Minetest logo

IRC log for #minetest-dev, 2018-07-01

| Channels | #minetest-dev index | Today | | Google Search | Plaintext

All times shown according to UTC.

Time Nick Message
00:10 paramat joined #minetest-dev
00:14 twoelk left #minetest-dev
02:07 Sokomine joined #minetest-dev
04:40 red-001 joined #minetest-dev
05:23 ANAND joined #minetest-dev
06:18 entuland joined #minetest-dev
08:20 Krock joined #minetest-dev
08:48 Krock joined #minetest-dev
09:21 nerzhul hello, when a coredev has time, #7511 & #7513 are ready for a review and are fairly trivial PR (there are many lines but it's trivial changes)
09:21 ShadowBot https://github.com/minetest/minetest/issues/7511 -- C converter cleanup by nerzhul
09:21 ShadowBot https://github.com/minetest/minetest/issues/7513 -- Backport Android cleanup parts from #7123 by nerzhul
09:39 Krock nerzhul, the chat API needs to be extended so it can specify what kind of message it is. However, in #7357 that's out of scope. Could you please review it once more? It's a small PR
09:39 ShadowBot https://github.com/minetest/minetest/issues/7357 -- Make the server status message customizable by SmallJoker
09:39 nerzhul yes, don't hesistate to do a pr to add optional message typ
09:39 nerzhul type
09:40 Krock so the chat message type args for the API must be added before this PR in your opinion?
09:41 nerzhul okay for me for that pr
09:41 nerzhul add it after merging this as a cleanup :)
09:42 nerzhul for the gradle thing, only the too long line ?
09:42 Krock yes, otherwise LGTM. didn't test
09:45 nerzhul it's just gradle :p we are not so many to build with it, i will try to backport one by one parts of my huge PR + additions to make makefile disappear
09:45 nerzhul this will take many time, but i prefer to have changes added one by one reviewable than a by fucking pr :D
09:46 Krock regarding the long line: was this auto-generated?
09:47 nerzhul no
09:47 Fixer joined #minetest-dev
10:23 nerzhul merging #7513 (i updated the too long line & re-tested it works)
10:23 ShadowBot https://github.com/minetest/minetest/issues/7513 -- Backport Android cleanup parts from #7123 by nerzhul
10:24 Krock will merge #7357 and #7491 in 5 minutes
10:25 ShadowBot https://github.com/minetest/minetest/issues/7357 -- Make the server status message customizable by SmallJoker
10:25 ShadowBot https://github.com/minetest/minetest/issues/7491 -- Log deprecated Lua function calls by SmallJoker
10:25 Krock FYI: chat message API (Lua) is currently WIP.
10:28 nerzhul no problem :)
10:28 nerzhul yeah 2 more PR merged
10:28 nerzhul i just opened #7519 it's the communication part
10:28 ShadowBot https://github.com/minetest/minetest/issues/7519 -- Add client weather communication by nerzhul
10:31 Krock merging...
10:32 Krock done. we're at 111 PRs again
10:32 nerzhul we can prevent that if you review #7511
10:32 nerzhul :p
10:32 ShadowBot https://github.com/minetest/minetest/issues/7511 -- C converter cleanup by nerzhul
10:35 Krock lack of float range check in read_v3s16
10:35 nerzhul exact
10:35 nerzhul hmm no
10:35 nerzhul read_v3d
10:36 nerzhul no checks
10:36 nerzhul okay i missed the thing. i cannot remove check_v3s16
10:36 nerzhul i was pretty sure it was good but not
10:36 Krock yes, but you replaced check_v3s16 with read_v3s16
10:36 nerzhul i can be strict yes
10:37 nerzhul lua_tonumber return 0 then i can use the range everywhere
10:37 nerzhul it's exact it should not broke anything
10:37 nerzhul it's strange we don't check ranges in many places
10:42 nerzhul okay it's fixed
10:42 nerzhul the check_float_range should be a noop for the v3s16 because if types, but at least it's there
10:46 Krock range checks are only important when the data will be sent to the client
10:46 Krock well, inf/nan checks are also important to ensure our functions work correctly
10:56 nerzhul okay, tell me if it's okay or not and what to do in the other case
11:02 Gael-de-Sailly joined #minetest-dev
11:25 ensonic joined #minetest-dev
11:30 Fixer_ joined #minetest-dev
11:31 Krock surely needs testing. afk
11:36 nerzhul for me it's okay, but i let another coredev to reverify it with more tests if needed
11:47 nerzhul i push a trivial dead code cleanup (the breath event whihc doesn't occur anymore on client)
11:48 red-001 !tell I was also working on a interface for the new chat message protocol, how far did you progress on it?
11:48 ShadowBot red-001: O.K.
11:49 red-001 !tell Krock I was also working on a interface for the new chat message protocol, how far did you progress on it?
11:49 ShadowBot red-001: O.K.
11:49 red-001 sorry ShadowBot
11:50 red-001 nerzhul, the server doesn't seem to use the protocol correctly?
11:51 red-001 the sender field doesn't seem to be used
11:51 nerzhul the server Lua API is currently using high level chat API instead of low level for system messages
11:51 nerzhul and yes the sender field is not populated by mods
11:51 red-001 I don't think I have any lua mods installed?
11:51 nerzhul it should be used by client chat bridge on the server
11:51 nerzhul do we have a function to detect if a player is inside a node in the core ?
11:52 red-001 https://pastebin.com/raw/E3m8sFjj
11:52 red-001 getting that without any mods
11:52 nerzhul red-001, strange
11:52 nerzhul then yes the server side should not be complete
11:52 nerzhul but it should if we wanted to handle chat in CSM mods properly :)
11:53 red-001 also I would suggest name isn't ignored for anything other than "raw"
11:53 nerzhul it seems Krock works on the system messages, you can work on the client message bridge
11:53 red-001 system messages?
11:53 nerzhul name has sense when a player send something no ? why a name in system messages ?
11:53 nerzhul if you got good examples i'm not against
11:53 red-001 could tell you what system sent it
11:54 nerzhul okayer, it's issuer more than player then :)
11:54 nerzhul okay for me
11:56 red-001 for announce it makes even more sense
11:57 red-001 eg. a server might have a mod that lets admins sent messages of that type
11:57 Krock joined #minetest-dev
11:57 nerzhul okay
11:57 nerzhul go
11:59 Krock red-001, can open a PR
12:03 red-001 https://github.com/red-001/minetest/tree/new_chat_interface
12:04 red-001 Krock, ok I only got I worked on client so far and didn't really have much testing done
12:05 red-001 I guess I will just review your one instead
12:05 Krock only "EnumString es_ChatMessageType" currently conflict in our verisons
12:05 Krock and the message -> text rename
12:05 red-001 msg->message just feels wrong
12:06 nerzhul thanks Krock, now builtin has deprecated getvelocity xD
12:06 nerzhul it's logged
12:06 nerzhul it's nice, :)
12:06 Krock nerzhul, oh lol xD
12:07 Krock I'll use parts of your PR and give you credits (at least in the PR description), if that's ok
12:07 red-001 sure
12:09 Krock I wonder why std::wstring is used all over the place. requires way too many conversions
12:12 red-001 rewrite chat messages to use utf-8?
12:15 red-001 Server::SendChatMessage seems broken
12:18 red-001 and Server::handleChat still uses the old system
12:19 Krock yes, I'm on it
12:19 Krock the CSM API is still unstable, right?
12:19 Krock yes. great. Will break that thingy
12:23 Krock red-001, heh nice. Your CSM would allow to manipulate timestamps of messages
12:24 Krock not sure if that's a good idea :/
12:24 red-001 that was the idea
12:25 red-001 it's only the display function
12:25 red-001 not the sender, that's just a string
12:43 red-001 Krock, pushed a fix for that method
12:43 red-001 if you are still working on it
12:43 Krock yes I am
12:44 Krock but slowly. it's way too hot to program efficiently
12:45 red-001 https://pastebin.com/raw/MfvDE0tw
12:46 behalebabo joined #minetest-dev
13:15 nerzhul https://www.youtube.com/watch?v=6Xe09W76EOw
13:25 Krock CLion looks somewhat similar to kdevelop
13:26 nerzhul oh it's an ide :p
14:03 red-001 is there any reason to keep the weird restriction on letting lua access the client version?
14:05 nerzhul i don't know, it was a hmmmism to ensure we are always comapt and mod never manage that
14:22 Krock nerzhul, read_chat_message() requires reading table fields, the string utils and wstring readParam.. uh well, that's quite much to do there
14:23 nerzhul okay :)
14:23 Krock especially providing table reading functions for all types
14:24 red-001 how is testing going?
14:26 AndroBuilder joined #minetest-dev
14:26 Krock the client and server must both be newer or be at this commit to have a correct chat handling
14:26 Krock but so far it looks promising
14:28 red-001 could you push your code somewhere?
14:32 red-001 Krock, client and server?
14:32 Krock red-001, PR is ready, if you were asking for that
14:32 red-001 why do you need client-side changes?
14:32 Krock CSM
14:33 red-001 yeah but you don't need that for chat to work
14:33 red-001 only the server had broken handling of the new protocol
14:33 red-001 sending everything as raw
14:34 Krock ah yes, sure. I thought I had changed more client-side. So only the server needs an update
14:34 red-001 Krock, oh I see you opened a PR
14:34 red-001 reviewing
14:37 Krock thanks :)
14:38 red-001 Krock, wouldn't it be better to deprecate the old server callbacks and add a new one that uses a chat message table?
14:39 red-001 just so new mods can use the same data structure everywhere
14:40 Krock yes it would.. I just have no idea how to make that new function look nice
14:40 Krock on_chatmessage2()?
14:40 red-001 on_chat?
14:41 red-001 bit confusing...
14:41 red-001 maybe on_message?
14:41 Krock it must contain "chat" to avoid collisions with "modchannel"
14:41 Krock I mean, to not confuse modders
14:42 red-001 well on_chat might be ok
14:45 red-001 Krock, could a new api be added for sending a  chat message object?
14:46 Krock ChatMessageRef ? Possible, yes.
14:46 red-001 or better it seems the read function supports either a string or a message
14:46 red-001 could chat_send_* just use that?
14:47 red-001 no need for a message type param then since if someone wants to use it they can just pass the right table
14:47 red-001 and it should be fully backwards compatibly
14:47 Krock well, it takes way more effort to create the right table than adding another function arg
14:48 red-001 {text = "hello", type = "system"}
14:50 red-001 plus it will let chat mods correctly pass on the username
16:05 Krock game#2160
16:05 ShadowBot https://github.com/minetest/minetest_game/issues/2160 -- Replace deprecated function calls by SmallJoker
16:10 nerzhul fuck irrlicht devs forgot a setter in 1.8.X series on particle affector, for speed
16:11 nerzhul it's okay on android but not for us...
16:19 nerzhul okay it's ported in irrlicht_changes, easily :p
16:46 srifqi joined #minetest-dev
16:49 nerzhul https://www.youtube.com/watch?v=w0FeV6S8Zxs welcome to the wind :)
17:30 stujones11 joined #minetest-dev
17:33 stujones11 #7483 is now ready for review
17:33 ShadowBot https://github.com/minetest/minetest/issues/7483 -- Move touchscreen input handling to base GUIModalMenu class by stujones11
17:34 rubenwardy so, I'm thinking that I'd like to add busted's assert library to Minetest somehow
17:34 rubenwardy so it can be used for integration tests and also be consistent with busted
17:34 rubenwardy any objections to this?
17:35 rubenwardy it's an MIT library that I'd have to rip out and add to a folder in builtin, and it could be hard to maintain
17:35 rubenwardy well, it's mostly bug free
17:35 nerzhul if you want a real unittest library just use cppunit it's a c++ industry standard
17:35 rubenwardy no
17:35 rubenwardy that's not for Lua
17:35 rubenwardy I'd like to contact the author first and see if they're able to separate the runner (which depends on ldir) with the unit tests
17:36 rubenwardy busted is the standard for lua unit tests
17:36 rubenwardy you want to be able to write your lua tests in lua
17:36 rubenwardy and also use the same API as standalone testing
17:37 rubenwardy busted has a lot of extra stuff like mocking and interception
17:37 rubenwardy not sure how this will work without investigation :/
17:38 rubenwardy sorry for not being clear, Busted is a lua library and tool which is the most popular one for writing unit tests in Lua [citation needed]
17:42 nerzhul no problem, i understand the scope is not the same :p
17:43 rubenwardy it would also be nice to use a standard testing library, but oh well
17:43 rubenwardy at least there are tests
17:43 rubenwardy *C++ testing library
18:12 Cornelia joined #minetest-dev
18:45 YuGiOhJCJ joined #minetest-dev
19:11 ensonic joined #minetest-dev
19:15 paramat joined #minetest-dev
19:16 paramat this will likely have to wait until after game2160 but, anyone for autocruise? game#2140
19:17 ShadowBot https://github.com/minetest/minetest_game/issues/2140 -- Boats: Add cruise mode (autoforward) by paramat
19:17 paramat and, white beds (possibly temporary) game#2150
19:17 ShadowBot https://github.com/minetest/minetest_game/issues/2150 -- Beds: Add white simple and fancy beds by paramat
19:17 paramat will review game2160
19:22 paramat +1 will merge in 5 mins
19:22 paramat game#2160
19:22 ShadowBot https://github.com/minetest/minetest_game/issues/2160 -- Replace deprecated function calls by SmallJoker
19:43 paramat uh forgot again, merging
19:44 paramat done
20:32 shacharr joined #minetest-dev
20:37 entuland joined #minetest-dev
20:59 AntumD joined #minetest-dev
21:07 Antum joined #minetest-dev
21:20 Wuzzy joined #minetest-dev
21:55 paramat joined #minetest-dev
22:26 stujones11 left #minetest-dev
23:48 Player-2 joined #minetest-dev

| Channels | #minetest-dev index | Today | | Google Search | Plaintext