Time  Nick      Message
08:03 red-001   #5113
08:03 ShadowBot https://github.com/minetest/minetest/issues/5113 -- [CSM] Add function to get the name of the player and functions to colour chat. by red-001
08:18 kaeza     red-001, you can use "\x1b" or "\27" instead of string.char(0x1b)
08:20 kaeza     not sure if the highlighting code will go in or not
08:21 kaeza     if yes, you should use something like :find(get_player_name(), 1, true) to avoid possible unintended use of pattern matching
08:23 kaeza     also maybe get_local_player_name() ?
08:26 nore      <kaeza> red-001, you can use "\x1b" or "\27" instead of string.char(0x1b) <-- no, we had problems with bundled lua
08:26 kaeza     hm?
08:26 nore      (it looks like it doesn't support escape sequences and you have to use string.char, while luajit works fine)
08:26 kaeza     ah
08:26 kaeza     # lua5.1 -e 'print"\x41"' --> x41
08:27 kaeza     "\27" should work though
08:27 nore      hmm
08:27 nore      I didn't try that one indeed
09:54 nrzkt     sfan5: i propose to add mismatched-tags as a blocking error in travis, do you agree ?
09:57 paramat   i'll merge #4805 #5085 #5102 in a few hours, in case anyone wants to look at those
09:57 ShadowBot https://github.com/minetest/minetest/issues/4805 -- Block spam messages before calling on_chatmessage callbacks by rubenwardy
09:57 ShadowBot https://github.com/minetest/minetest/issues/5085 -- Make entity on_punch same signature and behaviour as player on_punch by sapier
09:57 ShadowBot https://github.com/minetest/minetest/issues/5102 -- Dungeons: Add nodebox stairs to sandstone dungeons by paramat
09:57 nrzkt     paramat i merge #4805 now
09:57 ShadowBot https://github.com/minetest/minetest/issues/4805 -- Block spam messages before calling on_chatmessage callbacks by rubenwardy
10:10 paramat   ok
10:11 sfan5     nrzkt: it makes sense for travis but i don't think including it for normal compilation is a good idea
10:12 nrzkt     sfan5: why not ? coders should know they are doing wrong things
10:13 sfan5     yes that's why it should be in travis
10:13 sfan5     but it shouldn't break the build in other cases
10:18 nrzkt     but i don't understand why PR owners should wait for travis to break their builds ? struct/class mismatch is a real error like overloaded-virtual error
10:18 nrzkt     and it's only applied to MT, not deps
10:19 sfan5     dunno, what do the other devs think?
10:20 paramat   no idea :]
10:25 nrzkt     Zeno`: ?
10:29 nrzkt     sfan5: does if (WIN32) permits to match windows builds over gcc ?
10:29 nrzkt     (in cmake)
10:29 nrzkt     oh yes in fact, i should read more carefuly
11:00 VanessaE  are there any plans to use the new color=<foo> to handle mt_game wool?
11:00 VanessaE  (I don't mean param2, just the static node coloring)
11:01 paramat   ah, not yet
11:01 Calinou   that limits texture pack possibilities, and may look worse
11:01 VanessaE  Calinou: nah, if the base texture is done well enough, it'll look fine, but ok
11:01 VanessaE  I've put static coloring into use all over homedecor, to decent effect
12:17 paramat   #5109 is now well tested, i tested with fire, cart and general node sounds
12:17 ShadowBot https://github.com/minetest/minetest/issues/5109 -- OpenAL sound: Use a simpler distance model by paramat
13:28 red-001   could someone take a look at #5113?
13:28 ShadowBot https://github.com/minetest/minetest/issues/5113 -- [CSM] Add function to get the name of the player and functions to colour chat. by red-001
13:28 nrzkt     red-001: can you split colour chat & name getter ?
13:29 nrzkt     in two PR
13:29 red-001   well I suppose I could
13:29 red-001   but the colour stuff is only copying code from one place to another
13:29 nrzkt     ok
13:29 red-001   so maybe two commits?
13:30 nrzkt     no let it then
13:33 red-001   I think it might be a good time to rework the chat protocol a bit, client sided mods would benefit from knowing what player sent what without parsing the chat message, there is an existing comment in the code about splitting the name of the sender and the message apart, as the chat backend needs to have them separated
13:34 nrzkt     yes, what player and what message type
13:34 nrzkt     i will work on a draft today i think
13:34 nrzkt     maybe also message date
13:34 red-001   yeah that could be useful
13:42 red-001   I think I might give adding server <-> client
13:42 red-001   scripting communication a shot
13:43 VanessaE  ok for that new param2 coloring, is there some way I'm not seeing, to restrict the effect of param2 to just one of the node's textures?
13:45 nrzkt     red-001: wait a little bit for that, please look at current checkbox in main PR before if you want to work on it
13:45 VanessaE  ah HAH, figured it out
13:47 VanessaE  (for the logs:  palette = foo, paramtype2 = "color", tiles = { bar, baz, { name = bleh, color = 0xffffffff } } )
13:47 paramat   this? https://github.com/minetest/minetest/pull/4986#issuecomment-273640622
13:48 VanessaE  precisely that
13:48 VanessaE  I just didn't quite understand it as worded in lua_api.txt, at first.
13:50 paramat   i need to fix something in dungeongen before adding stairs to sandstone dungeons, so am delaying merging that PR
13:51 nrzkt     red-001: hmmm old chat is very raw xD
13:51 nrzkt     we will need a new packet for this
13:52 red-001   and protocol bump?
13:52 nrzkt     we have a protocol bump version in master currently but yes
13:52 nrzkt     client side mode will required new client version, there is no problem for that
13:54 nrzkt     i will deprecate TOCLIENT_CHAT_MESSAGE and rename it to TOCLIENT_CHAT_MESSAGE_OLD and create a new packet for new protocol
14:11 lisac     sofar, are you there?
14:12 lisac     http://imgur.com/4tXRYu4
14:12 lisac     ^Minimal Development test with latest Minetest on A8 7600
14:12 lisac     Smooth lighting is on
14:14 lisac     Should I report it as a bug?
14:15 sfan5     probably
14:20 lisac     Okay then.
14:26 Fixer     lisac: you are using newest mt?
14:26 lisac     Fixer, yes, I compiled the latest MT a few minutes ago
14:26 lisac     5116
14:26 lisac     #5116
14:26 ShadowBot https://github.com/minetest/minetest/issues/5116 -- LIghting not working at all (everything fully bright)
14:26 lisac     ^
14:26 lisac     I just reported the bug
14:26 Fixer     lisac: it seems sofar also had this bug, but it was fixed in smooth lighting pr
14:27 Fixer     that was merged
14:27 lisac     Yes, with shaders disabled, MT works fine. With them enabled, everything is fully bright.
14:27 Fixer     lisac: aha, yes, this one
14:28 Fixer     lisac: please report it
14:28 lisac     I reported it lol
14:28 Fixer     lisac: post link to a report here
14:28 lisac     #5116
14:28 ShadowBot https://github.com/minetest/minetest/issues/5116 -- Lighting not working at all (everything fully bright)
14:28 lisac     I did. :P
14:39 red-001   has duckbrain been seen recently?
14:39 nrzkt     red-001: i started a draft for chat protocol, this needs a little more things client side and proper detection & lua api server side #5117
14:39 ShadowBot https://github.com/minetest/minetest/issues/5117 -- Chat protocol rewrite by nerzhul
14:40 nrzkt     the goal is to merge it before client side modding permitting to start cleint side api after merge
14:41 red-001   whats message_type for?
14:42 nrzkt     chat, announcement, error
14:42 nrzkt     for example
14:42 nrzkt     permitting mods or client to know what to do (especially messages color depending on the message type)
14:42 fwhcat    are you gonna make both-side api, clients functions/events and server functions/events ?
14:42 nrzkt     fwhcat: for chatting protocol enhancements ?
14:43 fwhcat    no no I mean generally nrzkt
14:43 red-001   nrzkt, thats a pretty good idea
14:43 nrzkt     the client side API is worked on #5088
14:43 ShadowBot https://github.com/minetest/minetest/issues/5088 -- Client side scripting/modding by nerzhul
14:43 nrzkt     but only useful calls client side are added, and some are purely client side, many are purely server side
14:44 nrzkt     omg the chatting handling when sending messages to client is just... bad
14:45 red-001   that's one way to put it
14:49 fwhcat    Is committing allowed directly or should I start with a pull-request if I wanna help?
14:54 red-001   fwhcat, you need to make a PR of course
14:55 red-001   only core devs are able to push to the repository anyway
15:01 Fixer     got crash, will make full clean rebuild, and check again, maybe smth interesting
15:01 Fixer     at least in min dev game
15:06 Fixer     in mindevgame: NodeResolver: failed to resolve node name 'mapgen_river_water_source'
15:06 nrzkt     fwhcat: make a PR on the minetest:client_side_modding branch if you want to help
15:07 nrzkt     instead of minetest:master
15:07 nrzkt     red-001: is not a coredev but he contributes :)
15:22 paramat   eh i thought river water was added to minimal
15:30 Fixer     nope I think
15:30 Fixer     i'm building debug build to try catch that crash
15:32 red-001   Fixer, that shouldn't crash the server....
15:32 red-001   are you sure it's not something else?
15:32 Fixer     because crash is due to other bug
15:32 Fixer     it is something else ofc
15:32 Fixer     and it crashes only sometimes
15:33 paramat   #5119 my recent dungeon commit will cause rare segfaults near sandstone dungeons. here's the fix
15:33 ShadowBot https://github.com/minetest/minetest/issues/5119 -- Dungeongen: Fix out-of-voxelmanip access segfault by paramat
15:33 red-001   what does the channel part of a packet define mean?
15:33 red-001   definition*
15:33 Fixer     paramat: i seen dungeons near me in mindevgame, yes, but they were cobblestone
15:33 Fixer     i will get a backtrace anyway
15:34 paramat   yeah no sandstone dungeons in minimal (i think)
15:36 paramat   yes
15:36 nrzkt     red-001: oh you are right, i forgot channel part
15:36 red-001   what does it do? is it documented anywhere?
15:39 nrzkt     where did you see it ?
15:39 paramat   indeed river water was not added to minimal, not needed anyway, those warnings are harmless
15:54 paramat   #5119 is tested, can anyone review? fixes occasional segfaults near sandstone dungeons
15:54 ShadowBot https://github.com/minetest/minetest/issues/5119 -- Dungeongen: Fix out-of-voxelmanip access segfault by paramat
15:57 red-001   can any number in lua be converted back and forward to a string without data lose?
16:00 nrzkt     in pure Lua or using C functions ?
16:13 red-001   c functions
16:14 Fixer     paramat: i think i have dungeon related crash
16:15 red-001   I'm working on server <-> client scripting interface
16:16 Fixer     paramat: what do you think? http://pastebin.com/raw/k00GtKvK
16:17 paramat   yes that's what my PR fixes
16:17 paramat   exact same bug i just experienced
16:17 Fixer     paramat: aha! but it happened with cobblestone dungeons too :)
16:18 Fixer     in mindev
16:18 paramat   weird
16:19 nrzkt     red-001: lua <=> C have strict types
16:19 nrzkt     check with lua_isnumber and then handle
16:19 nrzkt     lua_tonumber
16:19 paramat   ah it was a desert dungeon
16:19 Fixer     oh, crashed again :}
16:21 Fixer     gdb crash
16:23 Fixer     paramat: yeah, i see desert stone in backtrace
16:23 Fixer     #4  0x00000000005f0b8b in MapgenBasic::generateDungeons (this=0xc223a90, max_stone_y=-34, stone_type=MGSTONE_DESERT_STONE)
16:23 paramat   ps.Z = 64, just outside the voxelmanip that extends from -48 to 63
16:23 ShadowBot https://github.com/minetest/minetest/issues/4 -- cppcheck warnings
16:23 paramat   heh
16:24 paramat   ok i had better merge my fix then
16:24 Fixer     paramat: shorter backtrace: http://pastebin.com/raw/6PTJtmVC
16:24 nrzkt     #3
16:24 ShadowBot https://github.com/minetest/minetest/issues/3 -- Furnace segfault
16:24 nrzkt     oh a CPP furnace :p
16:24 Fixer     lol
16:26 Fixer     paramat: same error? no need for me to test a PR fix?
16:26 paramat   yes shorter backtrace is exact same error
16:27 Fixer     i will compile with fix anyway
16:27 paramat   and this is exactly the bug i experienced that is fixed by 5119
16:28 paramat   problem is this bug is fairly rare, only happens when a corridor hits the voxelmanip edge
16:29 Fixer     yet i had it on first try
16:29 paramat   that's your bugfixing magic
16:34 paramat   just need to improve the commit message
16:39 red-001   has duckbrain given up on getting his PR merged?
16:39 nrzkt     red-001: which PR ? CSM ? no, but he should work with us if he wakes up, but atm he seeems away since a long time
16:48 red-001   wow mt is 197k lines of code
16:49 nrzkt     no
16:49 nrzkt     120k
16:50 red-001   your right looks like openhub includes blank lines and comments in loc...
16:51 nrzkt     i'm using cloc
16:52 nrzkt     http://pastebin.com/bcr5SWNF
16:52 nrzkt     133k lines, remove lua 121k
17:03 Fixer     nrzkt: this may be interesting for you https://github.com/celeron55/minetest/tree/client_player_physics_script
17:03 Fixer     nrzkt: iirc client-side physics
17:04 paramat   there's a video of that
17:06 Fixer     https://www.youtube.com/watch?v=CeTjR5-IabY
17:06 red-001   is that the one with which celeron55 added proper flying?
17:06 Fixer     worst title ever
17:15 celeron55 i guess i could change the title...
17:16 celeron55 the point wasn't really to make a useful video, that's where the title comes from
17:17 Fixer     please change the title to something nicer
17:19 paramat   "Nice Minetest Thing"
17:20 nrzkt     ugly mapblock loading
17:21 celeron55 Fixer: well, you tell me
17:22 celeron55 i don't know what the video is useful for
17:22 Fixer     celeron55: Client Side Physics in Minetest (concept) / or smth
17:23 paramat   found it #4907
17:23 ShadowBot https://github.com/minetest/minetest/issues/4907 -- Add default:river_water_* to minimal game by Rogier-5
17:23 red-001   Did anyone know how to get minetest to build in jenkins
17:23 red-001   does*
17:23 celeron55 Fixer: ok, it's now something like that
17:24 Fixer     much better
18:15 red-001   is #5088 suppose to include server transferring the files to the client?
18:15 ShadowBot https://github.com/minetest/minetest/issues/5088 -- Client side scripting/modding by nerzhul
18:20 nrzkt     not atm, we only need to fix remaining points to make it mergeable with an interesting preview
18:20 nrzkt     it will permit more users to test it, review it and then we will think about the best thing for transfer instead having a big huge pr
18:21 red-001   ohh so it's going to load client sided mods of the players own harddrive?
18:24 red-001   it might be a good idea then to add mod loading now so that security can have a proper test case
18:25 red-001   that is a rogue mod
18:26 nrzkt     yeah i'm thinking about loading using current API but it's very server centric this needs some cleanups to make a generic API
19:26 sfan5     merging #5119 in 5 mins
19:26 ShadowBot https://github.com/minetest/minetest/issues/5119 -- Dungeongen: Fix out-of-voxelmanip access segfault by paramat
20:24 sofar     game#1527 seriously???
20:24 ShadowBot https://github.com/minetest/minetest_game/issues/1527 -- Creative: Optimize some functions by motherfscker
20:24 sofar     fucking politics and swearing in a single commit
20:26 sfan5     you're overreacting
20:27 sapier    well Imho that name is a little bit problematic too
20:27 sofar     maybe, but you're underreacting
20:27 sofar     we can just say NO to stuff like that
20:27 sapier    especially as that guy doen's seem to have any history at all
20:29 sapier    if he did use that name for some time it'd be difficult but this way assuming that name is for trolling only is quite plausible
20:29 sofar     look, I know, the person probably thought "hey, I want to make a commit and I don't have an account. uhhhhh, what should I choose for account name?"
20:30 sapier    ans as result you use a obviously offending name?
20:30 sofar     some people? shrug
20:31 sapier    Well I'd let that pull stay open a while maybe github solves that issue for us
20:32 sofar     github aint going to do anything
20:32 sofar     I'm not even going to report his profile either
20:33 sofar     he can have the account all he wants
20:33 sapier    btw  not sure if that commit is correct anyway
20:33 sofar     I'm suspicious
20:33 sofar     it's a nonsense optimization if you ask me
20:33 sapier    se at the bottom end I'm not 100% sure but I believe that simplificytion changes behaviour
20:35 sapier    and using max floor ceil as local variables is at least a matter of taste
20:35 sfan5     no it's actually faster
20:36 sfan5     but this is not performance critical code so it's nonsense here indeed
20:36 sapier    faster is not always better
20:36 sfan5     no?
20:36 sapier    faster code beeing less readable in non performance relevant parts imho is worse then (slightly) slower code
20:37 sfan5     changing math.max to max won't make it any less readable
20:37 sapier    in this very special example maybe not
20:38 sapier    but e.g. I use local functions called MAX and MIN in some of my mods doing slightly different things as default max and min
20:38 sapier    different error handling mostly
20:38 sofar     good code or not, I would refuse this PR irregardless
20:39 sapier    true I'll not give an approval for someone creating a new account with that name a day before creating the pull too
20:39 sofar     account active: 5hrs
20:40 sofar     not even
20:40 sofar     "Joined 6 hours ago"
20:42 sofar     afk, lunch
20:58 nore      sapier: I think the removal of the "update_creative_inventory" calls *do* change behaviour however
20:58 nore      s/do/does/
20:58 sapier    those too yes
20:59 sapier    to be honest I don't fully understand if the obvious changes in behaviour are relevant or not but it's for sure no "style/cleanup" commit
21:01 sapier    but it wasn't supposed to be a no change commit so that might even be ok ... yet the commit name is not acceptable
21:01 sapier    commiter name
21:02 red-001   could someone take a look at #5113 ?
21:02 ShadowBot https://github.com/minetest/minetest/issues/5113 -- [CSM] Add function to get the name of the player and functions to colour chat. by red-001
21:02 nore      could someone look at game#1479 please?
21:02 ShadowBot https://github.com/minetest/minetest_game/issues/1479 -- Add creative.is_enabled_for by rubenwardy
21:03 sapier    hmm nit sure about 5113 wouldn't it be better to implement things like playername analogue to modapi for server?
21:03 nore      (also, it 'unfortunately' should cause a merge conflict with game#1527)
21:03 ShadowBot https://github.com/minetest/minetest_game/issues/1527 -- Creative: Optimize some functions by motherfscker
21:07 red-001   ?
21:09 nore      red-001: #5113 looks good to me but untested
21:09 ShadowBot https://github.com/minetest/minetest/issues/5113 -- [CSM] Add function to get the name of the player and functions to colour chat. by red-001
21:52 sfan5     sofar: "please close this and lock conversation." what?
21:54 sofar     he's not going to close it, and before you know it the SJW's are going to harass everyone including him
21:55 sfan5     lol no
21:55 sofar     frankly I'm disappointed you guys didn't just say "please submit from an account with a reasonable username and we'll discuss that"
21:56 sfan5     you are correct that his name is not the most appropriate, however i don't think the contribution should be flat out rejected
21:57 sofar     so tell me how you're going to merge it - plainly falsify the commit author?
21:57 sofar     I'm also not saying it should be rejected
21:58 sofar     what part of
21:58 sofar     "please submit from an account with a reasonable username and we'll discuss that"
21:58 sofar     is not clear?
21:58 sofar     (or even impolite)
21:58 sofar     instead you game him credibility by ignoring the problem and discussing the technical issues first
21:59 sofar     you don't tell someone is a troll and then feed them
21:59 sofar     sure makes me look like an idiot for sticking up to some level of quality
21:59 sfan5     i'm not saying that he is
22:00 sfan5     and IMO the name issue is secondary to the code changes itself
22:01 sofar     if someone really cares about the project, making a new account would be immaterial
22:01 sofar     so he obviously cares about the account name
22:01 sofar     so obviously he is happy to be the troll
22:01 sfan5     you're assuming malice
22:01 sfan5     instead of carelessness
22:02 nore      sofar: well, I told him both the issues *and* that I wouldn't allow a merge if he doesn't change the name
22:02 sofar     I assumed stupidity at first, and perhaps I overreacted, but after his last reply, I no longer assume stupidity, and I know he's actually trolling
22:02 sofar     he has no intention of making a new account
22:02 sofar     so, clearly, it was not an accident
22:02 sofar     he made it to troll
22:03 nore      but since I saw the last comment, I have to agree with you
22:03 nore      he cares about the username
22:03 celeron55 yeah please assume people are dumb, not malicious
22:03 nore      so, no merge
22:03 sofar     so again, please, consider closing and locking the topic
22:03 sfan5     how does locking the discussion help?
22:03 sofar     once trolling has been established, the only recourse to prevent escalation is locking the topic
22:04 sofar     perhaps the risk isn't that high
22:04 sofar     but at this point the conversation isn't overly heated yet
22:04 sofar     count your blessings and make sure it won't?
22:05 sfan5     yes
22:05 sofar     goddammit man, I teach minetest stuff at a middle school myself at the moment
22:06 sofar     leave the fucking politics out of it
22:06 * sofar   out for coffee
22:06 celeron55 that username and avatar is, like, actually ridiculous though
22:07 celeron55 it's probably one of your students
22:07 sfan5     i really wonder why politics matter here
22:07 celeron55 8)
22:07 nore      well, I closed the conversation anyway, and asked him to reopen with a better username if he cares about it; I haven't locked (and I can't), but I don't think it's necessary now
22:07 sfan5     github avatars are not bound to git history
22:07 sofar     celeron55: nope, the kids are 12-14 and can barely write a for loop
22:08 sofar     celeron55: but, that did make me chuckle
22:08 sfan5     oh and his commit email doesn't even match the github one so the avatar will not show up anywhere
22:11 celeron55 i'm fine with the avatar and others should also be, see what sfan5 said
22:13 celeron55 the way to make politics matter the least is to not react to politics
22:13 celeron55 on the other hand if you selectively react to politics sometimes, you're being even more political
22:14 sofar     note in the thread I never commented or mentioned politics
22:14 sofar     only his username
22:14 sofar     tbh, if you want to protect your OSS project you shouldn't allow obviously fake email addresses either
22:15 sofar     there's no way to validate that the contribution is actually legally sound
22:15 sofar     (yes, I know, it's derivative of gpl2 etc. but there are still circumstances that could spell out legal problems for users)
22:15 celeron55 i don't think we have a habit of allowing obviously fake email addresses but maybe it should be written as a rule
22:16 sofar     I also deal with this type of stuff (code authenticity, license, origin) at work on a daily basis, so perhaps I'm much more cautious than others on the project
22:23 celeron55 it's probably impossible to find out the actual origin and authenticity of all of minetest's codebase but i guess it's legally kind of important to still try to do a reasonable job at maintaining it
22:23 celeron55 otherwise it's probably considered just reckless
22:25 celeron55 of course, all of that matters on who is considering it to be anything