Time  Nick          Message
04:05 bigfoot547    Night guys!
05:40 nerzhul       merging #5921 which is fairly trivial in  ~10 mins
05:40 ShadowBot     https://github.com/minetest/minetest/issues/5921 -- Remove old network code by red-001
05:52 nerzhul       !tell paramat stop trying to block unrelated PR just because you want external contributor work for you to find a solution on problem you want to be resolved. CODE THEM.
05:52 ShadowBot     nerzhul: O.K.
08:42 celeron55     VanessaE: certainly not soon because the way i'm going to revisit farmap is that i'm taking everything else than farmap from the farmap PR and making a PR out of that
08:43 celeron55     (actually, i already did that and rebased it, but it's still buggy and missing the PR)
08:52 kilbith       https://lut.im/d9bQgT6GaD/5A1iC3cMNtgYmj1N.png
08:54 nerzhul       kilbith, is this animated ?
08:54 nerzhul       sounds to be a nice torch
08:54 kilbith       yes
08:54 kilbith       but notice the bloom
08:56 nerzhul       interesting
08:57 DS-minetest   is the bloom on whole torch or only at flame part?
08:58 kilbith       on the flame
09:01 kilbith       subtle effect on the snow too: https://lut.im/mCTfxV8oO9/7GTZ7IJzQlKXET8v.png
09:01 kilbith       and clouds
09:02 DS-minetest   nice!
09:03 DS-minetest   the only thing i can find about it is this: http://kidscode.com/en
09:05 kilbith       stalker ;)
09:08 Thomas-S      celeron55, could you maybe please take a look at my meta-set noddef PRs, as they are based on code originally written by you? (#5895, #5902, #5903) Thanks in advance for your efforts!
09:08 ShadowBot     https://github.com/minetest/minetest/issues/5895 -- Meta-set nodedef: Part 1 by Thomas--S
09:08 ShadowBot     https://github.com/minetest/minetest/issues/5902 -- Meta-set nodedef: Part 2 by Thomas--S
09:08 ShadowBot     https://github.com/minetest/minetest/issues/5903 -- Meta-set nodedef: Part 3 by Thomas--S
09:21 nerzhul       just a question Thomas-S why do you split it in 3 parts ?
09:21 Thomas-S      To allow the first parts to be merged earlier, so that too much rebases can be avoided.
09:22 kilbith       https://lut.im/oN4xs8afjG/YOyRmxBqGXvdWyYv.png
10:14 nerzhul       i'm starting to work on CSM blacklist flavours
10:21 sfan5         i don't think a feature blacklist is a good idea
10:21 sfan5         what if someone wants to use an ambience CSM mod? wouldn't work without get_node
10:22 red-NaN       ¯\_(ツ)_/¯
10:22 red-NaN       can't be helped
10:22 red-NaN       have to stop those dangerous cheaters
10:22 nerzhul       sfan5, it depend
10:23 nerzhul       if CSM is less stupid than SSM: play_sound_if_node for example
10:23 sfan5         no that's stupid and very limiting
10:23 red-NaN       nerzhul, maybe you could have the client have a builtin blacklist, if the server wants to restrict csm
10:23 Thomas-S      Why not giving nodes the ability to pretend to be another node? So that for example a iron ore node can pretend to be a default:stone node.
10:23 sfan5         it is possible to limit get_node without removing it entirely
10:23 nerzhul       sfan5, i think get_ndoe is not really the problem, but more find_node_near no ?
10:24 red-NaN       and make is possible for the server to request extra functions to be added or removed from the blacklist?
10:24 sfan5         well if you limit find_node_near you also need to limit get_node
10:24 nerzhul       having flavour is one of the required  features
10:24 nerzhul       it permit to blacklist a set of functions
10:24 nerzhul       after we can have more fine grain
10:24 nerzhul       it's not easy to handle
10:24 red-NaN       that way we can safely add new dangerous functions without having to worry about old servers
10:24 sfan5         outright blacklisting features such as find_node_near is a bad idea
10:24 red-NaN       btw I updated my player list PR
10:25 nerzhul       sfan5, i agree, but how do you want to handle that ?
10:25 nerzhul       replace all ore with no ore ?
10:25 nerzhul       (in CSM calls)
10:25 red-NaN       limit the range>
10:25 red-NaN       *?
10:25 nerzhul       we can limit range, limit API calls per timer
10:25 sfan5         e.g. only allow get_node (+ related funcs) withhin player interaction range and only for nodes that have a non-solid neighbor
10:26 sfan5         limiting api calls is stupid
10:26 nerzhul       it prevent scans, it's quite stupid like this cheating problem but we should find a solution
10:26 nerzhul       i started to prepare flavour transport https://github.com/nerzhul/minetest/commit/83ee5abec6a6e0729fe7db86097e93e4d500a078
10:27 nerzhul       we should find a consensus and release a 0.4.17 with that fix directly
10:27 sfan5         the problem is not *that* pressing
10:28 sfan5         also yes that's a good start but outright blacklisting essential functions is not good
10:28 nerzhul       allowing get_ndoe in player interaction area is very limiting for ambience, yo ucannot hear water which is a little bit far
10:28 nerzhul       sfan5, if not blacklisting maybe limiting then ?
10:28 nerzhul       :)
10:28 sfan5         yes limiting
10:28 nerzhul       s/blacklist/limit in my pr then
10:28 sfan5         it would be okay to blacklist e.g. chat message sending
10:28 sfan5         but not get_node
10:33 nerzhul       ok i renamed all to limits, it's less restrictive :p
10:34 nerzhul       i opened #5930 for this issue
10:34 ShadowBot     https://github.com/minetest/minetest/issues/5930 -- [WIP] [CSM] Add flavour limits controlled by server by nerzhul
10:34 nerzhul       sfan5, don't forget my 3 mergeable PR (C++11 part 3/4 + gitlab CI)
10:36 nerzhul       i will work on this limitation during the next days between 2 maintenance PRs :) it's not difficult to handle, the more difficult is to find the better way to limit that
10:40 red-NaN       I don't think chat message sending should be blacklisted
10:41 DS-minetest   +
10:41 DS-minetest   imo it would be enough to limit get_node_near
10:42 * DS-minetest doesn't like that csm should be made more limited
10:43 nerzhul       red-NaN, what usecase do you see ?
10:44 DS-minetest   nerzhul: sending eg. teleport commands to server
10:45 nerzhul       you mean a button sending commands ?
10:45 DS-minetest   yeah
10:45 nerzhul       the release process forgot client_lua_api.md...
10:46 nerzhul       DS-minetest, if server has such feature it's its responsibilty to not disable that no ?
10:46 nerzhul       the goal is to not disable/limit anything and let server owner choose which flavour enable
10:47 DS-minetest   but server owners can be so mean in such things >_<
10:48 nerzhul       yes i know, if they doesn't want their users to be happy, they will have less user and a better server will take users, seems lgtm no ? :)
10:52 nerzhul       i added basic handling for the two existing flavours in #5930, remaining issues are: configuring the flag server side & implement call limitations on node getters
10:52 ShadowBot     https://github.com/minetest/minetest/issues/5930 -- [WIP] [CSM] Add flavour limits controlled by server by nerzhul
10:54 sfan5         pushing minor texture fix (black -> transparent for unused part) to minetest_game
10:54 nerzhul       nice
11:03 red-NaN       nerzhul, sending commands, custom chat, etc
11:03 red-NaN       if we add a http api to csm then translations
11:05 red-NaN       I don't really see a downside to sending chat
11:05 red-NaN       expect maybe easier spamming but chat history already made that pretty simple
11:12 cx384         Why is the stack_max of bucket:bucket_empty defined as 99? https://github.com/minetest/minetest_game/blob/master/mods/bucket/init.lua#L115 (It is not possible to change the stack_max of this item by changing the default stack_max.)
11:13 sfan5         new windows build for minetestmapper: https://github.com/minetest/minetestmapper/releases/tag/20170606
11:14 red-NaN       custom chat is basically the sort of thing that seems to be a decent usecase for local csm mods
11:20 red-NaN       the worse it can do is make it easier for players to use colour codes
11:51 nerzhul       sfan5, nice
11:51 nerzhul       no need for python file heh, or move it to this project :)
11:52 nerzhul       red-NaN, for me custom chat is customizing chat appearance/colors and doesn't require send_chat_message API
11:52 nerzhul       but i agree it's interesting to have admin panel in CSM
11:53 nerzhul       worldedit with server side commands can be nice too
11:53 nerzhul       imagine a mass selector to define your area client side and this send worldedit messages to design the area, a paint for changing nodes, etc :p
11:54 red-NaN       I meant as in sending custom chat to the server
11:55 red-NaN       there is nothing currently preventing clients from adding colour codes to thier messages
11:55 nerzhul       exact but it's not read by non modded clients no ?
11:55 red-NaN       no clients can use colour codes in the same way IRC clients can
11:56 nerzhul       if a server xwants to forbid API and make player regular chat he can no ? it's its own resposnbility to make its players happy
11:56 nerzhul       use regular chat*
11:56 red-NaN       CSM just makes it easier to do, before you had to find a way to type the escape character
11:57 red-NaN       I say a good way to stop people from abusing colours would be to add a server-sided setting to removing colour codes from incomming messages
11:58 nerzhul       yeah i see but if a server admin wants to have regular chat not customized he can (with the downside that commands are not sent)
11:59 red-NaN       if we limit chat sending we should add something like #3856
11:59 ShadowBot     https://github.com/minetest/minetest/issues/3856 -- Keyboard shortcuts for chat commands by celeron55
12:05 VanessaE      celeron55: well that's one way to do it I guess :)
12:07 red-NaN       maybe a compromise would be to allow the server to set the rate limiting on messages sent by CSM?
12:07 red-NaN       that would make it useless for spamming
12:07 nerzhul       red-NaN, we can send a such thing but it's not compatible with my flavours except if i add an extra configuration parameter :p
12:08 red-NaN       well it might be best to get the opinion of some server owners first
12:09 nerzhul       yes :)
12:10 nerzhul       if it's important maybe the flavour can have a specicial rate limit send with flavours and used by global configuration
12:10 celeron55     VanessaE: it means you don't get what you wanted though
12:11 VanessaE      celeron55: I know.
12:12 VanessaE      one thing at a time I guess
12:15 nerzhul       merging #5912 #5906 & #5922 in ~20 mins (ty Zeno` )
12:15 ShadowBot     https://github.com/minetest/minetest/issues/5912 -- Main Menu: Add get_clientmodpath API by octacian
12:15 ShadowBot     https://github.com/minetest/minetest/issues/5906 -- C++11 patchset 3: remove Atomic/GenericAtomic and use std::atomic by nerzhul
12:15 ShadowBot     https://github.com/minetest/minetest/issues/5922 -- C++11 patchset 4: use std::(recursive)mutex directly & drop compat layer by nerzhul
12:24 nerzhul       currently we are talking about messages but our main concern is node cheat
12:25 red-NaN       ^
12:25 nerzhul       i added a flavour which enable limitations in API, now we need to define what is the limit for each call (find_node_near, get_node(or_nil)
12:25 nerzhul       range limit ?
12:25 red-NaN       I would suggest only limiting functions for getting large numbers of nodes
12:26 red-NaN       get_node is too slow to be useful for cheating
12:26 red-NaN       at least not in any useful way I can think of
12:26 Zeno`         well it might be useful for a dos
12:27 sfan5         wut
12:27 red-NaN       ?
12:27 sfan5         just because cheating is slow doesn't mean it's not cheating
12:27 sfan5         ratelimitings are useless for get_node and friends
12:32 nerzhul       Zeno`, get_node doesn't involve server in CSM it read local map and doesn't generate a mapblock fetch from server
12:34 nerzhul       erf 5922 requires rebase due to 5906 :) i rebase it
12:35 Zeno`         yep I realised that as soon as I made my comment ;)
12:41 nerzhul       waiting for ci on 5922 after rebasing
12:44 glut32        hi all
12:45 nerzhul       hi glut32
12:48 glut32        I would like to compile Minetest under MinGW/Linux. I installed MinGW on my Ubuntu but when I want to execute "./util/buildbot/buildwin32.sh ." I have compile error about c++11 std:thread... Do you know how to use "buildwin32.sh" ?
12:50 glut32        ps : I have the same errors in MinGW/Windows
12:51 numzero       does it say that that header requires C++11 to be enabled?
12:53 glut32        how to enable it on Linux? I know how to do it on Windows with CMake GUI
12:53 nerzhul       glut32, which ubuntu ?
12:53 glut32        I must to use -std=c++11
12:53 glut32        16.10
12:53 nerzhul       drop your cmake cache and rebuild i think
12:53 nerzhul       it didn't switched
12:53 numzero       you may use cmake -DCMAKE_CXX_FLAGS=-std=c++11
12:54 nerzhul       don't do this, cmake does it for you if you drop your cache
12:54 nerzhul       it's a solution but not the proper solution :p
12:54 nerzhul       glut32, did you update buildbot ?
12:54 sfan5         -std=c++11 is enabled by default isn't it?
12:54 nerzhul       sfan5 provided a recent mingw gcc 7 which works perfect on travis and on docker with all ubuntu version priori to 14.04
12:54 nerzhul       sfan5, yes
12:54 sfan5         the buildbot will build from a clean state anyway
12:55 sfan5         glut32: which version is your mingw installation?
12:55 nerzhul       (travis & gitlab proof it's working as intended)
12:55 glut32        I used the lastest with "apt install mingw-w64" this morning
12:56 nerzhul       use sfan5 build
12:56 sfan5         ubuntu or debian often have old versions
12:56 nerzhul       glut32, => https://github.com/nerzhul/minetest/blob/668de0e963016ae992aa5f92e29fcfad1e75b541/.gitlab-ci.yml
12:57 nerzhul       look at this part
12:57 nerzhul       https://github.com/nerzhul/minetest/blob/668de0e963016ae992aa5f92e29fcfad1e75b541/.gitlab-ci.yml#L236
12:57 sfan5         glut32: you can use these https://github.com/minetest/minetest/blob/master/util/travis/before_install.sh#L22 https://github.com/minetest/minetest/blob/master/util/travis/before_install.sh#L30
12:57 nerzhul       on my last link you have exactly the build process for windows
12:57 nerzhul       (used by Gitlab CI)
12:57 nerzhul       (it's on xenial, like you :p)
12:58 nerzhul       do the before_script in generic_win_template and execute script in build/win{32,64}]
13:01 nerzhul       the source script is the link sfan5 show you
13:05 red-NaN       #5933
13:05 ShadowBot     https://github.com/minetest/minetest/issues/5933 -- [CSM] Add function to get player privileges by red-001
13:06 glut32        sed -e "s|%PREFIX%|${WIN_ARCH}-w64-mingw32|" -e "s|%ROOTPATH%|/usr/${WIN_ARCH}-w64-mingw32|" < util/travis/toolchain_mingw.cmake.in > ${TOOLCHAIN_OUTPUT}
13:06 glut32        What is the values of PREFIX and ROOTPATH?
13:06 glut32        What are
13:07 glut32        https://github.com/nerzhul/minetest/blob/668de0e963016ae992aa5f92e29fcfad1e75b541/.gitlab-ci.yml#L246
13:10 nerzhul       it's a sed, it's keys in the source files
13:10 nerzhul       you should only set WIN_ARCH here (which is in jobs if you look later in the file)
13:10 glut32        I downloaded the mingw.7z archive
13:11 glut32        I must to extract it on /usr?
13:11 glut32        in "usr" ?
13:11 glut32        "/
13:11 glut32        "/usr"
13:12 nerzhul       as you want, but follow the build process shown in gitlab CI it extract in /usr and replace paths properly in cmake etc
13:12 glut32        ok
13:12 calcul0n      not recommanded to extract an archive in /usr
13:12 calcul0n      better use /usr/local or /opt
13:13 glut32        Where is the generated cmake file ? What is the name of this file?
13:13 nerzhul       calcul0n, yeah it's not optimal but it's done like this atm
13:14 calcul0n      i see
13:30 glut32        thanks nerzhul sfan5 calcul0n and the minetest team. it works !
13:36 Thomas-S      https://github.com/minetest/minetest/commit/599e13e95e81aadb959c9f3715aec9b425ede084 seems to have unintended side-effects related to clear_craft: When the technic mod tries to unregister the sandstone->sand recipe, the mesecons recipe for luacontrollers also gets unregistered.
13:37 Thomas-S      ^ sfan5, nerzhul, ShadowNinja
13:47 nerzhul       Thomas-S, don't tell me it's because before those id were static and now it's dynamic we unregister previous static node id
13:48 Thomas-S      I don't know what the exact reason is, but  I left everything the same, and it doesn't work with the mentioned commit. One commit earlier is fine.
13:52 nerzhul       can you open a bug report on gh ?
13:54 Thomas-S      I opened a revert PR instead: https://github.com/minetest/minetest/pull/5934
14:02 nerzhul       reverting is not the best solution it's better to know the cause
14:03 nerzhul       also you reverted all minetest master branch...
14:04 Thomas-S      So, I'd better close it again. Thanks for the explanation! (I used the Github revert commit button)
14:04 nerzhul       seems github revert button is shit
14:04 nerzhul       xD
14:05 Thomas-S      Here you are: https://github.com/minetest/minetest/issues/5935
14:06 nerzhul       ty
14:07 Thomas-S      I'll try to find more information, although I'm not familiar with this part of the code.
14:08 nerzhul       i think nobody except celeron55 is :p
14:10 Thomas-S      I found some more information: The first commit of #5891 is fine.
14:10 ShadowBot     https://github.com/minetest/minetest/issues/5891 -- Remove deprecated code segments by Thomas--S
14:11 Thomas-S      This here is the problem: https://github.com/minetest/minetest/pull/5891/commits/b7ee8147db8880bb9eafb756b38499f8e6f3593e
14:11 celeron55     wtf are you doing here https://github.com/minetest/minetest/commit/599e13e95e81aadb959c9f3715aec9b425ede084#diff-a3d43cc239efc733a8a046ff0a6bc91eL934
14:12 celeron55     if you stop supporting an old version, then throw an exception or something, don't just break stuff when loading
14:12 celeron55     i suppose that commit does similar things elsewhere too
14:14 celeron55     it's probably best to remove it as it will break map data on disk
14:15 celeron55     permanently breaking stuff on disk is generally kind of not wanted
14:15 celeron55     i mean, revert
14:16 Thomas-S      nerzhul, the removal of the   if(name == "node" || name == "NodeItem" || name == "MaterialItem3" || name == "craft" || name == "CraftItem")   block was the problem.
14:18 Thomas-S      I think that celeron55 is right: It's best to revert my commit and then someone else who knows the code better does it again (correctly).
14:18 nerzhul       ok i push the revert
14:19 Thomas-S      Thanks!
14:20 nerzhul       seems more complicated than intended
14:21 nerzhul       merging #5922 in ~10 mins
14:21 ShadowBot     https://github.com/minetest/minetest/issues/5922 -- C++11 patchset 4: use std::(recursive)mutex directly & drop compat layer by nerzhul
14:21 nerzhul       travis was so slow to compile heh
14:21 nerzhul       GCC is slow since GCC 5.2...
14:22 Thomas-S      I think that the comments are a bit misleading, too: They declare still needed parts as obsolete.
14:32 red-NaN       #373 seems to have returned
14:32 ShadowBot     https://github.com/minetest/minetest/issues/373 -- Copy-paste doesn't work on GNU/Linux, from another program to Minetest
14:32 nerzhul       irrlicht 1.8 unmaintained powa
14:45 red-NaN       #5936
14:45 ShadowBot     https://github.com/minetest/minetest/issues/5936 -- Fix typos/mistakes in the documentation for colour related functions. by red-001
14:51 nerzhul       sfan5, i think we can merge it :)
14:51 sfan5         yeah
14:51 nerzhul       if you are okay i will do it in ~10 mins :)
15:16 nerzhul       who uses all travis slots, i'm pretty sure it's red-NaN
15:18 red-NaN       decent chance that it is
15:28 Thomas-S      I think #5925 can be closed: Not a bug and duplicate of #5572
15:28 ShadowBot     https://github.com/minetest/minetest/issues/5925 -- nodebox connects_to issue
15:28 ShadowBot     https://github.com/minetest/minetest/issues/5572 -- Connected nodeboxes rendered incorrectly
17:26 Fixer         wtf with boat these days, occasional jerkiness in singleplay up to boat abrupt stopping
17:48 nerzhul       SAO code didn't change maybe due to mod ?
17:54 nerzhul       i'm pushing trivial fix on bump_version.sh and i fix client_lua_api with correct MT version
20:28 paramat       hi nerzhul, you're best qualified to judge #5908 , can i merge it?
20:28 ShadowBot     https://github.com/minetest/minetest/issues/5908 -- (Re)spawn players within mapgen edges by paramat
20:29 red-NaN       could someone review #5852?
20:29 ShadowBot     https://github.com/minetest/minetest/issues/5852 -- Improve the path select GUI by red-001
20:29 red-NaN       I updated #5924
20:29 ShadowBot     https://github.com/minetest/minetest/issues/5924 -- Have the server send the player list to the client by red-001
21:01 nerzhul       paramat, let me 30 mins to look at this and test it
21:01 nerzhul       i push a minor ifdef fix for #5937
21:01 ShadowBot     https://github.com/minetest/minetest/issues/5937 -- DragonflyBSD latest git fails to compile
21:02 paramat       ok thanks :]
21:03 paramat       i carefully tested it so you might be able to trust that
21:04 paramat       i know the subject more than anyone, it's more the inter-file code that i'm unsure of
21:09 sfan5         paramat: the ambience mod example is for cases where the server does not provide such a mod
21:09 sfan5         also coding "not being able to see through nodes with get_node" in a perfect way is too computationally intensive
21:18 paramat       yeah i know :]
21:38 DeadWave      hey everyone
21:39 DeadWave      are here any people online?
21:39 rubenwardy    -> #minetest
21:40 nerzhul       paramat, i reviewed your PR
21:40 DeadWave      can everyone help me with isPlayer() method?
21:41 nerzhul       ?
21:41 DeadWave      i can use isLocalPlayer() only
21:41 rubenwardy    DeadWave, due to some weirdness in the api, you should do this:
21:41 rubenwardy    entity:is_player and entity:is_player()
21:42 paramat       thanks
21:44 DeadWave      but
21:44 DeadWave      i can use isLocalPlayer() without entity
21:45 DeadWave      getPosition works roo
21:45 DeadWave      too*
21:46 DeadWave      ClientActiveObject *obj = objects[i].obj;                              obj->isPlayer()
21:46 DeadWave      this
21:46 DeadWave      std::vector<DistanceSortedActiveObject> objects;
21:47 DeadWave      ~_~
22:07 paramat       if this is for modding it should go to minetest channel
22:09 nerzhul       seems minetest APK reviews on play store are better than before we moved from 3.48/5 to 3.53/5
22:09 nerzhul       :p
22:10 nerzhul       (3544 notes)
22:22 sfan5         the android ui still leaves a lot to be desired though :P
22:22 nerzhul       yeah controls are not very good :(
22:25 Fixer         nerzhul: in 100 years it will approach 4.0
22:26 nerzhul       sfan5, i hope you can look at #5923 to make it merged and let windows build be automatic :p
22:26 ShadowBot     https://github.com/minetest/minetest/issues/5923 -- Implement GItlab CI daily builds for windows platform (32 & 64) by nerzhul