Time  Nick        Message
06:29 Emerald2    Hey muke.
06:29 muke_       hi
06:29 Emerald2    Ahhh an aussie!
06:29 muke_       i forgot ips were public
06:30 Emerald2    Unless you ask freenode staff for a cloak.
06:30 muke_       oh well
06:30 Emerald2    Hey neighbour from across the ditch. :p
06:31 muke_       pondhopping isn't my style.
06:34 Emerald2    Gonna get some food and hope the exile winter passes so we can get off the volcano. :p
12:22 specing     Hmm, is there a command line switch to make run-in-place minetest build use ~/.minetest as its home?
12:23 sfan5       no, this choice is compiled in
12:23 specing     Hmm, or is run-in-place the thing that controls this, and I should make a normal build
12:23 sfan5       yes
12:24 sfan5       if you want stuff to live in ~/.minetest do a RUN_IN_PLACE=0 build
12:33 specing     sfan5: out of tree builds are not properly working, at least the mt binary is deposited into source dir
12:34 sfan5       don't ask me why but that's intentional
12:34 specing     rip, also git clean nuked my first baby CSM
12:35 specing     strangely it did not nuke another one that I had installed
13:57 specing     Hmm, minetest is no longer loading any CSMs...
13:58 specing     just as I write this...
14:10 specing     Hmm, Minetest says that 'io' is nil (in CSM) even though I do luaL_openlibs(m_luastack); unconditionally
14:11 specing     (commented out clientOpenLibs)
14:14 * specing   adds io to clientOpenLibs instead
14:14 specing     monkey strategy.. :/
14:29 specing     { LUA_IOLIBNAME,   luaopen_io      }, is also not working
14:30 specing     Is there some other sanitation going on?
14:33 * specing   notices src/script/cpp_api/s_security.cpp
15:39 specing     Well, renaming lua's io to rio did not work either
15:39 specing     { "rio",   luaopen_io      },
18:25 tango_      hello all
18:26 tango_      came back to minetest after a long while away, nice to see the improvements, my kids love it too
18:26 tango_      couldn't get it to run it on my Android 10 device though, gets stuck on boot loop (got the apk from fdroid)
18:26 tango_      pity
18:28 tango_      (I'm not even sure where this bug should be reported: is there a repo specific for the android version?)
18:29 tango_      hm I believe this patch would be aimed among other things at fixing the issue https://git.minie.re/minetest/minetest/commit/87b25e583dcb2bc1079a7b5906b97a787ac04be8
18:31 tango_      oh it was merged
18:31 tango_      (dunno why google led me to that site instead of GH, https://github.com/minetest/minetest/pull/9900 seems to be merged)
18:31 tango_      I guess i'll have to wait for the fdroid update then
18:40 rubenwardy  tango_: F-droid won't update for a long time
18:40 rubenwardy  you could try using the APK itsel
18:40 rubenwardy  https://github.com/minetest/minetest/releases/download/5.3.0/app-armeabi-v7a-release.apk
18:46 specing     rubenwardy: why wont they update?
18:47 rubenwardy  the build system was changed and now depends on prebuilt binaries
18:47 rubenwardy  F-Droid forbids prebuilt binaries
19:08 majochup    tango_, You can also use an older version of minetest on f-droid.
19:08 majochup    I for example use 5.2.0.26
19:10 rubenwardy  doesn't support Android 10+
19:10 majochup    Ohh
19:11 tango_      why does the new build system depend on prebuilt binaries?
19:14 rubenwardy  in order to get the requried libraries
19:14 rubenwardy  Minetest used to have complicated and fragile makefiles to build the libraries, this was replaced
19:14 rubenwardy  the eventual aim is to reintroduce building the dependencies in CI
21:06 specing     Hello
21:07 specing     Any clues as to why io.open does not work in CSMs? (specifically, io is nil). I've commented out openClientLibs so that lual_openlibs is called for client mods, and I also tried adding LUA_IOLIBNAME into openClientLibs to no avail
21:07 specing     I've also tried loading io as "rio" by substituting "rio" for LUA_IOLIBNAME in openClientLibs
21:07 specing     also did not work
21:08 specing     I've also commented out all the parts related to the io library from the ...security.cpp
21:08 * specing   ran out of ideas
21:12 DS-minetest specing: io is nil on purpose for security reasons, afaik. I don't know what openClientLibs is.
21:12 DS-minetest specing: use modstorage
21:12 specing     I don't want to use modstorage for this
21:12 specing     Where is io made nil?
21:13 rubenwardy  you need to copy all of io in security.cpp
21:13 rubenwardy  it's not made nil, it's just not copied
21:13 specing     void ScriptApiBase::clientOpenLibs(lua_State *L)  line ..
21:13 specing     ah rubenwardy, thanks
21:13 rubenwardy  sandboxing needs to whitelisted, not blacklisted, so it creates the environment from scratch by copying
21:14 specing     rubenwardy: though io should not be nil after adding it to clientOpenLibs, just certain functions from it
21:14 specing     from what I read in security
21:24 rschulman   How do people playing with v7 mapgen (so some ores generate REALLY deep) remember how to get down to where they were digging and back up?
21:25 rubenwardy  mineshafts with water, mostly
21:37 specing     rubenwardy: travelnets, elevators
21:42 MTDiscord   <12W​arr1024> ziprunes :-D
21:48 rschulman   I'm not smart enough for NodeCore, though. :D
21:51 MTDiscord   <12W​arr1024> haha, people sometimes say they're not smart enough for it, but the game will let you take it at your own pace, especially in singleplayer
21:51 MTDiscord   <12W​arr1024> apparently people have also been doing some work on the wiki now
21:52 rschulman   Oh, I know, I've actually played through a small bit of it.
21:52 rschulman   I was just joking.
22:26 tango_      damn, iron ore is hard to find
22:26 tango_      is there a way to know which mapgen was used, as a user?
22:27 SwissalpS   the seed, yes
22:46 specing     tango_: really?
22:46 specing     tango_: my experience is that it's really commonplace past -500 or so
22:47 tango_      yeah but you have to dig to that depth first
22:47 specing     true, but that's usually easy
22:48 tango_      it's a bit boring
22:48 specing     yes
22:48 specing     not just a bit, a lot
22:48 specing     ore should be in veins, like in real life
22:48 tango_      especially if all you want is just a little bit of iron to build a frigging screwdriver
22:48 Corey[m]    Has anyone experienced a bug where right-clicking and then passing the cursor over an entity triggers that entity's right click (as in you clicked on something different, and then moved the crosshair over the entity, doesn't even have to be immediately)?
22:48 specing     so instead of finding iron 20 nodes apart, you should find it in massive veins, but the veins being rare
22:49 tango_      that aside, how does farming work?
22:49 specing     tango_: also no such problem with TechPack autosieves :)
22:49 tango_      8-P
22:49 tango_      meaning, how can i have more than 12 or 13 light for the field? stack lots of torches around it?
22:52 MTDiscord   <04L​one_Wolf> @Corey yes, it's a bug with the bindable LMB/RMB PR they merged. Lemme grab the link
22:52 MTDiscord   <04L​one_Wolf> Or wait
22:52 MTDiscord   <04L​one_Wolf> Are you in 5.4?
22:53 MTDiscord   <04L​one_Wolf> Sounds like proper behavior now that I reread that
22:53 specing     tango_: I personally use SuperGlowGlass from moreblocks
22:53 specing     tango_: you can also use ultra plastic from sci-fi nodes
22:55 SwissalpS   meselight-slabs or super glow glass slabs do the job for me in most setups
22:55 Corey[m]    Lone_Wolf: I always run the git master
22:56 SwissalpS   only rarely do I need the full nodes
22:56 Corey[m]    And I've noticed that the console `/` binding broke as well
22:56 Corey[m]    At this point,  I would suggest reverting the change, unless something else is relying on it, because this is hella broken
22:58 specing     you can also use trap super glow glass, so you can move through the field
22:58 specing     or slabs on ceiling
23:01 specing     rubenwardy: specing | rubenwardy: though io should not be nil after adding it to clientOpenLibs, just certain functions from it
23:02 specing     I'm sorry if I had missed an answer to this, but ...
23:03 specing     When IO is loaded to client libs, the safe io functions should be copied to the new environment in src/script/cpp_api/s_security.cpp
23:03 specing     But my problem is that there is no io (it is nil)
23:04 specing     ah no wait, there is a initializeSecurityClient that has no mention of io
23:04 specing     nvm
23:26 tango_      s there a way to know at what depth one is?
23:27 DS-minetest tango_: press F5
23:31 tango_      oh nice