Time  Nick             Message
08:34 nrzkt            hi -dev ! trivial but a long #4574 ready for review
08:34 ShadowBot        https://github.com/minetest/minetest/issues/4574 -- Replace various std::map with UNORDERED_MAP + various cleanups by nerzhul
08:47 sfan5            nrzkt: lgtm
08:53 nrzkt            sfan5: merging then :)
08:56 nrzkt            done
09:18 Hijiri           can someone check #4128
09:18 ShadowBot        https://github.com/minetest/minetest/issues/4128 -- Prevent attached models from disappearing during parent reload by Foghrye4
09:22 nrzkt            Hijiri: i'm doing it
09:22 Hijiri           thanks
09:24 nrzkt            done,
09:24 nrzkt            some spaces, some code lines to change
09:25 nrzkt            and i ask you to change one MT container type because you are manipulating it here to generic type permitting to switch to C++11 optimized set for this use case
09:27 Hijiri           It's not my PR
09:27 nrzkt            oh i see :)
13:35 red-001          intersting you can search the posts of users who no longer have an account
13:36 Krock            seems like you had some strange number overflow in your name, which started with 1 instead of 0
13:44 _Megaf           Krock: not binary based
13:44 Krock            yes, thus, strange.
13:44 _Megaf           Krock: is base 10, he goes to 999 and resets to 1
13:44 _Megaf           I know, should start in 0
13:44 _Megaf           [08:27:11] --> red-999 (red-999@109.77.13.221) has joined #minetest-dev
13:45 red-255          so much for my cloak
13:46 red-255          :)
13:49 * _Megaf         stops talking before someone says it's not minetest development related
13:54 * red-255        says it's not minetest development related
14:16 nrzkt            guys, release was in apr/may. it's time to release no ? is there any blocker to release MT 0.4.15 ?
14:16 nrzkt            many fixes has been added and some little feature
14:17 Fixer            i see reason for 0.4.14.1 not for 0.4.15 :(
14:31 nrzkt            we didn't do sub patch
14:31 nrzkt            major.minor.patch
14:31 nrzkt            6 months of commit it's a patch :)
14:31 nrzkt            we didn't need to be impressive if it's only user experience improvements
16:10 paramat          Fixer we're very aware of bugs and do care, the issue is simply lack of dev time. please stop being so negative, it's actually harmful to the project. how about improving your coding and starting to contribute since you care so much and have so much time to report bugs and complain? :]
16:14 paramat          your nick is funny
16:14 * paramat        changes your nick to Whiner
16:32 Zeno`            https://www.youtube.com/watch?v=1DZmyAWul9Y
16:35 paramat          i'll merge games 1306 1307 soon
16:37 paramat          sofar sfan5 nore rubenwardy i need game#1303 for new biomes, please can i merge that too?
16:37 ShadowBot        https://github.com/minetest/minetest_game/issues/1303 -- Default: Add grey sand by paramat
16:37 sfan5            grey sand is a little weird as a description
16:37 Zeno`            #1306
16:37 sfan5            how about "dark sand" or "cold sand"?
16:37 ShadowBot        https://github.com/minetest/minetest/issues/1306 -- Fix line run on, making lua-api.txt more readable. by hoodedice
16:37 Zeno`            +1
16:38 Zeno`            #1307
16:38 ShadowBot        https://github.com/minetest/minetest/issues/1307 -- Add minetest.register_on_punchplayer by Bremaweb
16:38 sofar            can I get a sec to see if I can come up with a better name for grey/gray sand?
16:38 nore             paramat: same comment as sfan5: "Grey sand" looks weird, but once this is changed, +1
16:38 Zeno`            err 1307 is already merged
16:39 Zeno`            oh games
16:39 Zeno`            oops :)
16:40 * Zeno`          listens to Evie song again
16:40 nore             (also: since we are starting to have quite diverse versions of basically the same thing (for sand, trees, ...), it would be nice to make them somehow a bit different)
16:40 paramat          sure we can improve the name
16:40 nore             (would be harder for sand than for wood though)
16:41 paramat          texture seems fine to me, it's a lovely design
16:42 sofar            fluvial sand, alluvial sand, volcanic sand... trying to find better names...
16:42 paramat          i chose 'grey sand' to make it more flexible in use, not restricted to cold desert, but sure i don't mind a new name
16:43 Zeno`            volcanic sand is normally black. Although it might be grey if there is coral reef involved as well. The merging of the two is... hmm... variable
16:43 sofar            eolian sand would do well, but it's a difficult word
16:43 sofar            Zeno`: volcanic sands often turn gray after a while when exposed to sunlight
16:43 Zeno`            sofar is the geologist... we should listen to him!
16:44 sofar            but sure, many people think volcanic sands are dark, and they mostly are
16:44 sofar            I'
16:44 Zeno`            sofar, I have a black pearl (which is really grey)
16:44 sofar            I'm not actually looking for a name that declares the origin of the sand
16:44 sofar            just posing ideas
16:44 Zeno`            from an area which has volcanic sands
16:44 sofar            lol
16:45 Zeno`            apparently it's probably worth some $$
16:45 Zeno`            but I aint selling it
16:46 Zeno`            I dunno if alluvial sand would be correct... wouldn't that depend of the alluvium?
16:46 paramat          btw this biome will extend from 'cold desert' into (dry) tundra climate
16:47 Zeno`            auxiliary sand
16:47 Zeno`            :D
16:47 sofar            Zeno`: `alluvial` means `in a river dominated geomorphology`
16:48 sofar            e.g. non-marine
16:48 Zeno`            sofar, yeah, but wouldn't the sand colour then depend on what is being eroded into the river/stream?
16:48 sofar            eolian can mean near seas as well (dunes) or desert
16:49 sofar            alluvial sands are usually more rich in detritus and clay, so yes, browner
16:49 Zeno`            hmm ok
16:49 sofar            that's why aeolian fits better
16:49 Zeno`            what about in sandstone areas? :D
16:50 Zeno`            aeolian is a bit hard to type
16:50 sofar            we could call it `dune sand` to simplify
16:50 Zeno`            yeah that'd work
16:51 Zeno`            i guess :/
16:51 Zeno`            what about sand dunes? lol
16:51 Zeno`            (i mean beach ones)
16:51 sofar            at least it avoids the `gray` vs `grey` sand problem
16:51 Zeno`            but this is a bit silly hehehe
16:51 Zeno`            desert sand?
16:52 sofar            how about
16:52 sofar            `silver sand`
16:52 sofar            avoid the whole genealogy debate
16:52 paramat          silver was my first idea
16:52 sofar            isn't weird like grey, perhaps?
16:52 Zeno`            so, "grey" is the issue?
16:52 Zeno`            just call it "dark sand"
16:53 sofar            in dutch it's even a word I think
16:53 paramat          the biome is inspired by high cold deserts in pakistan
16:53 sofar            silversand (loosely translated)
16:53 sofar            I'd be quite content with it being called `silver sand`
16:53 sofar            it fits the biome, it's a bit more descriptive and catching
16:53 paramat          it's a prettier word
16:54 Zeno`            sounds nice
16:54 sofar            do it
16:55 paramat          ^ nore sfan5 ?
16:55 sfan5            yup
16:55 sfan5            though it should be "silversand" not "silver sand"
16:55 paramat          https://pbs.twimg.com/media/B_vmUKfXEAA1c2_.jpg
16:56 paramat          nope silver_sand, node names now use seperated word
16:56 paramat          http://www.pakistantoursguide.pk/wp-content/uploads/2015/11/Cold-desert-at-dusk.%E2%80%94-S.M.Bukhari.jpg
16:56 sfan5            i meant the description
16:58 paramat          ah i would think Silver Sand is correct english
16:58 sofar            english usually splits out words
16:58 sfan5            it is
16:58 sofar            german/dutch concatenate
16:58 sofar            I'd split it out myself
16:58 paramat          yeah, ok thanks everyone
16:59 T4im             "Police helps dog bite victim."
16:59 sfan5            i'm aware that "silversand" is not correct english, i'd just name it that way because it's not really quite literally supposed to be "just" silver colored sand
16:59 redyellow        sand maked from sliver dust?
17:00 redyellow        made*
17:00 paramat          i see 'silversand' is often used as a fancy name
17:00 Krock            as a name for an infrumental album from 1985 for example
17:00 Krock            *instrumental
17:00 paramat          heh
17:13 sfan5            serverlist stats are in << https://kitsunemimi.pw/tmp/serverlist_stats_2016-10-05.txt
17:13 sfan5            could be interesting for coredevs
17:15 paramat          it is
17:15 twoelk           modernhousebuild?
17:15 sfan5            android clones have quite silly names
17:16 twoelk           I didn't know you could actually identify them
17:18 T4im             passive os fingerprinting, i assume :)
17:19 paramat          mobile is 93%!
17:19 twoelk           16 older than winXP? wonder what they ran on
17:20 T4im             oh, that's simple user agent parsing?
17:20 sfan5            yes it is
17:21 T4im             any chance of minetest unrelated bots being part of that?
17:21 T4im             you know, exploit scaners
17:22 T4im             nvm, i should've read the rest first x)
17:23 twoelk           so most users by far use some sort of multicraft :-(
17:23 T4im             it's impressive how many old clients are still in play
17:26 Krock            but wow.. two million ingame server list fetches
17:27 Krock            and that in less than a month
17:27 twoelk           hm, this is for ten days right? so we have a quarter million clients per day? If some of those are rejoins or such this still gives minetest some  hundreds of thousands unique users per day
17:27 Krock            yeah, imperssive
17:28 T4im             doesn't the same client update the lists, too after a while? or at least when switching tabs?
17:28 Krock            no, only when removing or adding favourites
17:28 twoelk           creating clicks by server hopping?
17:28 T4im             well switching tabs updates server status here though, Krock
17:29 Krock            ah right, yes
17:30 Krock            right now I noticed that I submit wrong data by using my browser to fetch the list. x64 linux gets some clicks more by that xD
17:31 sfan5            T4im: F5 and take a look at the effective version distribution
17:31 sfan5            then you'll be much happier
17:32 T4im             ah :D
17:34 T4im             why are some absolute numbers smaller in the effective distribution now though?
17:34 T4im             shouldn't all of them be larger if you distribute the third-party applications onto them?
17:35 T4im             eh nvm
17:35 T4im             theyare
17:37 twoelk           I wonder wether the desktop/mobile ratio is tipped towards mobile devices because they reconnect much more often due to bad internet and such
17:37 Krock            could be
17:38 twoelk           the last time I tried to play from a phone I was in an overland bus with a bad wlan connection
17:39 sfan5            i could count unique ip addresses in a time frame of 30 minutes
17:39 sfan5            but then i'd need to do all the stats again
17:39 sfan5            nahhh
17:42 twoelk           I did notice on some servers that sometimes players are disconnected and reconnect in groups, like a netsplit on irc
17:43 Fixer            sfan5: do it, is not it is interesting to confirm 7% desktop share?
17:43 sfan5            no i'm going to have some ice cream now
17:43 twoelk           yummy
17:44 * T4im           wonders if mobile clients aren't essentially just crowd sourced bots
17:44 Fixer            i go eat too
17:44 twoelk           well looking at their behaviour does let one wonder
18:58 nrzkt            will push a little compilation fix since recent commit that travis didn't point
18:59 nrzkt            pushed
19:52 redyellow        will minimal be included in the next release?
19:55 _Megaf           probably
19:58 sfan5            minimal will always be included unless we find a reason to omit it
19:58 redyellow        It doesn't really make sense to include it
19:59 sfan5            paramat: greysand now has 2 approvals
19:59 sfan5            why not?
19:59 redyellow        only the sort of people that complie minetest from source have a use for it
20:00 redyellow        half of the game is code for testing stuff
20:01 paramat          ok
20:01 est31            thats why it says "this is minimal and only intended for developers"
20:02 paramat          minimal is probably useful in stable
20:02 paramat          i always keep a 'last stable' version around for testing and comparing
20:06 _Megaf           minimal is good for making mods too
20:08 redyellow        you can't make a lot of mods using minimal
20:08 redyellow        most mods depend on default
20:13 paramat          i'll merge game#1310 also, a trivial tuning of ABM speed to match previous ABM version
20:13 ShadowBot        https://github.com/minetest/minetest_game/issues/1310 -- Default/functions: Tune speed of grass spread ABM by paramat
20:15 paramat          ok will merge games 1303 1306 1307 1310 in 20mins
20:21 sfan5            paramat: why can't we add bushes btw
20:35 paramat          we can, that's in the pipeline
20:36 Krock            that pipeline with the valve on the one side?
20:36 nrzkt            sfan5, i'm working on a worker for C++11 builds using gcc #4578, i will change our exclusion mode to inclusion matrix
20:36 ShadowBot        https://github.com/minetest/minetest/issues/4578 -- [WIP] Changes on travis build matrix by nerzhul
20:36 Krock            sometimes it flushes through, sometimes it gets stuck
20:36 paramat          with the closed valve yes
20:37 sfan5            wait didn't you say that you don't want/like bushes?
20:37 sfan5            also for which biomes are bushes planned?
20:39 paramat          est31 hmmmm nore nrzkt sfan5 sofar Tesseract rubenwardy we're considering a release for around winter solstice Dec 21st, 2.5 months from now
20:39 paramat          erm we disagreed on some details, like bushes in jungles
20:40 est31            paramat, okay with me
20:40 nrzkt            paramat, we're ? who ? because in the list you poke everybody except you in team :p
20:40 sfan5            lemme dig up the bushes pr again
20:40 paramat          but they're good in savanna and maybe grassland too
20:40 paramat          yeah there's new discussion
20:42 paramat          and this time i think we can manage the release together
20:43 sofar            I'd settle for bushes like mine in savannah's. I'd be exhilerated with small bush clumps in grassland
20:43 paramat          yes i like the 3x3x2 savanna bushes
20:44 sfan5            nrzkt: see my comments
20:44 sofar            how about a pine needle version of that in grassland?
20:44 nrzkt            sfan5, did you remove the condition fix comment ? i saw it but after refresh... disappear (for the ||)
20:44 paramat          those with generic 'bush leaves' nodes, in savanna and grassland is fine by me
20:44 sfan5            yes i added a more comprehensive one instead
20:44 paramat          i'd rather keep pine nedles to pine trees
20:44 paramat          (needles)
20:45 nrzkt            nice idea for the compiler var :)
20:45 sfan5            keeps the code compact :)
20:45 paramat          nrzkt by 'we' i meant a discussion a few devs had a few days ago
20:46 nrzkt            paramat, who ? :)
20:46 nrzkt            sfan5, yes ofc
20:46 nrzkt            i think we could remove llvm install no ?
20:47 nrzkt            it's included in base ?
20:47 nrzkt            i will test it
20:47 paramat          erm me, est, sfan .. can't remember
20:47 nrzkt            ty paramat
20:47 nrzkt            because apt-get install $COMPILER i think i can remove the clang + llvm 3.1
20:47 sfan5            not sure, try that before
20:47 sfan5            i think we used that because ubuntu 12.04 had outdated versions
20:48 nrzkt            okay, but 16.04 is now out
20:48 nrzkt            and 14.04 :)
20:48 sfan5            indeed
20:48 sfan5            we need travis to update
20:48 sfan5            but i'm sure you can set a newer version in travis
20:49 sfan5            i might try this with minetest/minetestmapper travis today
20:49 nrzkt            will try it
20:49 nrzkt            for this PR i can do some cleanups :)
20:49 nrzkt            yes i can remove -std=c++11 because GCC 6.2 enable std=c++14
20:49 nrzkt            i will test it by adding a std::unordered_map in map t obe sure
20:49 sfan5            paramat: so we can agree on acacia bushes whose trunks should drop sticks? and also the leaves should be a seperate node somehow?
20:50 paramat          yes, the leaves nodes might drop sticks, we might use a trunk node at the centre to provide wood
20:51 paramat          (for grassland)
20:51 sofar            just do trunk + leaves
20:51 sfan5            why should the leaves drop sticks?
20:51 sofar            I don't see why you'd make it more complex
20:51 sfan5            IMO there isn't much material in bush leaves
20:51 sofar            keeps it simple for the players
20:51 paramat          we could use the generic 'tree' node for the trunk
20:51 sfan5            so you can't justify it dropping sticks
20:51 sofar            I don't see the need for it
20:51 sfan5            just make the trunk drop sticks
20:51 paramat          fine by me
20:51 sofar            why sticks?
20:52 paramat          wood
20:52 sofar            just make it drop a log like normal
20:52 sfan5            nah
20:52 sfan5            bushes are way less dense than normal trees
20:52 sfan5            it would be kinda weird if they dropped comparable amount of resources
20:52 sofar            it's a blocky game with everything being 1m3
20:52 sofar            lol
20:52 sofar            sorry, I think it's overdoing it :)
20:53 paramat          i know it's weird but the idea was to provide enough wood to make a tool, in grassland
20:54 sfan5            can't we make it drop half a wood node? :D
20:54 sofar            how many bushes are you going to put in mapgen?
20:54 sofar            that's really the question
20:54 paramat          i guess it could directly drop a small amount of wood
20:55 sofar            if you are going to make a significant amount, then maybe, it makes sense to limit resources
20:55 sofar            if you are going to scatter a few very far apart, then it doesn't matter
20:55 paramat          i was thinking quite scattered and sparse in grassland, but larger clumps in savanna
20:55 sofar            say, 5 bushes in a 200x200 space
20:55 sofar            each 1 trunk
20:56 sofar            that's 5 trunks in a 40000 node surface
20:56 paramat          in savanna density is not an issue because trees
20:56 sofar            who cares it's 5 logs instead of 10 sticks?
20:56 sofar            it's still not enough to build a house!
20:56 sofar            (20 planks doesn't make a shed, even)
20:57 sofar            sure, it's 80 sticks, so I can make a nice long ladder and some basic tools
20:57 paramat          the issue is, whether to provide wood for tools in grassland, if not sticks might be more suitable
20:57 paramat          i'm not too bothered either way
20:57 sofar            either way you look at it, if you make it sparse enough it's not enough to do major construction
20:58 sofar            so don't bother making it drop even less artificially
20:58 nrzkt            okay then there is g++-6 usage https://travis-ci.org/minetest/minetest/jobs/165354980 but i need to ensure it work by adding c++11 container explicitely to code :)
20:58 paramat          yeah
20:59 paramat          the new biomes mean that there's more chance player is a long way from any forest
20:59 paramat          personally i think that's good for exploration
21:00 Fixer            i really miss this: possibility to see rough map of a given seed before world creation
21:00 Fixer            not all seeds are nice
21:00 Fixer            some are boring
21:01 paramat          my latest core mapgen has a way to shrink the whole world down to a 3D map
21:01 paramat          very useful for devving
21:01 paramat          but could be used by players too
21:02 paramat          you can zoom to any scale
21:02 Fixer            lets say i want seed with island in ocean
21:03 Fixer            minecraft has nice configuration menu for world, ore settings, water level, here it is hard to adjust water level
21:03 Fixer            i want ocean world, that means i need to write my mod
21:03 Fixer            or arctic only world
21:03 sofar            water level can be tuned
21:05 Fixer            on my first try I increased water level up and made arctic only, but then I discovered underground trees :)
21:05 Fixer            underwater*
21:06 nrzkt            if that build pass https://travis-ci.org/minetest/minetest/jobs/165356862 we have working c++11 CI. and then i will squash my commits. Can you review a last time sfan5 please ? (except the std::unordered change which is just for testing)
21:07 sfan5            in a few mins, sure
21:10 nrzkt            nice build pass
21:11 paramat          yes changing water level screws up any non-mgv6 mapgen unless you reregister all biomes and decorations
21:11 paramat          better to use noise params to lower terrain
21:12 sfan5            nrzkt: see comments
21:12 sfan5            if you fix that you have my +1
21:12 nrzkt            fixed the two conditions :)
21:12 nrzkt            strange, unittests are not working with C++11
21:13 nrzkt            somes are unconditonally failing
21:13 sfan5            is it pcg random?
21:13 sfan5            or something with streams & floats?
21:13 nrzkt            [FAIL] testAllSettings - 1ms
21:13 nrzkt            [FAIL] testAtomicSemaphoreThread - 7ms
21:14 nrzkt            only those
21:14 paramat          it possible to write mods that create miniature 3D 'seed previews' of the core mapgens
21:14 sfan5            never seen those fail
21:15 nrzkt            i know but two builds failed exactly here
21:15 nrzkt            comparing const char* vs std::string
21:15 nrzkt            it's not good in c++11
21:16 nrzkt            on my pc cannot reproduce easily, coredump on floating point exception
21:16 sfan5            same here
21:17 nrzkt            systematic on the test after testPcgRandom
21:19 paramat          (will merge those 4 game PRs in a moment)
21:22 nrzkt            sfan5, in fact in debug mode i have the unit tests problem mentionned by travis and in release mode the floating point failure
21:24 nrzkt            sfan5, the problem is the std::unordered_map i used in a previous commit sfan5 for the settings
21:24 nrzkt            the generated string is different
21:25 sfan5            ah yes
21:26 sfan5            serialization gets non-deterministic with unordered maps
21:26 nrzkt            i don't know how fix this... except remove this test
21:26 sfan5            which one? (github code link please)
21:27 paramat          merging
21:27 nrzkt            sfan5, https://github.com/minetest/minetest/blob/master/src/unittest/test_settings.cpp#L200
21:28 nrzkt            this line
21:28 sfan5            hm yes that's hard
21:28 sfan5            comment it out for now i guess
21:29 nrzkt            sfan5, or ifndef c++11 like ?
21:29 nrzkt            to use it in old builds but not in c++11 ?
21:29 sfan5            that works too
21:31 paramat          merge complete
21:31 nrzkt            lol
21:32 nrzkt            testAtomicSemaphoreThread problem is only due to uninitalized variable and c++11 doesn't like it
21:32 nrzkt            random value in atomic :)
21:34 I_wish_I_was_a_l #3900
21:34 ShadowBot        https://github.com/minetest/minetest/issues/3900 -- Add a button for disabling all mods to world config. by red-001
21:34 I_wish_I_was_a_l #3900
21:34 ShadowBot        https://github.com/minetest/minetest/issues/3900 -- Add a button for disabling all mods to world config. by red-001
21:40 _Megaf           T4im: [22:29:52] <MinetestBot> [git] t4im -> minetest/minetest_game: Fix sapling placement handling of on_rightclick and CONTENT_IGNORE 32f7051 https://git.io/vPnSd (2016-10-05T22:22:15+01:00)
21:40 _Megaf           awesome!
21:41 nrzkt            sfan5, i found the problem i think
21:41 nrzkt            0 % 0
21:41 nrzkt            i think
21:43 nrzkt            very very strange
21:43 nrzkt            if (bound == 0)
21:43 nrzkt            return next();
21:43 nrzkt            but
21:43 nrzkt            i have this line after
21:43 nrzkt            std::cout << bound << " test2" << std::endl;
21:43 nrzkt            and it works
21:43 nrzkt            ...
21:43 nrzkt            and bound = 0
21:45 nrzkt            sfan5, i'm waiting for travis but did you agree with last version ?
22:26 _Megaf           Some of the latest commits broke minetestserver!! #4579
22:26 ShadowBot        https://github.com/minetest/minetest/issues/4579 -- SEG FAULT on entering anything on the curses terminal.
22:27 _Megaf           sfan5: nrzkt paramat ˆ
22:28 nrzkt            merging #4578
22:28 ShadowBot        https://github.com/minetest/minetest/issues/4578 -- Changes on travis build matrix by nerzhul