nore       Krock: ah, but that is only testing for
15:23 nore       now
15:24 nore       I didn't code the mintest.translate call ye, that's all
15:24 Krock      oh, then it's no problem :)
15:24 burli      I guess I'll open an github issue
15:25 Krock      soon: 837 issues
15:27 Krock      "wha are there seven hundred and twelve open issues on github"
15:27 Krock      november 2016
15:28 nore       "there are only hundreds of issues now, go fix them before there are thousands" -- c55, a while ago
15:28 nore       almost a thousand :/
15:28 nore       (now)
15:29 Krock      yeah..
15:34 paramat    luckily many are feature requests
15:34 paramat    i suspect 'find nodes in area' was coded for small areas and a mapchunk is too big
15:34 paramat    for a mapchunk you would use a lua voxelmanip instead
15:35 Krock      but C++ would be quite faster, if implemented correctly
15:38 paramat    the node 'count' is a u16, max 65536 values
15:38 Krock      oh, overflows.
15:39 burli      oops
15:39 Krock      changing this to an u32 shouldn't be a problem
15:39 burli      that explains a lot
15:39 Krock      just to make sure this doesn't end in weird results
15:40 paramat    https://github.com/minetest/minetest/blob/master/src/script/lua_api/l_env.cpp#L653
15:41 paramat    yes will need testing, a table of 512000 positions
15:43 Krock      should I make a PR to clean up this code?
15:43 burli      I can split it to mapblock size, but makes it a little bit more difficult
15:44 paramat    u16 entries is a 40^3 volume
15:45 paramat    cleanup should be a separate commit
15:46 paramat    there are some unnecessary indents in the triple loop
15:46 paramat    a few spaces needed
15:47 paramat    i'd be ok with a u32, if tested to be ok
15:50 red-002    #5255
15:50 ShadowBot  https://github.com/minetest/minetest/issues/5255 -- Add option to also check the center to `find_node_near` by red-001
15:54 paramat    a lua table of 512000 v3s16s will be a lot of memory though
15:55 paramat    if possible it would be better to process the volume using a LVM
15:57 paramat    maybe u16 was chosen to avoid problems caused by excessive-sized volumes?
15:57 paramat    we should add a limit
15:57 paramat    just larger than a mapchunk
16:01 red-002    #5276
16:01 ShadowBot  https://github.com/minetest/minetest/issues/5276 -- Fix the documentation for `minetest.is_yes` by red-001
16:02 red-002    this should at least be documented
16:02 red-002    if fixing it is too complex
16:05 paramat    +1
16:11 Zeno`      looks kind of complicated to me
16:12 Zeno`      merged
16:13 burli      paramat, I want to avoide LVM because it makes it slower and more difficult to search for groups for example
16:15 burli      I think the limit of find_nodes_in_area is ok. Maybe it's better to reduce the size to mapblocks
16:18 burli      paramat, you are right. u16 would be nice for 40^3, but I have no guarantee that the chunk size is always 5
16:28 paramat    even if we decide against u32 a cleanup and a limit that throws an error will still be good
16:28 Krock      "With u32 the resulting list would go crazy"
16:28 Krock      even now it goes crazy, no matter how many times it overflows
16:31 paramat    it's just a case of how much memory use is acceptable, lua tables have slow garbage collection so memory use might be quite high, it might trigger OOM with LuaJIT
16:33 paramat    need to test with a mapchunk's worth of nodes, in on-generated
16:34 Krock      limit by volume or by count?
16:35 paramat    volume
16:35 paramat    as that is the limit of co-ordinates stored and a limit of any individual count
16:35 Krock      right
16:36 paramat    there may be a function for 'volume of' somewhere
16:36 Krock      found on in the vmanip code.. not useful in this case
16:36 Krock      *one in
16:45 Fixer      i have OOMs all the time with dreambuilder
16:45 Fixer      while vanessa don't
16:53 VanessaE   Fixer: it helps that I have 16GB of RAM :)
16:54 Fixer      VanessaE: it is complicated with that damn luajit
16:55 VanessaE   yeah that is a problem for some people
16:55 VanessaE   I don't know wtf the author was thinking
16:58 Fixer      and as usual I accumulate all the problems around
16:58 paramat    that's because you're 'Fixer'
17:02 juhdanad   Next step: connect the accumulated problems to make Minetest crash so badly that the map becomes unrecoverable. Then open an issue.
17:05 paramat    #5278 suggestion
17:05 ShadowBot  https://github.com/minetest/minetest/issues/5278 -- Biome API: Heat and humidity gradient parameters
17:07 Fixer      get biome info
17:08 paramat    crumbs i guess that's my job now
17:09 paramat    it's on my large todo list
17:12 juhdanad   Will the client also see the biome info?
17:13 paramat    well i mean APIs for 'get heat/humidty/biome at pos'
17:29 paramat    over the next week i might test #4967 , then i will be able to +1 . needs review from others though
17:29 ShadowBot  https://github.com/minetest/minetest/issues/4967 -- New bulk node light update by juhdanad
21:01 rubenwardy #5283
21:01 ShadowBot  https://github.com/minetest/minetest/issues/5283 -- Fix wrong meta key in item meta on ItemStack construction by rubenwardy
21:09 paramat    epic title
21:10 rubenwardy it's not a great commit message :D
21:10 rubenwardy better than "fix stuff" though
21:25 red-002    could someone review #5152 or #5255 ?
21:25 ShadowBot  https://github.com/minetest/minetest/issues/5152 -- [CSM] Add `get_wielded_item` by red-001
21:25 ShadowBot  https://github.com/minetest/minetest/issues/5255 -- Add option to also check the center to `find_node_near` by red-001
21:26 nerzhul    red-002, in fact i reviewed 5152 & merge it to CSM branch
21:55 kaeza      it would be nice to have a way to specify a human-readable name for mods
21:55 kaeza      is anyone working in something like this?
21:56 kaeza      for e.g. "Mods" tab, advanced settings, etc
21:57 kaeza      maybe a "title" field in mod.conf
21:57 Fixer      please read personal appeal of fixer
21:57 Fixer      wikipedia guy picture
21:58 Fixer      modpacks are broken, i can't use them properly if there are same named mods in both modpacks, many other scenarious
21:58 Fixer      please at least allow it as load_folder_mod if mod is in some folder
21:59 Fixer      or load_folder_folder_mod... etc
21:59 nerzhul    Fixer, maybe have modpackname_modname could be better ?
22:00 Fixer      could be
22:00 Fixer      any suggestions and ideas here: https://github.com/minetest/minetest/issues/4183
22:01 Fixer      so when I select THIS mod, it will be THIS MOD
22:02 kaeza      that will be ugly
22:02 kaeza      modpacks have no ID like mods
22:03 kaeza      +do
22:03 kaeza      IMHO that should be fixed at the modpack level (see e.g. mesecons and firearms)
22:04 kaeza      modders should use a prefix for submods IMHO
22:04 Fixer      you are overcomplicating
22:05 Fixer      just remember exact path to mod you enabled, problem solved
22:05 Fixer      relative path*
22:05 Hijiri     we should also have mod capabilities/alternatives system (like I think rw suggested at some point?)
22:06 Hijiri     then we won't need "X redo" to have the same mod name as the old mod
22:06 Fixer      AND
22:06 Fixer      minetest.conf should not be cluttered with some mod garbage
22:06 kaeza      >overcomplicating
22:07 kaeza      looks like the other way around
22:07 Fixer      currently it is broken, you can't leave it like that
22:08 Fixer      also, if dreambuilder subgame wants special settings, they should be specified in subgame config, not in main minetest.conf*