Time  Nick           Message
02:46 hmmmm          just because settings code is unused it means it's bad?
02:47 est31          it certainly makes life hard if you want to e.g. make a faster settings patch
02:47 est31          been there done that
05:09 paramat        hmmmm now that we can hide mapgens i would like to add the fractal mapgen in hidden mode #3088 it's playable but unfinished and has a mysterious bug described in the thread. i've had to manually rebase it many times, it will be easier for me to develop it if it's merged
05:09 ShadowBot      https://github.com/minetest/minetest/issues/3088 -- Mapgen: Add 4D fractal mapgen by paramat
05:09 hmmmm          ahh
05:09 hmmmm          it looks like a good version of proller's fractal mapgen
05:10 paramat        indeed
05:10 hmmmm          looks pretty standard
05:10 hmmmm          i don't see anything obviously wrong
05:10 hmmmm          looks good to me.  +1
05:11 hmmmm          err
05:11 hmmmm          I added U16_MAX and so on, you might want to modify the (u16)-1s
05:11 paramat        ah of course
05:12 paramat        thanks
05:12 hmmmm          also FYI there are "v3f" types :-)
05:12 hmmmm          I see you have a lot of:
05:12 hmmmm          scale_x
05:12 hmmmm          scale_y
05:12 hmmmm          etc.
05:12 paramat        ok combine
05:13 hmmmm          oh the destructor for MapgenFractalParams needs to be virtual
05:13 hmmmm          same with MapgenFractal
05:14 hmmmm          oh crap none of them have that
05:14 hmmmm          leave it go for now, I should fix that up later for all of them at the same time
05:14 paramat        ok
05:15 paramat        any idea on the weird bug described?
05:15 paramat        i will try playing with casts
05:16 hmmmm          i wouldn't be able to figure it out without trying it myself
05:16 paramat        yeah no worries i'll try a few things
05:16 hmmmm          I think some memory is being accessed out of bounds - just a hunch
05:18 paramat        if so what could cause that within a mapgen?
05:19 hmmmm          anything
05:19 paramat        heh ok
05:19 hmmmm          it's useless wondering about hypothetical situations
05:19 hmmmm          you need more information
05:20 hmmmm          try setting some printfs around places to see if variables are what you expect them to be
05:20 paramat        yeah
05:21 paramat        that was my plan
05:39 paramat        now merging 3088
05:49 paramat        merge complete woohoo
07:53 hmmmm          hey guys, I think it might be worthwhile to merge a couple of pull requests and do a release
07:53 hmmmm          just so things don't start lagging behind
07:53 hmmmm          let me know what you think.
07:55 sfan5          sounds like a good idea
07:55 sfan5          what pull requests did you have in mind for merging?
07:55 hmmmm          just as many as we can
07:55 hmmmm          there's a lot of good stuff in there already
08:01 sfan5          hmmmm: what do you think about #3210
08:01 ShadowBot      https://github.com/minetest/minetest/issues/3210 -- WoW-style Autorun by duane-r
08:01 hmmmm          eh
08:01 hmmmm          never played WoW
08:02 sfan5          me neither
08:02 hmmmm          what is this useful for exactly...?
08:02 sfan5          "+1 very useful for dev work on mapgen when flying to explore terrain, or long journeys on the flat or on paths with ramps (on foot or in a vehicle), or in a boat."
08:02 sfan5          (https://github.com/minetest/minetest/pull/3210#issuecomment-144129742)
08:02 hmmmm          what do you think?
08:03 sfan5          it might not be that useful for many people
08:03 nrzkt          i played on wow and having this feature could good yes
08:03 sfan5          but as long as it doesn't confuse players when they accidentally press F its ok
08:03 sfan5          (it's not bound to a key by default)
08:05 hmmmm          yeah, probably for the best.
08:06 hmmmm          hrmm
08:06 hmmmm          it needs a default setting set for disable_mouse_wheel otherwise it'll crash
08:09 sfan5          nrzkt: is #3220 ready or does it still need work (est's suggestion)?
08:09 ShadowBot      https://github.com/minetest/minetest/issues/3220 -- Improve ABM performance: pre-resolve node ids by nerzhul
08:11 nrzkt          sfan5 it needs some work, maybe somebody can improve it, i'm a little bit busy this week to solve this problem
08:14 nrzkt          I also thing RealBadAngel Meta performance improve should be improved and merged if you want to do a release, this change is very useful to same some FPS
08:14 VanessaE       and that damn VBO idea needs to be finished, too
08:18 VanessaE       ..and sometime really soon, texture atlases.
08:18 VanessaE       (yes, I know.  "who's gonna code it?"  I don't know.)
08:19 VanessaE       as for autorun, frankly if that got turned on by accident, I would personally find it very confusing.
08:19 hmmmm          kahrl can code it =]
08:19 hmmmm          he was the one who removed it
08:19 hmmmm          hehe
08:19 VanessaE       heh
08:27 TenPlus1       hi folks
08:27 VanessaE       hi
08:27 TenPlus1       hi vanessa
08:28 TenPlus1       does anyone know of a simple way to remove ALL metadata from a node position?
08:28 TenPlus1       even after removing node the metadata remains (apart from infotext)
08:29 VanessaE       odd, set_node() is supposed to overwrite the node.
08:29 VanessaE       sounds like a bug; this is precisely what the swap_node() call was added for
08:29 TenPlus1       it overwrites the node but any pre-defined metadatas remaings
08:29 TenPlus1       *remains
08:30 TenPlus1       even placing a schematic block over the top leaves metadata intact...  I dont get it
08:33 TenPlus1       *hint hint* devs maybe add a meta:clearall command :P
08:34 TenPlus1       I can only imagine how many empty nodes still have a ton of data contained inside, bloating the map
12:27 * OldCoder     reports that some combinations of username and password result in unexpectedly long hash strings; bug in the core or builtin. Causes some mods to break. Affects 5% to 10% of players.
12:36 * celeron55    picks his nose and points to github issues
12:37 * Krock        picks the nose and runs away
14:19 neoascetic     Hi all. Please tell me where NDEBUG flag is defined? I am building with CMAKE_BUILD_TYPE=Release, but my server freeze after redis connection exception instead of fail: ERROR: An unhandled exception occurred: Redis command 'MULTI' failed: Connection reset by peer
14:20 nrzkt          NDEBUG is defined when you choose Release mode
14:21 neoascetic     Oh, well. Everything to make app not to fail, right?
14:23 neoascetic     https://github.com/minetest/minetest/blob/e7282816fffc433861e7e83b0bf1de701d3ea2d1/src/debug.h#L44-L51
14:24 neoascetic     hmm, my server doesn't abort with error; instead, it just echoes error and waiting. Wrong comment or wrong behavior?
15:58 est31          neoascetic where are you?
15:58 est31          Bc I wonder which commit you are on
15:58 est31          this sounds like a bug fixed recently
16:00 est31          This commit should have fixed this: https://github.com/minetest/minetest/commit/e7282816fffc433861e7e83b0bf1de701d3ea2d1
16:00 est31          with NDEBUG turned off, and with NDEBUG turned on.
16:01 est31          can you tell whether you can reproduce the behaviour with that commit in place?
16:01 est31          because then it would be a bug
16:02 est31          OldCoder, what you see isn't unexpected behaviour or a bug, its srp hashes
16:02 est31          there was no other way than making them so long
16:06 est31          which reminds me, I've talked wrong stuff in the docs about SRP
16:06 est31          its not compatible to the test vectors of RFC 2945
16:06 est31          it is compatible to RFC 2945 itself, but has another setup
17:09 est31          pushing in 10 minutes https://github.com/est31/minetest/commit/8edc8a14e4507e4ef78b1a0d39d338ee2d4a721d
17:10 est31          my bad, I've documented it wrongly initially
17:10 est31          it isnt RFC 5054 compatible bc it uses SHA-256
17:10 est31          and we dont use TLS xD
17:11 est31          and we use SHA-256 and not SHA1 like claimed
17:11 JohnnyComeL8ly Shouldn't we use security for the chat?
17:11 est31          you mean encryption?
17:11 JohnnyComeL8ly Yes, sorry.
17:11 est31          perhaps, yes.
17:11 est31          I won't be against :)
17:12 JohnnyComeL8ly (I'm no expert) but, I don't see how it would be that much draw on resources.
17:12 est31          computers these days are blazingly fast
17:15 Krock          Minetest was designed to run on older computers too, just a side notice
17:38 JohnnyComeL8ly I know that MT was designed that way, but wasn't able to run MT on a PIV PC (with iGPU, of course):  it is just too slow/skippy and I put the settings as low as possible.
18:16 hmmmm          you know, there's a reason why MT doesn't currently encrypt the entire protocol
18:16 hmmmm          mainly because of the lossiness of UDP
18:19 Calinou        Xonotic has full AES encryption for everything
18:19 Calinou        but nobody uses it :p
18:19 Calinou        it's not enabled by default
18:20 hmmmm          I'm just saying, encryption over UDP isn't exactly trivial or it would've been done already
18:20 hmmmm          we already do a key exchange with the server by virtue of using SRP
18:20 hmmmm          the problem is the implementation of the cipher
18:33 JohnnyComeL8ly Oh.
19:03 Krock          Strange. There are no problems with mining stone in undergorund but why is there a connection delay whenever I dig and place wheat and seeds
21:19 paramat        nrzkt #3233
21:19 ShadowBot      https://github.com/minetest/minetest/issues/3233 -- Android/Mapgen: Add fractal mapgen file to android makefiles by paramat
21:20 nrzkt          paramat, thanks, awaiting for jenkins android build. If works, you could merge directly to fix every new commit
21:20 paramat        ok
21:27 paramat        checks passed
21:29 paramat        now merging
21:34 paramat        done